Arthroplasties for hip fracture in adults

Sharon R Lewis; Richard Macey; Martyn J Parker; Jonathan A Cook; Xavier L Griffin

doi:10.1002/14651858.CD013410.pub2

Artroplastias para la fractura de cadera en adultos

Declaraciones de intereses de los autores

Versión publicada: 14 febrero 2022 Historial de versiones

https://doi.org/10.1002/14651858.CD013410.pub2

Contraer todo Desplegar todo

Resumen

disponible en

Antecedentes

Las fracturas de cadera son un importante problema sanitario que supone un reto y una carga para las personas y los sistemas sanitarios. El número de fracturas de cadera en el mundo está en aumento. La mayoría de estas fracturas se tratan quirúrgicamente. Esta revisión evalúa la evidencia acerca de los tipos de artroplastia: las hemiartroplastias (HA), que reemplazan parte de la articulación de la cadera; y las artroplastias totales de cadera (ATC), que la reemplazan en su totalidad.

Objetivos

Determinar los efectos de diferentes diseños, articulaciones y técnicas de fijación de las artroplastias para el tratamiento de las fracturas de cadera en adultos.

Métodos de búsqueda

En julio de 2020 se realizaron búsquedas en CENTRAL, MEDLINE, Embase, otras siete bases de datos y un registro de ensayos.

Criterios de selección

Se incluyeron los ensayos controlados aleatorizados (ECA) y cuasialeatorizados que compararon diferentes artroplastias para el tratamiento de las fracturas de cadera intracapsulares por fragilidad en adultos mayores. Se incluyeron ATC y HA insertadas con o sin cemento, y comparaciones entre diferentes articulaciones, tamaños y tipos de prótesis. Se excluyeron los estudios de personas con fracturas de cadera con patologías específicas a excepción de la osteoporosis y con fractura de cadera resultantes de traumatismos de alto impacto.

Obtención y análisis de los datos

Se utilizaron los procedimientos metodológicos estándar previstos por Cochrane. Se recopilaron datos de siete desenlaces: actividades cotidianas, estado funcional, calidad de vida relacionada con la salud, movilidad (todos ellos tempranos: dentro de los cuatro meses siguientes a la cirugía), mortalidad temprana y a los 12 meses de la cirugía, delirio confusional y vuelta al quirófano no planificada al final del seguimiento.

Resultados principales

Se incluyeron 58 estudios (50 ECA, ocho ensayos controlados cuasialeatorizados) con 10 654 participantes con 10 662 fracturas. Todos los estudios informaron sobre fracturas intracapsulares, excepto un estudio de fracturas extracapsulares. La media de edad de los participantes en los estudios varió entre los 63 y los 87 años; el 71% fueron mujeres.

A continuación, se proporcionan los resultados de tres comparaciones que representan el conjunto de evidencia más importante de la revisión. También se han realizado otras comparaciones, pero con muchos menos participantes.

Todos los estudios tenían riesgos inciertos de sesgo en al menos un dominio y tenían un alto riesgo de sesgo de detección. Se disminuyó la certeza de muchos desenlaces por la imprecisión, y por los riesgos de sesgo cuando el análisis de sensibilidad indicó que el sesgo a veces influía en el tamaño o la dirección de la estimación del efecto.

HA: cementada versus no cementada (17 estudios, 3644 participantes)

Hubo evidencia de certeza moderada de un efecto beneficioso con la HA cementada consistente con diferencias clínicamente pequeñas a grandes en la calidad de vida relacionada con la salud (CdVRS) (diferencia de medias estandarizada [DME] 0,20; IC del 95%: 0,07 a 0,34; tres estudios, 1122 participantes), y en la reducción del riesgo de mortalidad a los 12 meses (RR 0,86; IC del 95%: 0,78 a 0,96; 15 estudios, 3727 participantes). Se encontró evidencia de certeza moderada de poca o ninguna diferencia en el desempeño de las actividades cotidianas (AC) (DME ‐0,03; IC del 95%: ‐0,21 a 0,16; cuatro estudios, 1275 participantes), y la movilidad autónoma (RR 1,04; IC del 95%: 0,95 a 1,14; tres estudios, 980 participantes). Se encontró evidencia de certeza baja de poca o ninguna diferencia en el delirio confusional (RR 1,06; IC del 95%: 0,55 a 2,06; dos estudios, 800 participantes), la mortalidad temprana (RR 0,95; IC del 95%: 0,80 a 1,13; 12 estudios, 3136 participantes) o el regreso no planificado al quirófano (RR 0,70; IC del 95%: 0,45 a 1,10; seis estudios, 2336 participantes). Para el estado funcional, hubo evidencia de certeza muy baja que no mostró diferencias clínicamente importantes.

Los riesgos de la mayoría de los eventos adversos fueron similares. Sin embargo, las HA cementadas dieron lugar a menos fracturas periprotésicas intraoperatorias (RR 0,20; IC del 95%: 0,08 a 0,46; siete estudios, 1669 participantes) y posoperatorias (RR 0,29; IC del 95%: 0,14 a 0,57; seis estudios, 2819 participantes), pero tuvieron un mayor riesgo de embolia pulmonar (RR 3,56; IC del 95%: 1,26 a 10,11; seis estudios, 2499 participantes).

HA bipolar versus HA unipolar (13 estudios, 1499 participantes)

Se encontró evidencia de certeza baja de poca o ninguna diferencia entre las HA bipolares y unipolares en la mortalidad temprana (RR 0,94; IC del 95%: 0,54 a 1,64; cuatro estudios, 573 participantes) y en la mortalidad a los 12 meses (RR 1,17; IC del 95%: 0,89 a 1,53; ocho estudios, 839 participantes). No hay seguridad acerca del efecto en el delirio confusional, la CdVRS ni el regreso no planificado al quirófano, que indicaron poca o ninguna diferencia entre las articulaciones, porque la certeza de la evidencia fue muy baja. Ningún estudio informó sobre el desempeño temprano en las AC, el estado funcional ni la movilidad.

El riesgo global de eventos adversos fue similar. El riesgo absoluto de luxación fue bajo (aproximadamente 1,6%) y no hubo evidencia de diferencias entre los tratamientos.

ATC versus HA (17 estudios, 3232 participantes)

La diferencia en el riesgo de mortalidad a los 12 meses fue consistente con efectos beneficiosos y perjudiciales clínicamente relevantes (RR 1,00; IC del 95%: 0,83 a 1,22; 11 estudios, 2667 participantes; evidencia de certeza moderada). No hubo evidencia de una diferencia en el regreso no planificado al quirófano, pero esta estimación del efecto incluye efectos beneficiosos clínicamente relevantes de la ATC (RR 0,63; IC del 95%: 0,37 a 1,07; favorece la ATC; diez estudios, 2594 participantes; evidencia de certeza baja). Se encontró evidencia de certeza baja de poca o ninguna diferencia entre la ATC y la HA en cuanto al delirio (RR 1,41; IC del 95%: 0,60 a 3,33; dos estudios, 357 participantes), y la movilidad (DM ‐0,40; IC del 95%: ‐0,96 a 0,16, a favor de la ATC; un estudio, 83 participantes). No hay seguridad acerca del efecto en el estado funcional temprano, las AC, la CdVRS ni la mortalidad, que indicaron poca o ninguna diferencia entre las intervenciones, porque la certeza de la evidencia fue muy baja.

Los riesgos generales de eventos adversos fueron similares. Hubo un mayor riesgo de luxación con la ATC (RR 1,96; IC del 95%: 1,17 a 3,27; 12 estudios, 2719 participantes) y no hubo evidencia de una diferencia en la infección profunda.

Conclusiones de los autores

En las personas que se someten a una HA por fractura de cadera intracapsular, es probable que una prótesis cementada produzca un mejor desenlace global, especialmente en términos de CdVRS y mortalidad. No hay evidencia que indique que una HA bipolar sea superior a una prótesis unipolar. Es probable que cualquier efecto beneficioso de la ATC en comparación con la hemiartroplastia sea pequeño y no sea clínicamente apreciable. Se anima a los investigadores a que se centren en los implantes alternativos en la práctica clínica actual, como los cojinetes de doble movilidad, sobre los que la evidencia disponible es limitada.

PICO

Population

Intervention

Comparison

Outcome

El uso y la enseñanza del modelo PICO están muy extendidos en el ámbito de la atención sanitaria basada en la evidencia para formular preguntas y estrategias de búsqueda y para caracterizar estudios o metanálisis clínicos. PICO son las siglas en inglés de cuatro posibles componentes de una pregunta de investigación: paciente, población o problema; intervención; comparación; desenlace (outcome).

Para saber más sobre el uso del modelo PICO, puede consultar el Manual Cochrane.

Resumen en términos sencillos

disponible en

Cirugía de prótesis de cadera en adultos

Esta revisión evaluó la evidencia de los ensayos controlados aleatorizados (ECA) y cuasialeatorizados, sobre los efectos beneficiosos y perjudiciales de los diferentes tipos de prótesis de cadera utilizados para tratar la fractura de cadera en adultos.

Antecedentes

Una fractura de cadera es una rotura en la parte superior del hueso de la pierna. Este tipo de fracturas son frecuentes en los adultos de edad avanzada cuyos huesos pueden ser frágiles debido a una enfermedad llamada osteoporosis. Un método de tratamiento consiste en sustituir la cadera rota por una artificial. Esto puede hacerse mediante una hemiartroplastia (HA), que sustituye una parte de la articulación de la cadera (la parte esférica de la articulación). Estos reemplazos pueden ser unipolares (una sola articulación artificial), o bipolares (con una articulación adicional dentro de la HA). Como alternativa, la cirugía puede reemplazar toda la articulación de la cadera, que también incluye la cavidad en la que se asienta la parte esférica de la articulación de la cadera, que constituye una artroplastia total de cadera (ATC). Ambas articulaciones artificiales se pueden fijar en su lugar con o sin cemento óseo.

Fecha de la búsqueda

Se buscaron ECA (estudios clínicos en los que las personas se asignan al azar a uno de dos o más grupos de tratamiento), y ensayos controlados cuasialeatorizados (en los que las personas se asignan a los grupos por un método que no es aleatorio, como la fecha de nacimiento o el número de historia clínica), publicados hasta el 6 de julio de 2020.

Características de los estudios

Se incluyeron 58 estudios con 10 654 adultos con 10 662 fracturas de cadera. Los participantes de los estudios tenían entre 63 y 87 años, y el 71% eran mujeres, lo que es habitual en las personas que sufren este tipo de fractura de cadera.

Resultados clave

HA cementadas comparadas con HA no cementadas (17 estudios, 3644 participantes)

Se encontró que las HA cementadas mejoran la calidad de vida relacionada con la salud (CdVRS) y reducen el riesgo de muerte a los 12 meses de la cirugía. La magnitud de estos efectos beneficiosos varió entre un efecto pequeño y uno grande. Es posible que haya poca o ninguna diferencia entre los tratamientos en cuanto a la capacidad de usar la cadera (estado funcional), pero esta evidencia era muy incierta. El hecho de que la HA esté o no cementada probablemente suponga una diferencia mínima o nula en el desempeño de las actividades cotidianas (AC) o en la capacidad de caminar de forma autónoma, ya que muchas personas presentan confusión después de la cirugía (delirio confusional), mueren en los cuatro meses siguientes a la cirugía o necesitan ser operados de nuevo. La mayoría de los riesgos de complicaciones fueron similares, pero se observó que algunos riesgos relacionados directamente con la cirugía de prótesis de cadera (como que se produzca una rotura durante la cirugía) aumentaron con las HA no cementadas.

HA bipolares comparadas con HA unipolares (13 estudios, 1499 participantes)

El tipo de HA probablemente no influye en la cantidad de personas que mueren en los cuatro meses o hasta los 12 meses posteriores a la cirugía, y puede influir poco o nada en la necesidad de una cirugía adicional. Ningún estudio informó sobre las AC ni el estado funcional a los cuatro meses. La evidencia fue muy incierta en cuanto a si el uso de una HA bipolar o unipolar supone alguna diferencia en el delirio o la CdVRS a los cuatro meses de la cirugía. Una vez más, los riesgos de complicaciones fueron similares, y no se encontró evidencia de una diferencia en el riesgo de luxación de la cadera.

ATC en comparación con HA (17 estudios, 3232 participantes)

No está claro si las AC, el estado funcional, la confusión, la movilidad o las muertes dentro de los cuatro meses o hasta los 12 meses después de la cirugía son diferentes entre estos tratamientos. La evidencia no mostró una diferencia en el riesgo de cirugía adicional, pero no fue posible excluir la posibilidad de un efecto beneficioso importante de la ATC. Aunque el riesgo de la mayoría de las complicaciones fue similar, la luxación de cadera aumenta con la ATC.

Certeza de la evidencia

La evidencia de muchas de las comparaciones se basa en solo unos pocos participantes, y muchos estudios utilizaron métodos que podrían no ser fiables. La mayoría de la evidencia de las AC, el estado funcional, la CdVRS y la marcha autónoma fue de certeza baja y muy baja, lo que significa que no se tiene confianza en los resultados. Se tiene una confianza limitada o moderada en los otros resultados de esta revisión.

Conclusiones

En las personas que tienen una HA, es probable que una prótesis cementada produzca un mejor desenlace en general que una prótesis no cementada. No hay evidencia que indique que una HA bipolar conduzca a desenlaces diferentes de una HA unipolar. Las diferencias entre un reemplazo total de cadera y un reemplazo parcial son pequeñas y podrían no ser clínicamente importantes.

Authors' conclusions

Implications for practice

For people undergoing hemiarthroplasty for intracapsular hip fracture, it is likely that a cemented prosthesis will yield an improved global outcome, particularly in terms of clinically appreciable improvements in HRQoL and mortality. For every 26 people treated with a cemented hemiarthroplasty, one more person will be alive at 12 months following surgery.

Currently, there is insufficient evidence to determine whether a bipolar hemiarthroplasty yields different outcomes compared to a unipolar prosthesis. Both are appropriate treatments for people with intracapsular hip fracture.

Any benefit of THA compared with hemiarthroplasty is likely to be small and not clinically appreciable.

Implications for research

Considerable research resources have been and are being committed to this field; we identified seven ongoing studies that may contribute data in future review updates. It is unlikely that future research will importantly alter our inferences about the relative clinical effectiveness of cemented and uncemented HAs, which now include data from a large multicentre study (Fernandez 2022). There is a relative paucity of evidence available from generally small studies for the comparison of bipolar and unipolar hemiarthroplasty. The estimates of any difference between total hip arthroplasty and hemiarthroplasty for some of the critical outcomes are imprecise. However, available data provide little to suggest that any effect is likely to be clinically meaningful. This is consistent with the findings of the large, international HEALTH 2019 study, and suggests that repeating such a study may not yield high‐value information.

We therefore encourage investigators interested in these comparisons to focus on conducting studies of alternative implant designs ‐ such as dual mobility bearings ‐ that are being incorporated widely into clinical practice, with scant evidence to support their use. We encourage investigators to address the limitations in the quality of the evidence in the field through better study design and clear reporting about methods of randomisation and allocation concealment, as well as attempting to minimise attrition for participant‐reported outcomes. We raise the awareness amongst investigators of the core outcome set for hip fracture that should be included in every RCT in hip fracture (Haywood 2014). To date, few studies have considered patient‐relevant outcomes, such as performance of activities of daily living, health‐related quality of life, mobility, or delirium.

Given the recommendations in Haywood 2014, we recommend that future studies are large enough to detect differences in HRQoL. Having reviewed the included studies, we estimate that the standard deviation for EQ‐5D at four months' post‐diagnosis is approximately 0.3. Assuming a minimal clinically important difference of 0.07 (Walters 2005), and an observed attrition in the included studies approaching 40%, we recommend future samples of not less than 1000 participants in order to yield sufficiently precise estimates.

Summary of findings

Open in table viewer

Summary of findings 1. Cemented versus uncemented hemiarthroplasty for hip fracture in adults

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Cemented versus uncemented hemiarthroplasty for hip fracture in adults
Patient or population: adults with displaced and undisplaced hip fractures; included studies were for intracapsular fractures, except for one study of extracapsular fractures Setting: hospitals; included studies were conducted in China, Croatia, Denmark, Italy, New Zealand, Norway, Pakistan, Slovenia, Sweden, the UK and USA Intervention: HA fixed with cement (included studies which used unipolar or bipolar articulations) Comparison: HA fixed without cement (included studies which used unipolar or bipolar articulations. Designs of HA in 6 studies were first‐generation, and in 2 studies were unknown. We categorised them as first‐generation.)
Outcomes	Risk with uncemented HA	Risk with cemented HA	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Activities in daily living, early (within 4 months): using GARS (range from 18 to 72), a social dependency scale (range of scores 1 to 9); lower values in these scales indicate more independence. Also using OARS‐IADL (range from 0 to 14) and a 5‐point Likert scale derived from EQ‐5D; higher values in these scales indicate more independence Follow‐up: time points in the included studies were at 3 months and 4 months	The mean GARS score in the uncemented group was 45.7. The mean social mobility scale score in the uncemented group was 4.6. The mean OARS‐IADL score in the uncemented group was 3.7. The mean Likert score in the uncemented group was 3.15.	SMD 0.03 lower (0.21 lower to 0.16 higher)	‐	1275 (4 studies)	⊕⊕⊕⊝ moderate^a	This effect did not indicate a clinically important difference, based on a 'rule of thumb' of: 0.2 for a small difference, 0.5 for a medium difference, and 0.8 for a large difference.
Delirium (end of follow‐up) Follow‐up: time points in the included studies were at 12 months and 5 years	Study population		RR 1.06 (0.55 to 2.06)	800 (2 studies)	⊕⊕⊝⊝ low^c
	40 per 1000^b	42 per 1000 (22 to 82)	RR 1.06 (0.55 to 2.06)	800 (2 studies)	⊕⊕⊝⊝ low^c
Functional status, early (within 4 months): using HHS (range from 0 to 100); higher values indicate better function Follow‐up: time points in the included studies were at 6 weeks and 3 months	The mean HHS scores in the uncemented groups ranged from 62.53 to 72.1.	MD 3.38 higher (0.05 higher to 6.70 higher)	‐	416 (3 studies)	⊕⊝⊝⊝ very low^d	This effect did not indicate a clinically important improvement (based on a MCID of 15.9 to 18 points). In addition, data were available in 1 study with extracapsular fractures which showed improvement with cemented HAs (MD 14.70, 95% CI 11.78 to 17.62; 85 participants). We noted that the CI in this effect may indicate a clinically important improvement with cemented HAs in extracapsular fractures (based on a MCID of 15.9 to 18 points).
HRQoL, early (within 4 months): using EQ‐5D (range 0 to 1), and SF‐12 (range 0 to 100); higher values indicate better quality of life. Follow‐up: time points in the included studies were at 3 months and 4 months	The mean EQ‐5D score in the uncemented group ranged from 0.31 to0.58. The mean SF‐12 score in the uncemented group was 33.8.	SMD 0.20 higher (0.02 higher to 0.10 higher)	‐	1122 (3 studies)	⊕⊕⊕⊝ moderate^a	The difference between fixation techniques was compatible with no effect or a clinically important benefit of cemented HAs based on a MCID for EQ‐5D of 0.07.
Mobility, early (within 4 months): able to walk outdoors using no more than 1 walking aid. Follow‐up: time points in the included studies were at 3 months and 4 months	Study population		RR 1.04 (0.95 to 1.14)	980 (3 studies)	⊕⊕⊕⊝ moderate^a
	354 per 1000^b	369 per 1000 (227 to 404)	RR 1.04 (0.95 to 1.14)	980 (3 studies)	⊕⊕⊕⊝ moderate^a
Mortality, early (within 4 months) Follow‐up: time points in the included studies were at hospital discharge, 7 days, 6 weeks, 3 months and 4 months	Study population		RR 95 (0.80 to 1.13)	3136 (12 studies)	⊕⊕⊝⊝ low^e
	143 per 1000^b	136 per 1000 (114 to 162)	RR 95 (0.80 to 1.13)	3136 (12 studies)	⊕⊕⊝⊝ low^e
Mortality at 12 months Follow‐up: time points in the included studies were at 12 months, 16 months, 18 months, and 24 months	Study population		RR 0.86 (0.78 to 0.96)	3727 (15 studies)	⊕⊕⊕⊝ moderate^a
	283 per 1000^b	243 per 1000 (221 to 272)	RR 0.86 (0.78 to 0.96)	3727 (15 studies)	⊕⊕⊕⊝ moderate^a
Unplanned return to theatre (end of follow‐up)^f Follow‐up: time points in the included studies were at 12 months, 2 years and 5 years	Study population		RR 0.70 (0.45 to 1.10)	2336 (6 studies)	⊕⊕⊝⊝ low^g
	39 per 1000^b	27 per 1000 (17 to 43)	RR 0.70 (0.45 to 1.10)	2336 (6 studies)	⊕⊕⊝⊝ low^g
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; EQ‐5D: EuroQoL 5 Dimensions instrument; GARS: Groningen Activity Restriction Scale;HA: hemiarthroplasty; HRQoL: health‐related quality of life; HHS: Harris Hip Score; MCID: minimal clinically important difference; MD: mean difference; OARS‐IADL: Older Americans Resources Scale of Instrumental Activities of Daily Living; RR: risk ratio; SF‐12: Short‐form 12; SMD: standardised mean difference; THA: total hip arthroplasty
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.
^aWe downgraded by one level for study limitations because included studies had some high/unclear risks of bias. ^bDerived from the pooled estimate of the uncemented HA group ^cWe downgraded by two levels: one level for imprecision because we noted a wide CI in the estimate, and one level for study limitations because the studies had unclear risks of bias. ^dWe downgraded by three levels: one level for imprecision because we noted a wide CI in the estimate, and two levels for study limitations because some studies had unclear risks of bias, and we found during sensitivity analyses that the estimate was influenced by these studies. ^eDowngraded by two levels: one level for imprecision because the CI included possible benefits and possible harms, and one level for study limitations because the studies had unclear risks of bias. ^fSome re‐operations were because of infection, acetabular wear, dislocation, periprosthetic fracture or loosening. We noted that types of re‐operation included replacement with THA, Girdlestone and drainage of infection. ^gWe downgraded by two levels for study limitations because some studies had unclear risks of bias and all studies were at high risk of detection bias.

Open in table viewer

Summary of findings 2. Bipolar hemiarthroplasty compared with unipolar hemiarthroplasty for hip fracture in adults

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Bipolar hemiarthroplasty compared with unipolar hemiarthroplastyfor hip fracture in adults
Patient or population: adults with displaced and undisplaced hip fractures Setting: hospitals; included studies were conducted in Australia, Egypt, Finland, India, Norway, Sweden, the UK and USA Intervention: bipolar HA. These were fixed with cement in 9 studies, without cement in 3 studies, and at the discretion of the surgeon in 1 study. Comparison: unipolar HA. These were fixed with cement in 9 studies, without cement in 3 studies, and at the discretion of the surgeon in 1 study.
Outcomes	Risk with unipolar HA	Risk with Bipolar HA	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Activities of daily living, early (within 4 months)	‐	‐	‐	‐	‐	No studies reported this outcome
Early delirium Follow‐up: postoperative period	Study population		RR 0.48 (0.09 to 2.58)	261 (1 study)	⊕⊝⊝⊝ very low^b
Early delirium Follow‐up: postoperative period	31 per 1000^a	15 per 1000 (3 to 81)	RR 0.48 (0.09 to 2.58)	261 (1 study)	⊕⊝⊝⊝ very low^b
Functional status, early (within 4 months)	‐	‐	‐	‐	‐	No studies reported this outcome
HRQoL, early (within 4 months): using EQ‐5D (range 0 to 1); higher values indicate better quality of life Follow‐up: 4 months	The mean EQ‐5D score in the unipolar group was 0.54	MD 0.08 higher (0.03 lower to 0.19 higher)	‐	115 (1 study)	⊕⊝⊝⊝ very low^b
Mobility, early (within 4 months)		‐		‐		No studies reported this outcome
Mortality, early (within 4 months) Follow‐up: time points in the included studies were during hospital stay, at 3 months and at 4 months	Study population		RR 0.94 (0.54 to 1.64)	573 (4 studies)	⊕⊕⊝⊝ low^c
	105 per 1000^a	99 per 1000 (57 to 173)	RR 0.94 (0.54 to 1.64)	573 (4 studies)	⊕⊕⊝⊝ low^c
Mortality at 12 months Follow‐up: time points in the included studies were at 6 months, 12 months, 13 months, and 24 months	Study population		RR 1.17 (0.89 to 1.53)	839 (8 studies)	⊕⊕⊝⊝ low^c
	184 per 1000^a	216 per 1000 (164 to 282)	RR 1.17 (0.89 to 1.53)	839 (8 studies)	⊕⊕⊝⊝ low^c
Unplanned return to theatre (end of follow‐up)^d Follow‐up: time points in the included studies were at 12 months, 24 months, 48 months, and 60 months	Study population		RR 1.08 (0.44 to 2.64)	532 (4 studies)	⊕⊝⊝⊝ very low^e
	57 per 1000^a	62 per 1000 (25 to 151)	RR 1.08 (0.44 to 2.64)	532 (4 studies)	⊕⊝⊝⊝ very low^e
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; EQ‐5D: EuroQoL 5 Dimensions instrument; HA: hemiarthroplasty; HRQoL: health‐related quality of life; MD: mean difference; RR: risk ratio; THA: total hip arthroplasty
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.
^aDerived from the unipolar HA group if results from a single study, or otherwise, from the pooled estimate of the unipolar group ^bWe downgraded by three levels: two levels for imprecision because the evidence included very few participants, and one level for study limitations because the included study had high and/or unclear risks of bias. ^cWe downgraded by two levels: one level for imprecision because we noted a wide CI in the effect estimate, and one level for study limitations because some of the included studies had unclear risks of bias. ^dSome re‐operations were because of dislocation, acetabular wear, pain, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, revised HA, open reduction and drainage of infection. ^eWe downgraded by three levels: one level for imprecision, and two levels for study limitations because studies had high and unclear risks of bias, which included high risks of detection bias.

Open in table viewer

Summary of findings 3. Total hip arthroplasty compared with hemiarthroplasty for hip fracture in adults

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Total hip arthroplasty compared with hemiarthroplasty for hip fracture in adults
Patient or population: adults with displaced and undisplaced hip fractures Setting: hospitals; included studies were conducted in Canada, China, Greece, Finland, India, Italy, the Netherlands, New Zealand, Norway, South Africa, Spain, Sweden, the UK and USA Intervention: THA Comparison: HA (in 1 of the included studies, this was a first‐generation design of HA)
Outcomes	Risk with HA	Risk with THA	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Activities of daily living, early (within 4 months): using Katz Index and an undefined measurement tool to identify people who were independent Follow‐up: time points in the included studies were at 3 months and 4 months	Study population		RR 1.03 (0.91 to 1.18)	225 (2 studies)	⊕⊝⊝⊝ very low^b
	764 per 1000^a	787 per 1000 (695 to 901)	RR 1.03 (0.91 to 1.18)	225 (2 studies)	⊕⊝⊝⊝ very low^b
Delirium (end of follow‐up) Follow‐up: time point in the included studies was 12 months	Study population		RR 1.41 (0.60 to 3.33)	357 (2 studies)	⊕⊕⊝⊝ low^c
	47 per 1000^a	67 per 1000 (28 to 158)	RR 1.41 (0.60 to 3.33)	357 (2 studies)	⊕⊕⊝⊝ low^c
Functional status, early (within 4 months): using HHS (range from 0 to 100) and Johansen hip score (range from 0 to 100); higher scores indicate better function Follow‐up: time points in the included studies were at 3 months and 4 months	The mean HHS scores in HA groups ranged from 69 to 77.5. The mean Johansen hip score in the HA group was 71.4.	SMD 0.27 higher (0.07 higher to 0.47 higher)	‐	395 (3 studies)	⊕⊝⊝⊝ very low^d	There appeared to be no clinically important difference in this effect, based on a MCID for HHS of 16 to 18
HRQoL, early (within 4 months): using EQ‐5D (range from 0 to 1); higher scores indicate better quality of life Follow‐up: time points in the included studies were at 3 months and 4 months	The mean EQ‐5D scores in the HA groups ranged from 0.61 to 0.67.	MD 0.03 higher (0.06 lower to 0.12 higher)	‐	279 (2 studies)	⊕⊝⊝⊝ very low^e	Compatible with no effect or a clinically important benefit of THA, based on a MCID for EQ‐5D of 0.07
Mobility, early (within 4 months): using a 9‐point mobility scale; lower scores indicate better mobility Follow‐up: time point in the included study was 3 months	The mean mobility score in the HA group was 3.8	MD 0.40 lower (0.96 lower to 0.16 higher)	‐	83 (1 study)	⊕⊕⊝⊝ low^f
Mortality, early (within 4 months) Follow‐up: time points in the included studies were at 1 week, 1 month, 2 months, and 4 months	Study population		RR 0.77 (0.42 to 1.42)	725 (6 studies)	⊕⊝⊝⊝ very low^g
	62 per 1000^a	48 per 1000 (26 to 89)	RR 0.77 (0.42 to 1.42)	725 (6 studies)	⊕⊝⊝⊝ very low^g
Mortality at 12 months Follow‐up: time points in the included studies were at 12 months and 24 months	Study population		RR 1.00 (0.82 to 1.34)	2667 (11 studies)	⊕⊕⊕⊝ moderate^h
	135 per 1000^a	135 per 1000 (112 to 165)	RR 1.00 (0.82 to 1.34)	2667 (11 studies)	⊕⊕⊕⊝ moderate^h
Unplanned return to theatre (end of follow‐up)ⁱ Follow‐up: time points in the included studies were at 12 months, 24 months, 48 months, 60 months, and 13 years	Study population		RR 0.68 (0.41 to 1.15)	2476 (9 studies)	⊕⊕⊝⊝ low^j
	84 per 1000^a	57 per 1000 (35 to 97)	RR 0.68 (0.41 to 1.15)	2476 (9 studies)	⊕⊕⊝⊝ low^j
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; EQ‐5D: EuroQoL 5 Dimensions instrument; HA: hemiarthroplasty; HHS: Harris Hip Score; MCID: minimal clinically important difference; RR: risk ratio; THA: total hip arthroplasty
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.
^aDerived from the pooled estimate of the HA group ^bWe downgraded by three levels: one level for imprecision because the evidence included very few participants, and two levels for study limitations because one of the studies had unclear risk of selection bias and we found during sensitivity analyses that this may influence the estimate. ^cWe downgraded by two levels: one level for imprecision because we noted a wide CI in the effect, and one level for study limitations because of unclear risks of bias. ^dWe downgraded by three levels: one level for imprecision because the evidence included few participants, and two levels for study limitations because some studies had high and unclear risks of bias and we found during sensitivity analysis that the direction of effect was influenced by these studies. ^eWe downgraded by three levels: two levels for imprecision because the evidence was compatible with no difference and a clinically meaningful difference (based on a MCID for EQ‐5D of 0.07), and one level for study limitations because studies had high and unclear risks of bias. ^fWe downgraded by two levels: one level for imprecision because the evidence included few participants, and one level for study limitations because the study included unclear risks of bias. ^gWe downgraded by three levels: two levels for imprecision because the evidence was consistent with both benefits and harms, and one level for study limitations because some included studies had high and unclear risks of bias. ^hWe downgraded by one level for study limitations because included studies were at high or unclear risks of bias. ⁱSome re‐operations were because of dislocation, acetabular wear, pain, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, open reduction, and internal fixation. ^jWe downgraded by two levels: one level for imprecision because the evidence was consistent with both benefits and harms, and one level for study limitations because included studies had high and unclear risks of bias which included high risks of detection bias.

Background

Description of the condition

Epidemiology

A hip fracture, or proximal femoral fracture, is a break in the upper region of the femur (thigh bone) between the subcapital region (the area just under the femoral head) and 5 cm below the lesser trochanter (a bony projection of the upper femur). The incidence of hip fractures increases with age, and are most common in the older adult population (Court‐Brown 2017; Kanis 2001). Hip fractures in younger adults are usually associated with poor bone health (Karantana 2011; Rogmark 2018). A small proportion of fractures occurring in younger people are a result of high‐energy trauma, such as road traffic collisions and sports injuries. Most hip fractures are fragility fractures associated with osteoporosis, and resulting from mechanical forces that would not ordinarily result in fracture. The World Health Organization (WHO) has defined fragility fracture as those sustained from injuries equivalent to a fall from a standing height or less (Kanis 2001). In the UK, the mean age of a person with hip fracture is 83 years and approximately two‐thirds occur in women (NHFD 2017).

Hip fractures are a major healthcare problem at the individual and population level, and present a huge challenge and burden to individuals, healthcare systems, and societies. The increased proportion of older adults in the world population means that the absolute number of hip fractures is rising rapidly worldwide. For example, in 2016 there were 65,645 new presentations of hip fracture to 177 trauma units in England, Wales, and Northern Ireland (NHFD 2017). Based on mid‐2016 population estimates for these regions, this equates to an incidence rate of 108 cases per 100,000 population (ONS 2018). By 2050, the annual worldwide incidence is estimated to be 6 million hip fractures (Cooper 2011; Johnell 2004). Incident hip fracture rates are higher in high‐income countries compared to low‐ or middle‐income countries. The highest hip fracture rates are seen across northern Europe and the USA, and the lowest in Latin America and Africa (Dhanwal 2011). There is also a north‐south gradient seen in European studies, and similarly, more fractures are seen in the north of the USA than in the south (Dhanwal 2011). The factors responsible for the variation in the incidence of hip fractures and osteoporosis are thought to be population demographics (with more elderly populations in countries with higher incidence rates), and the influence of ethnicity, latitude, and environmental factors such as socioeconomic deprivation (Bardsley 2013; Cooper 2011; Dhanwal 2011; Kanis 2012).

Burden of disease

Hip fractures are also associated with a high risk of death. For example, in England, Wales, and Northern Ireland, the 30‐day mortality rate in 2016 remained high at 6.7%, despite a decline from 8.5% in 2011 and 7.1% in 2015 (NHFD 2017). Mortality at one year following a hip fracture is approximately 30%. However, fewer than half of deaths are attributable to the fracture itself, reflecting the frailty of the individuals and associated high prevalence of comorbidities and complications (Parker 1991; SIGN 2009). Morbidity associated with hip fractures is similar to stroke in terms of impact, with a substantial loss of healthy life‐years in older people (Griffin 2015). As such, hip fractures commonly result in reduced mobility and greater dependency, with many people failing to return to their pre‐injury residence. In addition, the public health impact of hip fractures is significant: data from large prospective cohorts show the burden of disease due to hip fracture is 27 disability‐adjusted life years (DALYs) per 1000 individuals, which equates to an average loss of 2.7% of the healthy life expectancy in this population at risk of fragility hip fracture (Papadimitriou 2017).

The direct economic burden of hip fractures is also substantial. Hip fractures are among the most expensive conditions seen in hospitals, with an aggregated cost of nearly 4900 million US dollars (USD) for 316,000 inpatient episodes in the USA in 2011 (Torio 2013). In England, Wales, and Northern Ireland, people with hip fracture occupy 1.5 million hospital bed days each year, and cost the National Health Service (NHS) and social care 1000 million pounds sterling (GBP) (NHFD 2017). Combined health and social care costs incurred during the first year following a hip fracture have been estimated at USD 43,669, which is greater than the cost for non‐communicable diseases, such as acute coronary syndrome (USD 32,345) and ischaemic stroke (USD 34,772) (Williamson 2017). In established market economies, hip fractures represent 1.4% of the total healthcare burden (Johnell 2004).

Types of hip fracture

Hip fractures either involve the region of the femur that is enveloped by the ligamentous hip joint capsule (intracapsular), or that is outside the capsule (extracapsular).

Intracapsular fractures include subcapital (immediately below the femoral head), transcervical (across the mid‐femoral neck), or basicervical (across the base of the femoral neck). These injuries are also commonly termed fractures of the ‘neck of femur' (Lloyd‐Jones 2015). Intracapsular fractures can be further subdivided by fracture morphology using several different classification systems, such as the Garden (Garden 1961) or Pauwels classifications (Pauwels 1935). The reliability of these various classifications is poor (Parker 1993a; Parker 1998). A more appropriate grouping distinguishes only those fractures that are displaced, where the anatomy of the bone has been disrupted at the fracture site, and those that are undisplaced (Blundell 1998; Parker 1999). This system broadly corresponds with prognosis: the more displaced, the more likely the blood supply to the femoral head is compromised, which can lead to complications such as avascular necrosis and collapse of the femoral head. More recently, this classification has been refined with additional consideration of posterior tilt ‐ this is not a component of earlier classification, but may be useful in predicting poor outcomes from osteosynthesis (Palm 2009). Furthermore, displaced fractures are less stable, so that treatments involving fixation have a higher risk of failure compared with undisplaced fractures. Approximately 60% of hip fractures are intracapsular; of these, approximately 70% to 90% are displaced (Keating 2010; NHFD 2017).

Extracapsular fractures traverse the femur within the area of bone bounded by the intertrochanteric line proximally up to a distance of 5 cm from the distal part of the lesser trochanter. Several classification methods have been proposed to define different types of extracapsular fractures (AO Foundation 2018; Evans 1949; Jensen 1980). They are generally subdivided depending on their relationship to the greater and lesser trochanters, the two bony projections present at the upper end of the femur, and the complexity of the fracture configuration. It is increasingly clear that each of these classifications is limited in its generalisability since inter‐ and intra‐observer agreement is poor. Table 1 provides a description of the most recent classification of trochanteric fractures (AO Foundation 2018). For this Cochrane Review, we plan to use a pragmatic simplification of these classifications as follows.

Open in table viewer

Table 1. Trochanteric region fractures: type and surgical management (revised AO/OTA classification, January 2018)

Type

Features

Stability

Description

Simple, pertrochanteric fractures (A1)

Isolated pertrochanteric fracture
2‐part fracture
Lateral wall intact

Stable

The fracture line can begin anywhere on the greater trochanter and end either above or below the lesser trochanter. The medial cortex is interrupted in only 1 place.

Multifragmentary pertrochanteric fractures (A2)

With 1 or more intermediate fragments
Lateral wall may be incompetent

Unstable

The fracture line can start laterally anywhere on the greater trochanter and runs towards the medial cortex which is typically broken in 2 places. This can result in the detachment of a third fragment which may include the lesser trochanter.

Intertrochanteric fractures (A3)

Simple oblique fracture
Simple transverse fracture
Wedge or multifragmentary fracture

Unstable

The fracture line passes between the 2 trochanters, above the lesser trochanter medially and below the crest of the vastus lateralis laterally.

AO/OTA: Arbeitsgemeinschaft für Osteosynthesefragen (German for "Association for the Study of Internal Fixation") / Orthopaedic Trauma Association

Trochanteric fractures: those that lie mostly between the intertrochanteric line and a transverse line at the level of the lesser trochanter. These can be further divided into simple two‐part stable fractures and comminuted or reverse obliquity unstable fractures.
Subtrochanteric fractures: those that mostly lie in the region bordered by the lesser trochanter and 5 cm distal to the lesser trochanter.

Approximately 40% of hip fractures are extracapsular, of which 90% are trochanteric and 10% are subtrochanteric (NHFD 2017).

Description of the intervention

Internationally, many guidelines exist concerning hip fracture management (e.g. AAOS 2014; NICE 2011; SIGN 2009). Each recommends that early surgical management, generally within 24 to 48 hours, is the mainstay of care for most hip fractures. The overall goal of surgery in the older population is to facilitate early rehabilitation, enabling early mobilisation and the return to premorbid function while minimising the complication risk. This approach has been associated with reductions in mortality in many worldwide registries (Neufeld 2016; Sayers 2017). A proposed grouping of arthroplasty interventions is given in Table 2.

Open in table viewer

Table 2. Proposed grouping of different types of arthroplasty for hip fracture in adults

Implant category	Variable (articulation/fixation technique)	Implant subcategory	Examples^a	Description
Total hip arthroplasty	Articulation	Femoral head and acetabular bearing surface materials	Metal‐on‐polyethylene (MoP) Ceramic‐on‐polyethylene (CoP) Ceramic‐on‐ceramic (CoC) Metal‐on‐metal (MoM) Polyethylene material Highly cross‐linked (HCL) Not HCL	Bearing surfaces may be grouped into hard (ceramic and metal) and soft (polyethylene variants). Arthroplasties exist with many of the possible combinations of these bearing surfaces.
		Femoral head size	Large head ≥ 36 mm Standard small head < 36 mm	Over the development of hip arthroplasty, different sizes of femoral head have been used, from 22 mm to very large diameters approximating that of the native femoral head. The size of the head represents a compromise between stability and linear and volumetric wear at the articulation. The optimum size varies by indication and bearing materials. 36 mm is considered as a cut‐off between standard and large sizes.
		Acetabular cup mobility	Single Dual	A standard THA has a single articulating surface between the femoral head and acetabulum bearing surface. Alternative designs incorporate a further articulation within the structure of the femoral head.
	Fixation technique	Cemented	Exeter Hip System CPT Hip System	Both components are cemented with polymethylmethacrylate bone cement that is inserted at the time of surgery. It sets hard and acts a grout between the prosthesis and the bone.
		Modern uncemented	Corail Hip System Avenir Hip System Taperloc Hip System	Neither component is cemented but rely on osseous integration forming a direct mechanical linkage between the bone and the implant. The femoral prosthesis may be coated with a substance such as hydroxyapatite which promotes bone growth into the prosthesis. Alternatively, the surface of the prosthesis may be macroscopically and microscopically roughened so that bone grows onto the surface of the implant. The acetabular component may be prepared similarly and may or may not be augmented with screws fixed into the pelvis.
		Hybrid	Combinations	The femoral stem is cemented and the acetabular cup is uncemented.
		Reverse hybrid	Combinations	The acetabular cup is cemented and the femoral stem is uncemented.
Hemiarthroplasty	Articulation	Unipolar	Thompson Austin‐Moore Exeter Trauma Stem Exeter Unitrax Endo Femoral Head CPT Zimmer Unitrax	A single articulation between the femoral head and the native acetabulum. The femoral component can be a single ‘monoblock’ of alloy or be modular, assembled from component parts during surgery.
	Articulation	Bipolar	CPT modular bipolar Exeter modular bipolar Bateman Monk Centrax	The object of the second joint is to reduce acetabular wear. This type of prosthesis has a spherical inner metal head with a size between 22 to 36 mm in diameter. This fits into a polyethylene shell, which in turn is enclosed by a metal cap. There are a number of different types of prostheses with different stem designs.
	Fixation technique	First‐generation uncemented	Thompson Austin Moore	These prostheses were designed before the development of polymethylmethacrylate bone cement and were therefore originally inserted as a ‘press fit’. Long‐term stability through osseus integration was not part of the design concept.
		Cemented	Thompson Exeter Trauma Stem Exeter Hip System CPT Hip System	The femoral stem is cemented with polymethylmethacrylate bone cement that is inserted at the time of surgery. It sets hard and acts a grout between the prosthesis and the bone.
		Modern uncemented	Corail Furlong Avenir	The femoral stem relies on osseous integration forming a direct mechanical linkage between the bone and the implant. A prosthesis may be coated with a substance such as hydroxyapatite, which promotes bone growth into the prosthesis. Alternatively, the surface of the prosthesis may be macroscopically and microscopically roughened so that bone grows onto the surface of the implant.

^aThis list is not exhaustive.

Abbreviations:
CoC: Ceramic‐on‐ceramic
CoP: Ceramic‐on‐polyethylene
CPT: collarless polished tapered
HCL: Highly cross‐linked
MoM: Metal‐on‐metal
MoP: Metal‐on‐polyethylene
THA: total hip arthroplasty

Arthroplasty

Arthroplasty entails replacing part or all of the hip joint with an endoprosthesis: an implant constructed of non‐biological materials such as metal, ceramic, or polyethylene. Arthroplasties can be grouped into two main categories: hemiarthroplasty (HA) where only the femoral head and neck are replaced, and total hip arthroplasty (THA) where both the femoral head and the acetabulum or socket are replaced.

Hemiarthroplasty

Hemiarthroplasty involves replacing the femoral head with a prosthesis whilst retaining the natural acetabulum and acetabular cartilage. The type of HA can be broadly divided into two groups: unipolar and bipolar. In unipolar HAs, the femoral head is a solid block of metal. Bipolar femoral heads include a single articulation that allows movement to occur, not only between the acetabulum and the prosthesis, but also at this joint within the prosthesis itself.

The best known of the early HA designs are the Moore prosthesis (1952) and the FR Thompson Hip Prosthesis (1954). These are both monoblock implants and were designed before the development of polymethylmethacrylate bone cement. They were therefore originally inserted as a ‘press fit’. The Moore prosthesis has a square femoral stem, which is fenestrated and has a shoulder to enable stabilisation within the femur; this resists rotation within the femoral canal. It is generally used without cement and, in the long term, bone in‐growth into the fenestrations can occur. The Thompson prosthesis has a smaller stem without fenestrations and is now often used in conjunction with cement. Numerous other designs of unipolar HAs exist, based on stems that have been used for THAs.

In bipolar prostheses, there is an articulation within the femoral head component itself. In this type of prosthesis, there is a spherical inner metal head between 22 mm and 36 mm in diameter. This fits into a polyethylene shell, which in turn is enclosed by a metal cap. The objective of the second joint is to reduce acetabular wear by promoting movement at the intraprosthetic articulation rather than with the native acetabulum. There are a number of different types of prostheses with different stem designs. Examples of bipolar prostheses are the Charnley‐Hastings, Bateman, Giliberty, and the Monk prostheses, but many other types with different stem designs exist.

Total hip arthroplasty

Total hip arthroplasty (also known as total hip replacement) involves the replacement of the acetabulum in addition to the femoral head. The first successful THA was developed by John Charnley, using metal alloy femoral heads articulating with polyethylene acetabular components. Subsequently, the articulating materials have diversified, and designs using metal alloys, ceramics, and various polyethylenes in various combinations have all been used.

Component fixation

Irrespective of the nature of the articulating surfaces, the components must be fixed to the bone to ensure longevity of the arthroplasty. The two approaches used to achieve this fixation are cemented and uncemented designs.

Cemented systems

Polymethylmethacrylate bone cement may be inserted at the time of surgery. It sets hard and acts as a grout between the prosthesis and the bone at the time of surgery. Potential advantages of cement are a reduced risk of intraoperative fracture, later periprosthetic fracture, and not relying on integration of the prosthesis with osteoporotic bone. Major side effects of cement are cardiac arrhythmias and cardiorespiratory collapse, which occasionally occur following its insertion. These complications may be fatal, leading either to embolism from marrow contents forced into the circulation (Christie 1994), or a direct toxic effect of the cement.

Uncemented systems

Uncemented systems rely on osseous integration forming a direct mechanical linkage between the bone and the implant. A prosthesis may be coated with a substance, such as hydroxyapatite, which promotes bone growth into the prosthesis. Alternatively, the surface of the prosthesis may be macroscopically and microscopically roughened so that bone grows onto the surface of the implant.

The general complications of both types of arthroplasty are those general to surgical management of hip fracture ‐ namely, pneumonia, venous thromboembolism, infection, acute coronary syndrome, and cerebrovascular accident ‐ and those specific to arthroplasty, including dislocation of the prosthesis, loosening of the components, acetabular wear, and periprosthetic fracture.

Why it is important to do this review

This review replaces the Cochrane Review, Parker 2010a, on the same topic. We used up‐to‐date review methods and have optimised current relevance in terms of patient population, implants used, and outcomes for policymaking bodies, such as the National Institute for Health and Care Excellence (NICE) in the UK, as well as international audiences. Since Parker 2010a, clinical uncertainty remains as to the optimum implant for older adults. Moreover, further studies have been reported since the last literature search in September 2009.

Appraisal and synthesis of contemporary evidence may enable more robust conclusions to be made to better inform practice. Furthermore, for displaced intracapsular fractures, the recommended treatment is either HA or THA (Parker 2010a; Hopley 2010; NICE 2011). However, there is a lack of evidence regarding whether older adults experience better outcomes with THA or HA. Recent research has also found interhospital variation and systematic inequalities in the provision of THA (Perry 2016). Further evidence is necessary to verify which individuals gain the most from THA. For treatment of undisplaced intracapsular fractures, there is also a gap in the evidence that resulted in the recently updated NICE guideline being unable to make an evidence‐based recommendation on the best surgical management strategy (NICE 2011). Other reviews that will address other types of interventions are in preparation; we focus on arthroplasty in this review.

Objectives

To determine the effects of different designs, articulations, and fixation techniques of arthroplasties for treating hip fractures in older adults.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) and quasi‐RCTs that assessed surgical interventions for the management of people with hip fracture. Quasi‐RCTs are trials in which the methods of allocating people to a trial are not properly random, but are intended to produce similar groups (Cochrane 2018). We included trials published as conference abstracts, provided the trial authors reported sufficient data relating to the methods and outcomes of interest. We aimed to include unpublished data if identified in the searches.

Types of participants

We included adults undergoing surgery in a hospital setting for fragility (low‐energy trauma) hip fractures. We included displaced and undisplaced intracapsular or extracapsular fractures which we expected to be caused by low‐energy trauma.

We expected trial populations to have a mean age of between 80 to 85 years, and include 70% women, 30% with chronic cognitive impairment, and 50% with an American Society of Anesthesiologists (ASA) score greater than II, indicating that people may have a disease or condition affecting their fitness before surgery (NHFD 2017; NICE 2011). These characteristics would be representative of the general hip fracture population.

We excluded studies that focused exclusively on the treatment of participants: younger than 16 year of age; with fractures caused by specific pathologies other than osteoporosis; and with high‐energy traumas. However, we took a pragmatic approach to study inclusion criteria, and included studies with mixed populations (fragility and other mechanisms, ages, or pathologies). We expected that participants with standard fragility fractures were most likely to outnumber those with high‐energy trauma or local pathological fractures; therefore, the results will be generalisable to the fragility fracture population. If the data were reported separately for standard fragility fractures, we planned to use this subgroup data in our main analysis.

We did not pool studies in which the fracture type is mixed (intracapsular and extracapsular).

Types of interventions

We included all hip prostheses: unipolar HA, bipolar HA, or THA (small and large head), applied with or without cement. We included the following comparisons in the review.

Prostheses inserted with cement versus without cement (stratified by THA versus HA; HA group subgrouped by modern versus first‐generation uncemented stems).
Bipolar HA versus unipolar HA (subgrouped by cemented versus uncemented).
HAs versus other HAs (subgrouped by modern stem design (‘ODEP 3A rating') and first‐generation stem design (e.g. Austin‐Moore or Thompson).
THA versus HA (cemented or uncemented, subgrouped by old versus new, as described above);
Single versus multiple (dual/triple) articulations of THA.
Large‐head THA (36 mm diameter or larger) versus other arthroplasty (stratified by THA versus HA).

We created a detailed table of interventions, grouping them by characteristics, and indicating which are in worldwide use. We prepared this table for the protocol with clinical authors and with the International Fragility Fracture Network (www.fragilityfracturenetwork.org/), and we updated it during review preparation to include all implants used in the included studies (Table 2).

Types of outcome measures

Depending on the length of follow‐up reported, we categorised the endpoints for outcomes into 'early' (up to and including 4 months), 12 months (prioritising 12‐month data, but in its absence including data after 4 months and up to 24 months) and 'late' (after 24 months, up to the end of study follow‐up). We selected four months as the definition of 'early' because most of early recovery has been achieved at this time point (Griffin 2015). This decision is also in accordance with the core outcome set for hip fracture, which prioritises early outcome over late recovery (Haywood 2014). Although priority was given to early outcomes in the presentation of our data, we also included outcome data at the '12 months' and 'late' times points.

Critical outcomes

We extracted information on the following seven 'critical' outcomes.

Activities of daily living (e.g. Barthel Index (BI), Functional Independence Measure (FIM)).
Delirium using recognised assessment scores, such as Mini‐Mental State Examination (MMSE) mental test score and the four 'A's test (4AT).
Functional status (region‐specific) (e.g. hip rating questionnaire, Harris Hip Score, Oxford Hip Score).
Health‐related Quality of Life (HRQoL) (e.g. Short Form Health Survey (SF‐36), EuroQol‐5 Dimensions (EQ‐5D)).
Mobility (e.g. indoor/outdoor walking status, Cumulated Ambulation Score, Elderly Mobility Scale Score, Timed Up and Go (TUG) test, Short Physical Performance Battery, self‐reported walking scores (e.g. Mobility Assessment Tool ‐ short form)).
Mortality.
Unplanned return to theatre: secondary procedure required for a complication resulting directly or indirectly from the index operation or primary procedure.

Other important outcomes

We also reported the following 'important' outcomes.

Pain (verbal rating or visual analogue scale (VAS)).
Length of in‐hospital stay.
Discharge destination. We used study authors’ definitions, which were variably defined in the included studies.
Adverse events.

We grouped adverse events by relatedness to the implant or fracture, or both. We reported each adverse event type separately for maximum clarity, and included the following.

Damage to a nerve, tendon, or blood vessel.
Intraoperative periprosthetic fracture.
Postoperative periprosthetic fracture.
Loosening of prosthesis.
Wound infection. We used study authors' definitions, which often distinguished deep infection and superficial infection.
Dislocation.

Unrelated

Acute kidney injury.
Blood transfusion.
Cerebrovascular accident.
Chest infection/pneumonia.
Decreased cognitive ability.
Myocardial infarction/acute coronary syndrome.
Sepsis.
Urinary tract infection.
Venous thromboembolic phenomena.

Search methods for identification of studies

As well as developing a strategy for this review, we developed general search strategies for the large bibliographic databases to find records to feed into a number of Cochrane Reviews and review updates on hip fracture surgery (Lewis 2021; Lewis 2022a; Lewis 2022b; Lewis 2022c). We searched the main databases up to July 2020.

Electronic searches

We identified RCTs and quasi‐RCTs through literature searching with systematic and sensitive search strategies, as outlined in Chapter 4 of the Cochrane Handbook of Systematic Reviews of Interventions (Lefebvre 2019, hereafter referred to as the Cochrane Handbook). We applied no restrictions on language, date, or publication status. We searched these databases for relevant trials:

Cochrane Central Register of Controlled Trials (CENTRAL; CRS Web; 8 July 2020);
MEDLINE (Ovid; 1946 to 6 July 2020);
Embase (Ovid; 1980 to 7 July 2020);
Web of Science (SCI EXPANDED; 1900 to 8 July 2020);
Cochrane Database of Systematic Reviews (CDSR; Cochrane Library; 7 July 2020);
Database of Abstracts of Reviews of Effects (DARE; www.crd.york.ac.uk/CRDWeb/; 17 December 2018);
Health Technology Assessment (HTA) database (www.crd.york.ac.uk/CRDWeb/; 17 December 2018);
Epistemonikos (www.epistemonikos.org/; 9 July 2020);
Proquest Dissertations and Theses (Proquest; 1743 to 8 July 2020);
National Technical Information Service (NTIS, for technical reports; www.ntis.gov/; 10 July 2020).

We developed a subject‐specific search strategy in MEDLINE and other listed databases. We adapted strategies with consideration of database interface differences as well as different indexing languages. In MEDLINE, we used the sensitivity‐maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2019). In Embase, we used the Cochrane Embase filter (www.cochranelibrary.com/central/central-creation) to focus on RCTs. We ran the initial searches in November and December 2018, and a top‐up search in July 2020 in all databases except for DARE and HTA, in which no new records had been added since the initial search. At the time of the search, CENTRAL was fully up to date with all records from the Cochrane Bone, Joint, and Muscle Trauma (BJMT) Group's Specialised Register, and so it was not necessary to search this separately. We developed the search strategy in consultation with Information Specialists (see Acknowledgements) and the Information Specialist for the BJMT Group. Search strategies can be found in Appendix 1.

We scanned ClinicalTrials.gov (www.clinicaltrials.gov/) for ongoing and unpublished trials on 10 July 2020.

Searching other resources

We handsearched these conference abstracts from 2016 to November 2018:

Fragility Fractures Network Congress;
British Orthopaedic Association Congress;
Orthopaedic World Congress (SICOT);
Orthopaedic Trauma Association Annual Meeting;
Bone and Joint Journal Orthopaedic Proceedings;
American Academy of Orthopaedic Surgeons Annual Meeting.

To identify further studies, we screened the reference lists of studies included in Parker 2010a as well as the reference lists of eligible studies and systematic reviews published within the last five years that were retrieved by the searches.

Data collection and analysis

In order to reduce bias, we ensured that any review author who is a co‐applicant, study author, or has or has had an advisory role on any potentially relevant study, remained independent of study selection decisions, risk of bias assessment, and data extraction for their study.

Selection of studies

Two review authors independently screened titles and abstracts of all the retrieved bibliographic records in a web‐based systematic reviewing platform, Rayyan (Ouzzani 2016), and in the top‐up search using Covidence. Full texts of all potentially eligible records passing the title‐ and abstract‐screening level were retrieved and examined independently by two review authors against the eligibility criteria described in Criteria for considering studies for this review. We conducted full‐text screening using Covidence. We resolved disagreements through discussion or by adjudication of a third review author. We excluded duplicates and collated multiple reports of the same study so that each study, rather than each report, was the unit of interest in the review. We prepared a PRISMA flow diagram to outline the study selection process, numbers of records at each stage of selection, and reasons for exclusions of full‐text articles (Moher 2009). We reported in the review details of key excluded studies, rather than all studies that were excluded from consideration of full‐text articles.

Data extraction and management

All review authors conferred on the essential data for extraction. We designed a data extraction form that aligns with the default headings in the Characteristics of included studies (see Appendix 2). Two review authors independently piloted the form on five studies and compared results. We then made changes to the template following additional discussion with the review author team. For the remaining data extraction, one review author independently extracted data and a second review author checked all the data for accuracy. We extracted the following data.

Study methodology: publication type; sponsorship/funding/notable conflicts of interest of trial authors; study design; numbers of centres and locations; size and type of setting; study inclusion and exclusion criteria; randomisation method; number of randomised participants, losses (and reasons for losses), and number analysed for each outcome. (Collecting information relating to the participant flow helped the assessment of risk of attrition bias.)
Population: baseline characteristics of the participants by group and overall (age, gender, smoking history, medication, body mass index (BMI), comorbidities, functional status such as previous mobility, place of residence before fracture, cognitive status, American Society of Anesthesiologists (ASA) status, fracture type and displacement).
Interventions: details of each intervention (number and type, manufacturer details); general surgical details (number of clinicians and their skills and experience, perioperative care such as use of prophylactic antibiotics or antithromboembolics, mobilisation or weight‐bearing protocols).
Outcomes: all outcomes measured or reported by study authors; outcomes relevant to the review (including measurement tools and time points of measure); extraction of outcome data into data and analysis tables or additional tables in Review Manager 2014.

Assessment of risk of bias in included studies

We assessed risk of bias in the included studies using the Cochrane risk of bias tool (Higgins 2011a). We assessed the following domains.

Sequence generation (selection bias).
Allocation concealment (selection bias).
Blinding of participants, personnel (performance bias).
Blinding of outcome assessors (detection bias).
Incomplete outcome data (attrition bias).
Selective reporting (reporting bias).
Other risks of bias.

We considered risk of detection bias separately for: subjective outcomes measured by clinicians, objective outcomes measured by clinicians, and participant‐reported outcomes (e.g. pain and HRQoL). For each domain, two review authors judged whether study authors made sufficient attempts to minimise bias in their design. For each domain, we made judgements using three measures ‐ high, low, or unclear risk of bias ‐ and we recorded these judgements in risk of bias tables.

Measures of treatment effect

We calculated risk ratios (RRs) for dichotomous data outcomes with 95% confidence intervals (CIs); it was not appropriate to use Peto odds ratio (OR) to calculate effects because no outcomes had very low numbers of observed events. We expressed treatment effects for continuous data outcomes evaluated using the same measurement scales as mean differences (MD) with 95% CI. For outcomes measured using different scales, we used standardised mean differences (SMD) with 95% CI.

In the event that studies reported dichotomous data using more than one category, we selected these cut‐off points in the distribution of categories:

for functional status: we reported data for those with a score of excellent or good (using Harris Hip Score (HHS)) versus those with a score of moderate or poor;
for mobility: we reported data for those who were able to walk independently out of doors with no more than the use of one stick (NICE 2011), versus those who were more dependent;
for pain: we reported data for participants who reported no pain versus those who reported any category of pain;
for discharge destination: we reported data for participants who were discharged home versus those who were discharged to a care environment.

Unit of analysis issues

In preparation of the review, we encountered potential unit of analysis issues. We found that some studies reported number of hip fractures (or cases) as well as the number of participants, with a very small number of participants having two fractured hips. Often, differentiating the denominators within a report was challenging. In such studies, depending on the outcome, the unit of analysis was either the participant (for example, for outcomes such as mortality, discharge destination, or some adverse events), or the case (for example, for outcomes such as unplanned return to theatre). We noted this differentiation where applicable and used the unit of analysis (participants or case) that was appropriate for the outcome within these studies. One study included three intervention groups (Dorr 1986). We created a pairwise comparison by combining the data for the two HA groups (cemented and uncemented) and comparing these data with the THA group. Although the review included a comparison of cemented HA versus uncemented HA, we did not use data from these two study arms in this comparison because recruitment to these two groups was completed at different time points within the study period and thus it was not appropriate to compare these against one another.

Dealing with missing data

For each included study, we recorded the number of participant losses for each outcome. Unless reported otherwise, we assumed complete case data for mortality, unplanned return to theatre, and adverse events. For outcomes that required participant assessment at end of follow‐up (such as HRQoL), we prioritised intention‐to‐treat (ITT) data where these data were available. If ITT data were unavailable for these outcomes, and if study authors did not clearly report denominator figures for each group for the outcome, we reduced the denominator figure in each group to account for reported mortality. We did not impute missing data. We used the risk of bias tool to judge attrition bias. We judged studies to be at high risk of attrition bias if we noted large amounts of unexplained missing data, loss that could not be easily justified in the study population, or losses were not sufficiently balanced between intervention groups. If we included a study with high attrition bias, we explored the effect during sensitivity analysis. We completed sensitivity analysis only for critical review outcomes and only considered attrition for outcomes that may be affected by these losses.

We attempted to contact study authors of more recently published trials when we noted that data for critical outcomes appeared to be measured but not reported. Where standard deviations were not reported, we attempted to determine these from other reported data (such as standard errors, confidence intervals, or exact P values). We noted in the Characteristics of included studies when we could not use outcome data because they were insufficiently reported or because numbers of losses in each group were not clearly specified.

Assessment of heterogeneity

We used the I² statistic, automatically calculated in Review Manager 2014 software, to quantify the possible degree of heterogeneity of treatment effects between trials. We assumed moderate heterogeneity when the I² was between 30% and 60%; substantial heterogeneity when it was between 50% and 90%; and considerable heterogeneity when it was between 75% and 100%. We noted the importance of I² depending on: 1) magnitude and direction of effects; and 2) strength of evidence for heterogeneity. We did not have sufficient studies to investigate statistical heterogeneity (Deeks 2017).

We assessed clinical and methodological diversity in terms of participants, interventions, outcomes, effect modifiers, and study characteristics for the included studies to determine whether a meta‐analysis was appropriate; we used the information collected during data extraction (Data extraction and management).

We visually inspected forest plots to look at the consistency of intervention effects across included studies. If the studies were estimating the same intervention effect, there should be overlap between the CIs for each effect estimate on the forest plot, but if overlap is poor, or there are outliers, then statistical heterogeneity may be likely.

Assessment of reporting biases

We planned to investigate the potential for publication bias and explore possible small‐study biases using funnel plots. However, we had insufficient studies (fewer than 10 studies) for most outcomes (Sterne 2017). For outcomes with 10 or more studies, we constructed a funnel plot and interpreted the plot using a visual inspection and the Harbord modified test in Stata; for the critical review outcomes, we reported P values for the Harbord modified test. We incorporated this judgement into the assessment of publication bias within the GRADE assessment.

To assess outcome reporting bias, we screened clinical trials registers for protocols and registration documents of included studies that were prospectively published, and we sourced all clinical trials register documents that were reported in the study reports of included studies. We used evidence of prospective registration to judge whether studies were at risk of selective reporting bias.

Data synthesis

We conducted meta‐analyses only when meaningful; that is, when the treatments, participants, and the underlying clinical question were similar enough for pooling to make sense. We pooled results of comparable groups of trials using random‐effects models. We chose this model after careful consideration of the extent to which any underlying effect could truly be thought to be fixed, given the complexity of the interventions included in this review. We presented 95% CIs throughout. We found that some studies reported outcome data at more than one time point and we reported the data within three time point windows for these studies. Early data included data up to four months, with priority given to data closest to four months; 12‐month data included a window from later than four months up to 24 months, but with priority given to data at 12 months; and late data, which included data reported after 24 months at the latest time point reported by study authors. For studies that reported outcome data using more than one measurement tool, we selected the tool that was used most commonly by other studies in the comparison group, or which reported data for the largest number of participants.

We considered the appropriateness or otherwise of pooling data where there was considerable heterogeneity (I² statistic value of greater than 75%) that could not be explained by the diversity of methodological or clinical features amongst trials. We presented data from these studies in the analyses and clearly reported these observations in the text for the critical outcomes in the review.

If effect sizes were statistically significant, we considered whether the effect was clinically important. We based these decisions on established minimal clinically important differences (MCIDs) for the measurement tool, or used Cohen's effect sizes as a guide if MCIDs were unavailable (Schünemann 2019a).

Subgroup analysis and investigation of heterogeneity

Few outcomes provided evidence from at least 10 studies to justify subgroup analysis. Although we aimed to explore possible sources of heterogeneity between studies (key effect modifiers such as age, gender, cognitive impairment, and fracture displacement and location), these possible effect modifiers were insufficiently reported to allow for meaningful subgroup analysis.

We planned to subgroup prostheses according to whether a modern or first‐generation uncemented stem was used (see Types of interventions), and we reported the test for subgroup differences in outcomes that had at least 10 studies.

There is no explicit means of accounting for step changes in co‐interventions, certainly not one that would be applicable to the worldwide totality of the evidence. Therefore, we could not try to explain any heterogeneity by statistical test of subgroups defined by co‐intervention. However, we ordered forest plots by date of recruitment so that any temporal trend could be inspected visually and commented on.

Sensitivity analysis

We used sensitivity analysis to explore the effects of risks of bias on the review's critical outcomes. If pooled analyses had at least two studies, we excluded studies that were:

at high or unclear risk of selection bias for sequence generation (this included studies described as quasi‐randomised, or those that did not adequately describe methods used to randomise participants to intervention groups); or
at high risk of attrition bias (because studies reported a large number of losses that were unexplained or not justified for this population, or losses that were unbalanced between groups, and that we expected could influence outcome data).

We compared the effect estimates in the sensitivity analysis with the effect estimates in the primary analysis, and we reported the effect estimates from sensitivity analyses only if we noted a difference in our interpretation of the effect. We planned to conduct sensitivity analysis by excluding studies that had mixed populations, but these data were inadequately reported by study authors and did not allow for meaningful analysis. We also planned, but did not conduct, sensitivity analysis by excluding studies of interventions that are not currently in clinical use. We obtained the general view that all interventions at the major‐grouping level (implant sub‐category level in Table 2) remain in current use. Although some types of implant may no longer be manufactured, we believe the distinction between implants within the same category is marginal and that sensitivity analysis would not be meaningful.

Summary of findings and assessment of the certainty of the evidence

Two review authors used the GRADE system to assess the certainty of the body of evidence associated with the seven critical outcomes in the review (Schünemann 2019b):

activities of daily living (ADL);
delirium;
functional status;
health‐related quality of life (HRQoL);
mobility;
mortality (measured within four months of surgery, and at 12 months);
unplanned return to theatre.

For outcomes that were reported using more than one measurement tool, and that could not be combined in analysis, we assessed the certainty of the evidence for the outcome that used a measurement tool with the most participants.

The GRADE approach assesses the certainty of a body of evidence based on the extent to which we can be confident that an estimate of effect or association reflects the item being assessed. Evaluation of the certainty of a body of evidence considers within‐study risk of bias (study limitations), directness of the evidence (indirectness), heterogeneity of the data (inconsistency), precision of the effect estimates (imprecision), and risk of publication bias. The certainty of the evidence could be high, moderate, low or very low, being downgraded by one or two levels depending on the presence and extent of concerns in each of the five GRADE domains. We used footnotes to describe reasons for downgrading the certainty of the evidence for each outcome, and we used these judgements when drawing conclusions in the review.

We did not construct summary of findings tables for all comparisons in this review. Instead, we selected three comparisons that provided the most substantial body of evidence. These provided evidence for each of our comparison types in our review objectives (different fixation techniques, different articulations, and different designs). We therefore constructed summary of findings tables for the following comparisons in this review, using the GRADE profiler software (GRADEpro GDT).

Cemented HA versus uncemented HA.
Bipolar HA versus unipolar HA.
THA versus HA.

Results

Description of studies

See Characteristics of included studies, Characteristics of excluded studies, Characteristics of studies awaiting classification, and Characteristics of ongoing studies.

Results of the search

After the removal of duplicates from the search results, we screened 28,509 titles and abstracts, which included backward citation searches and searches of clinical trials registers. We reviewed the full texts of 1135 records and selected 58 studies (with 101 records) for inclusion in this review. We linked any references pertaining to the same study under a single study ID. We excluded 1023 records, and report the details of eight key studies from these excluded records. Four studies are awaiting classification, and we identified seven ongoing studies. See Figure 1.

Figure 1

PRISMA flow diagram

Included studies

See Characteristics of included studies. Two studies were reported only as abstracts with limited study characteristics (Moroni 2002; Patel 2008).

Types of studies and setting

Whilst most studies were RCTs, eight studies used methods to allocate participants to interventions which we assessed as quasi‐randomised (Abdelkhalek 2011; Dorr 1986; Iorio 2019; Livesley 1993; Ravikumar 2000; Santini 2005; Sonaje 2017; Stoffel 2013).

Eleven were multicentre studies, and the remainder were single centre studies. Eighteen studies were completed in the UK (Baker 2006; Brandfoot 2000; Calder 1995; Calder 1996; Davison 2001; Emery 1991; Fernandez 2022; Griffin 2016; Harper 1994; Keating 2006; Livesley 1993; Parker 2010c; Parker 2012; Parker 2019; Parker 2020; Ravikumar 2000; Sadr 1977; Sims 2018), six in Sweden (Blomfeldt 2007; Chammout 2017; Chammout 2019; Cornell 1998; Hedbeck 2011; Inngul 2015), four in South Asia (Malhotra 1995; Rehman 2014; Sharma 2016; Sonaje 2017), four in Italy (Cadossi 2013; Iorio 2019; Moroni 2002; Santini 2005), four in the USA (DeAngelis 2012; Dorr 1986; Macaulay 2008; Raia 2003), three in Norway (Figved 2009; Figved 2018; Talsnes 2013), three in China (Cao 2017; Ren 2017; Xu 2017), three in Australasia (Jeffcote 2010; Stoffel 2013; Taylor 2012), two in the Netherlands (Moerman 2017; Van den Bekerom 2010), two in Egypt (Abdelkhalek 2011; Rashed 2020), and two in South Korea (Kim 2012; Lim 2020). HEALTH 2019 was an international study conducted in Canada, Finland, the Netherlands, New Zealand, Norway, South Africa, Spain, the UK, and the USA. The remainder were conducted in other European countries (Kanto 2014; Mouzopoulos 2008; Movrin 2020; Sonne‐Holm 1982; Vidovic 2013), and one study did not report where the study was conducted (Patel 2008).

Studies were published between 1977 and 2020, and we include one study that we expect to be published in 2021. Approximately two‐thirds of the studies were published since 2010.

Types of participants

In total, 10,654 participants, with 10,662 hip fractures, were recruited across the 58 studies. All studies included only participants with intracapsular fractures, except for Cao 2017 (85 participants), which included only participants with extracapsular fractures. Blomfeldt 2007 is the only study to report the inclusion of undisplaced fractures, which was only 2% of the reported study population. Nine studies did not report whether the fracture was displaced (Cadossi 2013; Cao 2017; Dorr 1986; Malhotra 1995; Moroni 2002; Patel 2008; Santini 2005; Sonne‐Holm 1982; Xu 2017), and the remainder included displaced fractures only. One study recruited participants that had neglected fractures, more than 30 days old (Xu 2017).

Most studies specified a lower age limit for recruited participants of at least 50 years (HEALTH 2019; Macaulay 2008), 55 years (DeAngelis 2012; Dorr 1986; Rashed 2020), 60 years (Baker 2006; Fernandez 2022; Griffin 2016; Iorio 2019; Jeffcote 2010; Parker 2010c; Parker 2019; Rehman 2014; Sharma 2016; Sims 2018; Sonaje 2017; Xu 2017), 65 years (Cao 2017; Chammout 2017; Cornell 1998; Davison 2001; Kanto 2014; Lim 2020; Raia 2003; Ravikumar 2000; Santini 2005), 70 years (Blomfeldt 2007; Cadossi 2013; Figved 2009; Figved 2018; Moerman 2017; Patel 2008; Sonne‐Holm 1982; Taylor 2012; Van den Bekerom 2010; Vidovic 2013), 75 years (Moroni 2002; Movrin 2020; Talsnes 2013), and 80 years (Calder 1996; Chammout 2019; Hedbeck 2011; Inngul 2015). Only five studies applied an upper age limit, which was 79 years (Calder 1995; Chammout 2017; Davison 2001), 80 years (Rashed 2020), and 90 years (Blomfeldt 2007). Where reported, the mean ages of participants ranged from 63 years to 87 years.

Seven studies did not report the baseline sex of the participants (Griffin 2016; Livesley 1993; Patel 2008; Ravikumar 2000; Sonaje 2017; Sonne‐Holm 1982; Stoffel 2013). In studies that reported sex distribution, there were 6835 female participants, which represented 71% of the sample in these studies. Approximately one third of the studies specified the ability to walk prior to surgery as an inclusion criteria or required participants to be free of any cognitive impairment. Almost half of the studies stated that pathological fractures would not be included.

The mean follow‐up time period was 24.3 (SD ± 109) months, with a range from 1 week (Malhotra 1995), to 13 years (Ravikumar 2000).

Types of interventions

We included 21 studies with 4282 participants that compared prostheses that were cemented or uncemented; as part of treatment with a THA (Chammout 2017), a HA (Brandfoot 2000; Cao 2017; DeAngelis 2012; Emery 1991; Fernandez 2022; Figved 2009; Harper 1994; Moerman 2017; Movrin 2020; Parker 2010c; Parker 2020; Rehman 2014; Sadr 1977; Santini 2005; Sonne‐Holm 1982; Talsnes 2013; Taylor 2012; Vidovic 2013), or a mixture of either a THA or HA (Inngul 2015; Moroni 2002). We briefly summarise the characteristics of these studies and the critical review outcomes they report that are relevant to this review in Table 3.

Open in table viewer

Table 3. Implant and study characteristics. Prostheses implanted with cement versus without cement

Study ID	Type of cemented implant Type of uncemented implant	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Brandfoot 2000	1. Cemented, Thompson, unipolar 2. Uncemented Thompson, unipolar	RCT (91)	98	Mortality (16 months, 91)
Cao 2017	1. Cemented, stem type and uni/bipolar NR 2. Uncemented, stem type and uni/bipolar NR	RCT (85)	NR	Function (3 and 6 months, 85)
Chammout 2017	1. Cemented, modular CPT, 32 mm head, cemented cup 2. Uncemented, Bi‐Metric stem, 32 mm head, cemented cup	RCT (69)	100	ADL (3 months, 65; 24 months, 59) Function (24 months, 65) HRQoL (3 months, 64; 12 months, 62) Mortality (12 months, 69) Unplanned return to theatre (24 months, 69)
DeAngelis 2012	1. Cemented, VerSys stem, unipolar 2. Uncemented, beaded stem, unipolar	RCT (130)	100	Unplanned return to theatre (12 months, 130)
Emery 1991	1. Cemented, Thompson, bipolar 2. Uncemented, Moore, bipolar	RCT (53)	100	Mobility (3 months, 39) Mortality (3 and 17/18 months, 53)
Figved 2009	1. Cemented, Spectron, bipolar 2. Uncemented, Corail, bipolar	RCT (230 fractures, 223 participants)	100	ADL (3 months, 190; 12 months, 168) Function (3 months, 189; 12 months, 167) HRQoL (3 months, 143; 12 months, 113) Mobility (3 months, 190; 12 months, 168) Mortality (3 and 12 months, 213) Unplanned return to theatre (12 months, 217)
Harper 1994	1. Cemented, Thompson, unipolar 2. Uncemented, Thompson, unipolar	RCT (137)	100	Mortality (3 and 12 months, 137)
Inngul 2015	1. Cemented, Exeter stem, unipolar or 32mm, cemented cross‐linked polyethylene cup 2. Uncemented, HAC Bimetric stem, unipolar or 32 mm, cemented cross‐linked polyethylene cup	RCT (141)	100	Mortality (4 and 12 months, 141) Unplanned return to theatre (12 months, 141)
Moerman 2017	1. Cemented, Muller, bi/unipolar NR 2. Uncemented, DB10, bi/unipolar NR	RCT (201)	100	ADL (3 months, 114; 12 months, 96) HRQoL (3 months, 102; 12 months, 90) Mobility (3 months, 88; 12 months, 74) Mortality (12 months, 201) Unplanned return to theatre (12 months, 201)
Moroni 2002	1. Cemented, AHS prosthesis, unipolar or THA 2. Uncemented (HAC), Furlong, unipolar or THA	RCT (28)	NR	Function (24 months, 28) HRQoL (24 months, 28) Mortality (24 months, 28)
Movrin 2020	1. Cemented, Muller, bi/unipolar NR 2. Uncemented, DB10, bi/unipolar NR	RCT (158)	100	Function (3 month, 148; 24 months, 94) Mortality (7 days and 24 months, 158)
Parker 2010c	1. Cemented, Thompson, unipolar 2. Uncemented, Moore, unipolar	RCT (400)	100	Delirium (60 months, 400) Mobility (3 months, 327; 60 months, 64) Mortality (12 and 60 months, 400) Unplanned return to theatre (60 months, 400)
Parker 2020	1. Cemented, Exeter Trauma or CPT, unipolar 2. Uncemented, Furlong, unipolar	RCT (400)	100	ADL (4 months, 329; 12 months 283) Delirium (12 months, 400) Mobility (3 months, 329; 12 months, 282) Mortality (3 and 12 months, 400)
Rehman 2014	1. Cemented, Thompson, unipolar 2. Uncemented, Moore, unipolar	RCT (110)	100	Mobility (3 months, 110)
Sadr 1977	1. Cemented, Thompson, unipolar 2. Uncemented, Thompson, unipolar	RCT (40)	100	Function (17 months, 25) Mortality (6 weeks and 12 months, 40)
Santini 2005	1. Cemented, stem type NR, unipolar 2. Uncemented, stem type NR, unipolar	RCT (106)	NR	ADL (12 months, 106) Function (12 months, 106) Mobility (unknown time point, 106) Mortality (at hospital discharge and 12 months, 106)
Sonne‐Holm 1982	1. Cemented, Moore, unipolar 2. Uncemented, Moore, unipolar	RCT (112)	NR	Function (3 and 12 months, 75) Mobility (3 and 12 months, 75) Mortality (6 weeks, 112)
Talsnes 2013	1. Cemented, Landos Titan, bipolar 2. Uncemented, Landos Corail, bipolar	RCT (334)	100	Mortality (12 months, 334)
Taylor 2012	1. Cemented, Exeter, unipolar 2. Uncemented, Zweymuller Alloclassic, unipolar	RCT (160)	100	Mortality (6 weeks and 12 months, 160) Unplanned return to theatre (24 months, 160)
Vidovic 2013	1. Cemented, modular, unipolar 2. Uncemented, Moore, unipolar	RCT (79)	100	Function (3 months, 79; 12 months, 60) Mortality (12 months, 79)
Fernandez 2022	1.Cemented HA, stem and head at surgeon's preference 2.Uncemented HA, stem and head at surgeon's preference	RCT (1225)	99	ADL (4 months, 715; 12 months, 580) HRQoL (4 months, 877; 12 months, 876) Mobility (4 months, 715; 12 months, 583) HRQoL (4 months, 877; 12 months, 876) Unplanned return to theatre (12 months, 1225) Mortality (12 months, 1225)

ADL: activities of daily living
AHS: manufacturer's name for implant
CPT: collarless, polished, double‐taper design concept
DB: manufacturer's name for implant
HAC: hydroxyapatite‐coated
HRQoL: health‐related quality of life
N: total number randomised
n: number analysed
NR: not reported
RCT: randomised controlled trial

We included 13 studies with 1499 participants that compared a bipolar HA with a unipolar HA (Abdelkhalek 2011; Calder 1995; Calder 1996; Cornell 1998; Davison 2001; Figved 2018; Hedbeck 2011; Jeffcote 2010; Kanto 2014; Malhotra 1995; Patel 2008; Raia 2003; Stoffel 2013). We briefly summarise the characteristics of these studies and the outcomes they report that are relevant to this review in Table 4.

Open in table viewer

Table 4. Implant and study characteristics. Bipolar HA versus unipolar HA

Study ID	Type of HA bipolar Type of HA unipolar	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Abdelkhalek 2011	1. Mixed cemented/uncemented, bipolar; 2. Mixed cemented/uncemented, unipolar	Quasi RCT (50)	100	Function (4.4 years, 50) Unplanned return to theatre (24 months, 50)
Calder 1995	1. Monk, cemented, bipolar 2. Thompson, cemented, unipolar	RCT (73)	100	Pain (6 months, 73) Mobility (6 months, 73)
Calder 1996	1. Monk, cemented, bipolar 2. Thompson, cemented, unipolar	RCT (250)	100	Mortality (4 and 12 months, 250)
Cornell 1998	1. Cemented modular, bipolar 2. Cemented modular, unipolar	RCT (48)	100	Function (6 months, 48) Mobility (6 months, 48) Mortality (6 months, 48)
Davison 2001	1. Cemented, Monk, bipolar 2. Cemented, Thompson, unipolar	RCT (187)	100	Mortality (12 and 36 months, 187) Unplanned return to theatre (36 months, 187)
Figved 2018	1. Cemented, 28 mm cobalt chromium head and a SelfCentering Bipolar (DePuy) 2. Cemented, Modular Cathcart Unipolar (DePuy)	RCT (28)	100	Function (48 months, 19) HRQoL (12 months, 25; 48 months, 19) Mortality (3 and 12 months, 28)
Hedbeck 2011	1. Cemented, UHR (Stryker), from 42 to 72 mm, bipolar 2. Cemented, Exeter modular, unipolar	RCT (120)	100	ADL (12 months, 99) HRQoL (4 months, 115; 12 months, 99) Mortality (4 and 12 months, 120) Unplanned return to theatre (12 months, 120)
Jeffcote 2010	1. Cemented, Centrax, bipolar 2. Cemented, Unitrax, unipolar	RCT (51)	100	Mortality (24 months, 51)
Kanto 2014	1. Cemented, Vario cup, bipolar 2. Cemented, Lubinus, unipolar	RCT (175)	100	Mortality (during hospital stay and 5 years, 175) Unplanned return to theatre (5 years, 175)
Malhotra 1995	1. Uncemented, Bateman type, bipolar 2. Uncemented; Austin‐Moore; unipolar	RCT (68)	NR	Function (NR, 66)
Patel 2008	1. Uncemented, medical internation stem, bipolar 2: Uncemented, Thompson; unipolar	RCT (40)	100	Mortality (13 months, 40)
Raia 2003	1. Centrax, appropriate‐sized cemented Premise stem, bipolar 2. Unitrax; appropriate‐sized cemented Premise stem, unipolar	RCT (115)	100	Mortality (12 months, 115)
Stoffel 2013	1. Cemented, collarless polished stem, bipolar 2. Cemented, collarless polished stem, unipolar	RCT (294)	100	Delirium (12 months, 261) Function (12 months, 251) Mobility (12 months, 186)

ADL: activities of daily living
HA: hemiarthroplasty
HRQoL: health‐related quality life
N: total number randomised
n: number analysed
NR: not reported
RCT: randomised controlled trial
UHR: universal head system (manufacturer's name)

We included four studies with 1397 participants that compared different types of HAs. These comparisons were between a short stem and standard stem (Lim 2020), a Thompson and a Exeter Trauma Stem (Parker 2012; Sims 2018), and an Austin‐Moore and a Furlong (Livesley 1993). We briefly summarise the characteristics of these studies and the outcomes they report that are relevant to this review in Table 5.

Open in table viewer

Table 5. Implant and study characteristics. HAs versus other HAs

Study ID	Type of HA in each intervention group	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Lim 2020	1. Short stem, Bencox M stem, proximal Ti‐plasma spray microporous coating, uncemented, bipolar 2. Standard stem, Bencox ID stem, proximal Ti‐plasma spray microporous coating, uncemented, standard stem, bipolar	RCT (151)	100	ADL (24 months, 75) Mortality (24 months, 151)
Livesley 1993	1. HAC bipolar 2. Uncemented; press‐fit Moore‐bipolar	Quasi‐RCT (82)	100	Mortality (1 and 12 months, 82) Unplanned return to theatre (12 months, 82)
Parker 2012	1. Uncemented, Exeter, unipolar 2. Cemented, Thompson, unipolar	RCT (200)	100	Delirium (12 months, 200) Mortality (3 and 12 months, 200) Unplanned return to theatre (12 months, 200)
Sims 2018	1. Uncemented, Exeter, unipolar 2. Cemented, Thompson, unipolar	RCT (964)	100	HRQoL (4 months, 618) Mobility (4 months, 494) Mortality (4 months, 964) Unplanned return to theatre (12 months, 964)

ADL: activities of daily living
HA: hemiarthroplasty
HAC: hydroxyapatite‐coated
HRQoL: health‐related quality of life
N: total number randomised
n: number analysed
RCT: randomised controlled trial

We included 17 studies with 3232 participants that compared a THA with a HA (Baker 2006; Blomfeldt 2007; Cadossi 2013; Chammout 2019; Dorr 1986; HEALTH 2019; Iorio 2019; Keating 2006; Macaulay 2008; Mouzopoulos 2008; Parker 2019; Ravikumar 2000; Ren 2017; Sharma 2016; Sonaje 2017; Van den Bekerom 2010; Xu 2017). We briefly summarise the characteristics of these studies and the outcomes they report that are relevant to this review in Table 6.

Open in table viewer

Table 6. Implant and study characteristics. THA versus HA

Study ID	Type of THA Type of HA	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Baker 2006	1. 28 mm femoral head articulating with an all‐polyethylene Zimmer cemented acetabular cup 2. Endo Femoral Head (Zimmer); cemented; unipolar	RCT (81)	100	Mortality (39 months, 81)
Blomfeldt 2007	1. Modular Exeter femoral component; 28 mm head; OGEE cemented acetabular component 2. Bipolar; modular Exeter, 28 mm head	RCT (120)	100	ADL (4 months, 114; 12 months, 111) Delirium (4 months, 116) Function (48 months, 83) Mortality (4, 12 and 48 months, 120)
Cadossi 2013	1. Uncemented Conus stem and a large‐diameter femoral head 2. Uncemented, bipolar	RCT (96)	100	Mortality (12 and 36 months, 96)
Chammout 2019	1. Cemented 32 mm cobalt‐chromium head; cemented highly cross‐linked polyethylene acetabular component 2. Cemented, unipolar	RCT (120)	100	ADL (3 months, 111; 24 months, 99) Delirium (3 months, 111) Function (24 months, 103) HRQoL (3 months, 111; 12 months, 106) Mortality (24 months, 120) Unplanned return to theatre (24 months, 120)
Dorr 1986	1. 28 mm head size was used 2. Cemented (n = 37) or uncemented (n = 13), bipolar	RCT (89)	100	Unplanned return to theatre (48 months, 89)
HEALTH 2019	1. Surgeon's preference 2. Surgeon's preference	RCT (1495)	100	Function (24 months, 669) HRQoL (24 months, 844) Mobility (24 months, 535) Mortality (24 months, 1441) Unplanned return to theatre (24 months, 1441)
Iorio 2019	1. Dual mobility cup with cementless femoral stem 2. Cementless femoral stem with bipolar head	RCT (60)	100	Mortality (1 and 12 months, 60) Unplanned return to theatre (12 months, 60)
Keating 2006	1. NR 2. Bipolar, cemented	RCT (180)	100	Delirium (24 months, 168) Function (24 months, 168) HRQoL (4 and 12 months, 168) Mortality (24 months, 180) Unplanned return to theatre (12 months, 180)
Macaulay 2008	1. Surgeon's preference 2. Surgeon's preference	RCT (41)	100	Function (24 months, 40) HRQoL (12 months, 40) Mobility (12 and 24 months, 40) Mortality (24 months, 40)
Mouzopoulos 2008	1. Plus (DePuy) 2. Merete	RCT (86)	100	ADL (48 months, 43) Function (48 months, 43) Mortality (12 and 48 months, 86) Unplanned return to theatre (48 months, 49)
Parker 2019	1. CPCS stem (n=29), CPT Zimmer (n=23) 2. Monoblock Exeter Trauma Stem (n=22), CPT bipolar (n=4), CPT modular (n=27)	RCT (105)	100	ADL (12 months, 78) Delirium (12 months, 105) Mobility (12 months, 78) Mortality (4 and 12 months, 105) Unplanned return to theatre (12 months, 105)
Ravikumar 2000	1. Cemented with Howse II 2. Uncemented Austin‐Moore	RCT (180)	100	Mobility (13 years, 32) Mortality (4 and 12 months and 13 years, 180) Unplanned return to theatre (12 months, 180)
Ren 2017	1. Surgeon's preference 2. Cemented	RCT (100)	NR	Function (NR, 100)
Sharma 2016	1. NR 2. NR	RCT (80)	100	Mortality (1 week, 80)
Sonaje 2017	1. NR 2. NR	Quasi‐RCT (42)	100	Function (24 months, 40)
Van den Bekerom 2010	1. Cemented; 32 mm diameter modular head 2. Cemented, bipolar	RCT (281)	100	Mortality (12 and 60 months, 252) Unplanned return to theatre (60 months, 252)
Xu 2017	1. Uncemented prosthesis 2. Bipolar; uncemented	RCT (76)	NR	Function (60 months, 76) Mortality (60 months, 76)

ADL: activity of daily living
CPCS: collarless, polished, cemented stem
CPT: collarless, polished, double‐taper design concept
HA: hemiarthroplasty
N: total number randomised
n: number analysed
NR: not reported
OGEE: manufacturer's name for implant
RCT: randomised controlled trial
THA: total hemiarthroplasty

We included three studies with 244 participants that compared different types of THAs. These comparisons were between a single articulation and a dual‐mobility articulation (Griffin 2016; Rashed 2020), and a short stem and standard stem (Kim 2012). We briefly summarise the characteristics of these studies and the outcomes they report that are relevant to this review in Table 7.

Open in table viewer

Table 7. Implant and study characteristics. THAs versus other THAs

Study ID

Type of THA

Study design
(N)

Displaced fractures %

Critical review outcomes (time point, n)

Griffin 2016

1. Single articulation: surgeon's preference

2. Dual mobility: surgeon's preference for prosthesis, Novae DM acetabular component; uncemented

RCT (21)

100

Function (12 months, 19)

HRQoL (12 months, 19)

Mortality (12 months, 21)

Rashed 2020

1. Single articulation: cemented 32 mm head

2. Dual mobility: cemented dual‐mobility cup (Ecofit 2M)

RCT (108)

100

Function (12 months, 60)

Mortality (12 months, 62)

Kim 2012

1. Short stem: short, anatomical metaphyseal‐fitting cementless femoral component, 36 mm modular head, cementless acetabular component

2. Standard stem: anatomical medullary locking fully porous coated cementless femoral component, 36 mm Biolox delta ceramic modular head

RCT (161)

100

Function (24 months, 140)

Mobility (24 months, 142)

Mortality (12 months, 162)

DM: dual‐mobility
HRQoL: health‐related quality of life
N: total number randomised
n: number analysed
RCT: randomised controlled trial
THA: total hip arthroplasty

We found no studies of large‐head THAs compared with other arthroplasties.

Types of outcome measures

All studies in our main comparison groups reported data for at least one of our critical review outcomes.

Sources of funding and declarations of interest

Fourteen studies reported that they received no commercial or external funding (Baker 2006; Cadossi 2013; Calder 1996; Davison 2001; Emery 1991; Inngul 2015; Livesley 1993; Parker 2010c; Parker 2012; Parker 2019; Parker 2020; Santini 2005; Sonaje 2017; Van den Bekerom 2010), and six studies reported funding from independent sources such as research foundations (Chammout 2019; Fernandez 2022; Griffin 2016; HEALTH 2019; Keating 2006; Macaulay 2008). Eleven studies received support from manufacturers or insurance companies (Blomfeldt 2007; DeAngelis 2012; Dorr 1986; Figved 2009; Figved 2018; Hedbeck 2011; Raia 2003; Ravikumar 2000; Sims 2018; Talsnes 2013; Taylor 2012). Eight studies declared that the study investigators had no conflicts of interest (Cao 2017; Chammout 2017; Iorio 2019; Lim 2020; Movrin 2020; Rashed 2020; Vidovic 2013; Xu 2017). The remaining studies reported no information about their funding sources nor provided declarations about conflicts of interest.

Excluded studies

Because the searches in this review were designed to feed into a series of related Cochrane Reviews about the surgical management of hip fracture, we have not included a bibliographic list of all excluded studies. We excluded most studies because they were study designs that were not eligible for inclusion in this review, or were not treating participants with the type of fractures or with the types of interventions that were eligible for this review. Some of the excluded studies were eligible for inclusion in the related Cochrane Reviews.

Here, we report the details of eight key excluded studies (see Characteristics of excluded studies). We excluded five studies because they were abstracts with insufficient detail on the numbers of participants in each group, meaning extraction of outcome data was not feasible (Karpman 1992; Kavcic 2006; Rosen 1992; Stock 1997; Van Thiel 1988). We excluded one study that appeared to be randomised, but on closer inspection, we believed was not randomised (Somashekar 2013). We excluded one study that investigated the surgical approach rather than the type of intervention (Aydin 2009). We excluded one study from our clinical trials register search which was abandoned because of lack of funding, and its results are not reported (ISRCTN42349821).

Studies awaiting classification

We found four studies from the search of clinical trials registries that were registered as completed but do not have a published study report in the literature (NCT00800124; NCT00859378; NCT01432691; NTR1782). These studies potentially recruited 1204 participants and investigated the following comparison groups: cemented HA versus uncemented HA (NCT00800124; NCT00859378; NTR1782), and THA versus HA (NCT01432691). See Characteristics of studies awaiting classification.

Ongoing studies

We found seven ongoing studies (ChiCTR1800019531; ISRCTN15606075; NCT01109862; NCT01578408; NCT01787929; UMIN000011303; Wolf 2020). These studies have an estimated enrolment of 7199 participants, and evaluate the following comparison groups: THA versus HA (ChiCTR1800019531; NCT01109862; UMIN000011303), cemented HA versus uncemented HA (NCT01787929), cemented THA versus uncemented THA (NCT01578408), dual mobility THA versus standard THA (Wolf 2020), and single versus dual antibiotic cement HA (ISRCTN15606075). See Characteristics of ongoing studies.

Risk of bias in included studies

We conducted risk of bias according to outcomes relevant to this review. Blank spaces in the risk of bias figure for some detection bias domains indicate that risk of bias assessment was not applicable as the outcome category was not reported. See Figure 2.

Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Blank spaces indicate that risk of bias was not conducted because study authors did not report outcomes relevant to these domains.

Allocation

Twenty studies reported an adequate method to randomise participants to groups, and we judged these studies be at low risk of selection bias for random sequence generation (Calder 1996; Cao 2017; Chammout 2017; Chammout 2019; Davison 2001; Fernandez 2022; Figved 2009; Figved 2018; Griffin 2016; HEALTH 2019; Kim 2012; Lim 2020; Macaulay 2008; Parker 2020; Raia 2003; Rashed 2020; Sims 2018; Taylor 2012; Van den Bekerom 2010; Xu 2017). We judged 11 studies to be at high risk of selection bias for random sequence generation because study authors used quasi‐randomised methods to randomise participants to intervention groups or because other information within the study report indicated that selection bias may have been present (Abdelkhalek 2011; Cornell 1998; Dorr 1986; Iorio 2019; Keating 2006; Livesley 1993; Mouzopoulos 2008; Ravikumar 2000; Santini 2005; Sonaje 2017; Stoffel 2013). The remaining studies reported insufficient information and we judged risk of selection bias for random sequence generation to be unclear.

We judged 11 studies to be at high risk of bias for allocation concealment (Abdelkhalek 2011; Dorr 1986; Iorio 2019; Keating 2006; Livesley 1993; Mouzopoulos 2008; Ravikumar 2000; Santini 2005; Sharma 2016; Sonaje 2017; Stoffel 2013). Twenty‐four studies reported insufficient information and we judged risk of selection bias for allocation to be unclear (Baker 2006; Blomfeldt 2007; Brandfoot 2000; Cadossi 2013; Calder 1995; Calder 1996; Cao 2017; Chammout 2017; Chammout 2019; Cornell 1998; Davison 2001; DeAngelis 2012; Emery 1991; Harper 1994; Jeffcote 2010; Malhotra 1995; Moroni 2002; Patel 2008; Raia 2003; Ren 2017; Sadr 1977; Sonne‐Holm 1982; Talsnes 2013; Vidovic 2013). The remaining studies reported sufficient information, and we judged these to be at low risk of selection bias.

Blinding

It is not possible to blind the operating surgeon to the type of arthroplasty or the implant fixation methods used in these studies. In making judgements about performance bias, we considered whether surgeons were equally experienced with the types of implants used in their study. We judged only 19 studies to report this sufficiently well and we assessed these studies to be at low risk of performance bias (Baker 2006; Blomfeldt 2007; Chammout 2017; Hedbeck 2011; Inngul 2015; Jeffcote 2010; Keating 2006; Kim 2012; Livesley 1993; Moerman 2017; Movrin 2020; Parker 2010c; Parker 2012; Parker 2019; Parker 2020; Ren 2017; Santini 2005; Sharma 2016; Taylor 2012). We expected that all surgeons would aim for the same standard of performance when carrying out all surgical procedures. However, unless otherwise stated, we could not be certain in the remaining studies that surgeons were equally experienced with the prostheses and we judged risk of performance bias in these studies to be unclear.

For detection bias, we considered the type of outcome and who was measuring it. We found that most studies did not report who measured clinically‐assessed outcomes that may be influenced by subjective decisions (such as performance in ADL, hip function or unplanned return to theatre). In these cases, we assumed that these judgements were made by surgeons who were aware of the intervention, which could influence their decision‐making. Thus, we judged detection bias for these clinically‐assessed subjective outcomes to be at high risk of bias. Only six studies reported that assessment of these outcomes was made by personnel who were unaware of treatment, and we judged these six studies to be at low risk of bias for these outcomes (Cornell 1998; Kim 2012; Parker 2012; Sims 2018; Sonne‐Holm 1982; Taylor 2012). Although some participants may have been aware of the type of intervention used during their surgery, we did not expect that knowledge of this would influence how they reported information that contributed towards outcome data such as mobility, pain, and HRQoL. We therefore judged risk of detection bias for all participant‐reported outcomes to be at low risk. We also judged detection bias to be at low risk for objective outcomes (such as mortality and length of stay), and we therefore judged all studies reporting these outcomes to be at low risk of detection bias.

Incomplete outcome data

Because of the high mortality in study population, we expected a large and unavoidable loss of outcome data from participants for outcomes measured with a longer follow‐up. We judged most studies to be at low risk of attrition bias because other losses were few, were well‐explained by study authors, and were balanced between groups. We judged only seven studies to be at high risk of attrition bias (Cadossi 2013; Calder 1995; Fernandez 2022; HEALTH 2019; Inngul 2015; Lim 2020; Moerman 2017). In these seven studies, we noted a large number of losses for outcomes reported at the end of follow‐up (such as HRQoL or functional status) that could not be explained by death. In four studies, we could not be certain if data were complete, particularly for outcomes reported at the end of study follow‐up (DeAngelis 2012; Dorr 1986; Malhotra 1995; Moroni 2002), and in two studies we could not be certain if the number of losses were balanced between intervention groups because of limited information in the study report (Stoffel 2013; Van den Bekerom 2010). We judged risk of attrition bias in these six studies to be unclear.

Selective reporting

Most studies did not report whether they were registered with a clinical trials register and did not provide details of protocols published prior to the study. Nine studies were registered retrospectively with a clinical trials register (Chammout 2017; Chammout 2019; DeAngelis 2012; Figved 2009; Figved 2018; Inngul 2015; Kanto 2014; Parker 2019; Parker 2020). Two reported registration with a clinical trials register, but because they did not report registration numbers, we were unable to source the trials register documents (Talsnes 2013; Taylor 2012). It is not feasible to effectively assess risk of selective reporting bias without these documents, and we judged risk of bias in all of these studies to be unclear.

Only five studies reported prospective clinical trials registration, with outcomes listed in these documents which were consistent with those measured and reported in the study report (Fernandez 2022; Griffin 2016; HEALTH 2019; Moerman 2017; Sims 2018). We judged these four studies to be at low risk of selective reporting bias.

Other potential sources of bias

Two studies were reported as abstracts, with limited study details, and we judged these to be at high risk of other bias because the reports were not peer‐reviewed (Moroni 2002; Patel 2008). We judged other bias to be unclear in another study in which the study methods had limited detail and we could not be certain of bias (Ren 2017). We identified no other sources of bias in the remaining studies.

Effects of interventions

See: Summary of findings 1 Cemented versus uncemented hemiarthroplasty for hip fracture in adults; Summary of findings 2 Bipolar hemiarthroplasty compared with unipolar hemiarthroplasty for hip fracture in adults; Summary of findings 3 Total hip arthroplasty compared with hemiarthroplasty for hip fracture in adults

We report data available at three possible time points: early (within four months of surgery), 12 months (after four months and up to 24 months after surgery, prioritising data at 12 months if possible), and late (more than 24 months after surgery, at the latest time point reported by study authors). In the following, we report subgroup and sensitivity analyses only for comparisons for which they were appropriate and possible.

1. Prostheses implanted with cement versus without cement

Here we present three comparisons of cemented and uncemented prostheses ‐ stratified by the categories THA, HA, and a mixed intervention of THA and HA (participants were treated with either a THA or HA which is cemented, or a THA or HA which is uncemented). A summary of the implant and study characteristics is presented in Table 3. For outcomes measured with scales, we present range of scores and direction of effect for each scale in Appendix 3.

THA: cemented versus uncemented

This comparison includes data from only one study with 69 participants (Chammout 2017).

Here we report the effects for critical outcomes, and we summarise the effects of other important outcomes in Table 8. All outcomes in this comparison are reported without GRADE assessments.

Open in table viewer

Table 8. THA (cemented vs uncemented): effects of other important outcomes and adverse events

Other important outcomes
	Number of studies	Studies	Participants	Effect estimate
Pain at ≤ 4 months Using Pain Numerical Rating Score (range of scores from 0 to 11; lower scores indicate less pain)	1	Chammout 2017	64	MD ‐0.90, 95% CI ‐1.82 to 0.02 (favours cemented); Analysis 1.6
Pain at 12 months Using Pain Numerical Rating Score (range of scores from 0 to 11; lower scores indicate less pain)	1	Chammout 2017	63	MD 1.00, 95% CI 0.03 to 1.97 (favours uncemented); Analysis 1.6
Adverse events related to implant or fracture, or both
Intraoperative periprosthetic fracture	1	Chammout 2017	69	RR 0.14, 95% CI 0.01 to 2.59 (favours cemented); Analysis 1.7
Postoperative periprosthetic fracture	1	Chammout 2017	69	RR 0.97, 95% CI 0.06 to 14.91 (favours cemented); Analysis 1.7
Loosening	1	Chammout 2017	69	RR 0.32, 95% CI 0.01 to 7.69 (favours cemented); Analysis 1.7
Superficial infection	1	Chammout 2017	69	RR 0.32, 95% CI 0.01 to 7.69 (favours cemented); Analysis 1.7
Dislocation	1	Chammout 2017	69	RR 0.32, 95% CI 0.04 to 2.96 (favours cemented); Analysis 1.7

CI: confidence interval
MD: mean difference
RR: risk ratio

Critical outcomes

Activities of daily living (ADL)

Chammout 2017 provided scores for the ability to perform ADL but did not report the measurement tool used to score this data. We found no evidence of a difference in performing ADL at 3 months (mean difference (MD) 0.00, 95% confidence interval (CI) ‐0.17 to 0.17; 1 study, 65 participants; Analysis 1.1) and at 12 months after surgery (MD 0.10, 95% CI ‐0.22 to 0.42, favours uncemented; 1 study, 63 participants; Analysis 1.1).

Delirium

Chammout 2017 did not report this outcome.

Functional status

Chammout 2017 reported functional status measured with the Harris Hip Score (HHS), with higher scores indicating better function. We found no evidence of a difference in function at 3 months after surgery (MD 1.00, 95% CI ‐5.37 to 7.37, favours cemented; 1 study, 65 participants; Analysis 1.2) and at 12 months after surgery (MD ‐3.00, 95% CI ‐11.29 to 5.29, favours uncemented; 1 study, 65 participants; Analysis 1.2).

Health‐related quality of life (HRQoL)

Chammout 2017 also reported HRQoL, measured using the EQ‐5D (range of scores from 0 to 1; higher scores indicate better quality of life). We found no evidence of a difference in HRQoL at 3 months after surgery (MD 0.00, 95% CI ‐0.12 to 0.12; 1 study, 64 participants; Analysis 1.3) and at 12 months after surgery (MD 0.00, 95% CI ‐0.13 to 0.13; 1 study, 62 participants; Analysis 1.3).

Mobility

Chammout 2017 did not report this outcome.

Mortality

We found no evidence of a difference in mortality at 12 months (risk ratio (RR) 0.49, 95% CI 0.05 to 5.11, favours cemented; 1 study, 69 participants; Analysis 1.4).

Unplanned return to theatre

We found no evidence of a difference in return to theatre at 24 months (RR 0.11, 95% CI 0.01 to 1.93, favours cemented; 1 study, 69 participants; Analysis 1.5). Some re‐operations were because of dislocation, subsidence or pain. We noted that types of re‐operation included revision to HA and to change the liner to an elevated rim.

Other important outcomes

Effect estimates were imprecise; we found no evidence of a difference in pain or adverse events (intra‐ or postoperative periprosthetic fracture, loosening, superficial infection, and dislocation). We report the summary effects of important outcomes and adverse effects in Table 8.

HA: cemented versus uncemented

This comparison includes data from 17 studies with 3644 participants (Brandfoot 2000; Cao 2017; DeAngelis 2012; Emery 1991; Fernandez 2022; Figved 2009; Harper 1994; Moerman 2017; Movrin 2020; Parker 2010c; Rehman 2014; Sadr 1977; Santini 2005; Sonne‐Holm 1982; Talsnes 2013; Taylor 2012; Vidovic 2013). We analysed the data from Cao 2017 separately because this study includes only participants with extracapsular fractures.

Here we report effects for critical outcomes. Where pooled analyses included at least one study in each category, we subgrouped the analysis according to whether studies used a modern or a first‐generation, uncemented stem design in one of the intervention groups. Of the 16 studies including participants with intracapsular fractures, eight reported a comparison between cemented and modern uncemented HAs.

We used GRADE to assess the certainty of the evidence for the critical outcomes measured within four months of surgery (ADL, functional status, HRQoL, and mobility), within four months and at 12 months for mortality, and at the end of follow‐up for delirium and unplanned return to theatre. See summary of findings Table 1.

We summarise the effects of other important review outcomes in a table, which are not subgrouped according to the stem design; these outcomes are reported without GRADE assessments.

Critical outcomes

ADL

Seven studies reported performance of ADL. The uncemented stem designs in these studies were modern (DeAngelis 2012; Fernandez 2022; Figved 2009; Moerman 2017; Parker 2020), first generation (Brandfoot 2000), or the type of stem was unknown (Santini 2005).

Early:

Moerman 2017 used the Groningen Activity Restriction Scale (GARS) at four months, and Parker 2020 used a social dependency scale at four months. For both instruments, lower scores indicate more independence. DeAngelis 2012 used the Older Americans Resources Scale of Instrumental Activities of Daily Living (OARS‐IADL), and Fernandez 2022 used a five‐point Likert scale for 'usual activities' derived from the EQ‐5D utility index; we inverted the data in these studies to account for these instruments in which higher scores indicate more independence. We found no evidence of a difference in performance of ADL (SMD ‐0.03, 95% CI ‐0.21 to 0.16, favours cemented; 4 studies, 1275 participants; I² = 53%; moderate‐certainty evidence; Analysis 2.1). We downgraded the certainty of the evidence by one level for study limitations because studies had unclear risks of bias.
In addition, Figved 2009 reported the number of people who were independent at three months, defined as those scoring 19 or 20 on the Barthel Index Score. From these data, we found no evidence of a difference in performance of ADL (RR 0.88, 95% CI 0.65 to 1.19, favours uncemented; 1 study, 190 participants; Analysis 2.2).

At 12 months:

Moerman 2017 used the GARS and Parker 2020 used a social dependency scale. Santini 2005 used the Verona Elderly Care (VELCA) scoring system, DeAngelis 2012 used the OARS‐IADL, and Fernandez 2022 used a five‐point Likert scale; we inverted the data for these studies to account for those instruments in which higher scores indicate more independence. We found no evidence of a difference in performance of ADL (SMD ‐0.09, 95% CI ‐0.21 to 0.02, favours cemented; 5 studies, 1173 participants; I² = 0%; Analysis 2.3). Data in this analysis were reported at 12 months in all studies.
Figved 2009 reported the number of people who were independent at 12 months, defined as those scoring 19 or 20 on the Barthel Score. From these data, we found no evidence of a difference in performance of ADL (RR 0.79, 95% CI 0.61 to 1.04, favours uncemented; 1 study, 168 participants; Analysis 2.4).
In addition, Brandfoot 2000 reported ADL outcome at 16 months using a modification of the HHS, extracting responses from the instrument related to using stairs, socks, shoes, and bathing. We did not calculate an effect estimate because data were reported as a measure of the variance. See Appendix 4 for mean scores as reported by study authors.

Late:

Figved 2009 reported the number of people who were independent at five years, defined as those scoring 19 or 20 on the Barthel Score. From these data, we found no evidence of a difference in performance of ADL (RR 0.87, 95% CI 0.63 to 1.21, favours uncemented; 1 study, 80 participants; Analysis 2.5).

Delirium

Two studies reported data for delirium. The uncemented stem designs in these studies were modern (Parker 2020), and first generation (Parker 2010c). We found no evidence of a difference for this outcome (RR 1.06, 95% CI 0.55 to 2.06, favours uncemented; 2 studies, 800 participants; I² = 0%; low‐certainty evidence; Analysis 2.6). We downgraded the certainty of the evidence by two levels ‐ one level for imprecision because we noted a wide CI in the estimate, and one level for study limitations because the studies had unclear risks of bias.

Functional status

Eight studies reported data for functional status. The uncemented stem designs in these studies were modern (Cao 2017; Figved 2009; Movrin 2020), first generation (Brandfoot 2000; Sadr 1977; Sonne‐Holm 1982; Vidovic 2013), or unknown (Santini 2005).

Early:

Three studies reported function using the HHS (Figved 2009; Movrin 2020; Vidovic 2013). For this instrument, higher scores indicate better function. We found improved function with cemented HAs (MD 3.38, 95% CI 0.05 to 6.70, favours cemented; 3 studies, 416 participants; very low‐certainty; Analysis 2.7). We noted that this estimate did not indicate a clinically important improvement (based on a minimal clinically important difference (MCID)) of 16 to 18 points (Singh 2016). We downgraded the certainty of the evidence by one level for imprecision because the evidence included few studies, and two levels for study limitations because the studies had unclear risks of bias, and we found during sensitivity analysis that the effect estimate was influenced by these studies (see below).
In addition, Sonne‐Holm 1982 reported the number of participants with a maximum score on the D'Aubigne scale indicating good functional status (D'Aubigne 1954. From these data, we found no evidence of a difference in functional status (RR 1.15, 95% CI 0.84 to 1.59, favours cemented; 1 study, 75 participants; Analysis 2.8).
Cao 2017 reported function using the HHS at three months, and we found improved function with cemented HAs for extracapsular fractures (MD 14.70, 95% CI 11.78 to 17.62, favours cemented; 1 study, 85 participants; Analysis 2.9). We noted that the CI in this estimate may indicate a clinically important improvement with cemented HAs based on a MCID of 16 to 18 points (Singh 2016).

At 12 months:

Three studies reported this outcome using the HHS (Figved 2009; Movrin 2020; Vidovic 2013), Taylor 2012 used the Oxford Hip Score (OHS), and Santini 2005 used the VELCA scoring system. For all instruments, higher values indicate better function. The estimate was imprecise, including clinically relevant benefits in favour of cemented implants but also no difference between interventions (SMD 0.13, 95% CI ‐0.09 to 0.35, favours cemented; 5 studies, 494 participants; I² = 31%; Analysis 2.10). We pooled data reported at 12 months (Figved 2009; Santini 2005; Taylor 2012; Vidovic 2013), and at 24 months (Movrin 2020).
Sadr 1977 reported data using the HHS measured at 17 months; scores were categorically reported as excellent, good, medium or poor. We combined the good and excellent scores with maximum scores from Sonne‐Holm 1982, reported at 12 months. We found no evidence of a difference in functional status (RR 1.15, 95% CI 0.91 to 1.45, favours cemented; 2 studies, 100 participants; I² = 0%; Analysis 2.11).
Cao 2017 reported function using the HHS at 6 months, and we found improved function with cemented HAs for extracapsular fractures (MD 11.09, 95% CI 7.70 to 14.48, favours cemented; 1 study, 85 participants; Analysis 2.12). We noted that the CI is unlikely to be compatible with a clinically important improvement (Singh 2016).
In addition, Brandfoot 2000 reported this outcome at 16 months using the HHS. We did not pool data from this study in the analysis because the data were reported without any measures of variance. See Appendix 4 for mean scores as reported by study authors.

Late:

Figved 2009 also reported functional status outcome at five years, but the estimate was imprecise (MD ‐9.90, 95% CI ‐17.75 to ‐2.05, favours uncemented; 1 study, 78 participants; Analysis 2.13).

HRQoL

Three studies reported HRQoL (Fernandez 2022; Figved 2009; Moerman 2017). These studies used modern, uncemented stem designs. We extracted data for the physical component of the Short‐Form 12 (SF‐12) for Moerman 2017, and EQ‐5D for Fernandez 2022 and Figved 2009. In both scales, higher scores indicate better quality of life.

Early:

We found improved HRQoL for cemented HAs (SMD 0.20, 95% CI 0.07 to 0.34, favours cemented; 3 studies, 1122 participants; I² = 9%; moderate‐certainty evidence; Analysis 2.14). The outcome was measured at four months in Fernandez 2022 and at three months in the other studies. After converting this effect estimate onto the EQ‐5D utility scale, the difference between fixation techniques was compatible with clinically small and large benefits of cemented HAs (0.06, 95% CI 0.02 to 0.10); this was based on a MCID for EQ‐5D of 0.07 (Walters 2005). We downgraded the certainty of the evidence by one level for study limitations because the studies had high and unclear risks of bias.

At 12 months:

We found improved HRQoL for cemented HAs (SMD 0.12, 95% CI 0.00 to 0.24, favours cemented; 3 studies, 1079 participants; I² = 0%; Analysis 2.15). After converting this effect estimate onto the EQ‐5D utility scale, the difference was compatible with no effect or a clinically important benefit of cemented HAs (0.03, 95% CI 0.00 to 0.07); this was based on a MCID for EQ‐5D of 0.07 (Walters 2005).

Late:

Figved 2009 also reported HRQoL at five years, with no evidence of a difference between fixation techniques (MD ‐0.09, 95% CI ‐0.23 to 0.05, favours uncemented; 1 study, 71 participants; I² = 0%; Analysis 2.16).

Mobility

Eleven studies reported data for mobility. The uncemented stem designs in these studies were modern (Fernandez 2022; Figved 2009; Moerman 2017; Parker 2020; Taylor 2012), first generation (Brandfoot 2000; Emery 1991; Parker 2010c; Rehman 2014; Sonne‐Holm 1982), and unknown (Santini 2005).

Early:

Three studies reported the proportion of people who were able to walk independently at three months (Figved 2009; Sonne‐Holm 1982), and four months (Fernandez 2022). We found no evidence of a difference in mobility (RR 1.04, 95% CI 0.95 to 1.14, favours cemented; 3 studies, 980 participants; I² = 5%; moderate‐certainty evidence; Analysis 2.17). We downgraded the certainty of the evidence by one level for study limitations because included studies had unclear risks of bias.
Parker 2010c and Parker 2020 used a nine‐point mobility scale in which lower scores indicate better mobility, and Moerman 2017 used a nine‐point mobility scale in which higher scores indicate better mobility. We inverted the data in Moerman 2017 before pooling. We found that mobility was improved with cemented prostheses (SMD ‐0.26, 95% CI ‐0.40 to ‐0.12, favours cemented; 3 studies, 766 participants; I² = 0%; Analysis 2.18). This effect size is likely to be small to medium (Cohen 1988).
Rehman 2014 used a nine‐point mobility rating scale, in which higher scores indicated better mobility; these data were reported as mean reduction values. We found that mobility was improved with uncemented HA (MD ‐0.40, 95% CI ‐0.68 to ‐0.12, favours uncemented; 1 study, 110 participants; Analysis 2.19). This effect estimate was imprecise, including clinically relevant benefits and harms (Cohen 1988).

At 12 months:

Parker 2010c and Parker 2020 used a nine‐point scale in which lower scores indicate better mobility. Moerman 2017 used a nine‐point scale and Santini 2005 reported a six‐point subscale for walking abilities from the VELCA scoring system; for both scales, higher scores indicate better mobility. We inverted the data in Moerman 2017 and Santini 2005 before pooling. We found that mobility was improved with a cemented HA (SMD ‐0.24, 95% CI ‐0.42 to ‐0.06, favours cemented; 4 studies, 762 participants; I² = 32%; Analysis 2.20). This effect size is likely to be small to medium (Cohen 1988).
Three studies reported the proportion of people who were able to walk independently at 12 months (Fernandez 2022; Figved 2009; Sonne‐Holm 1982). We found no evidence of a difference in mobility (RR 0.98, 95% CI 0.70 to 1.37, favours uncemented; 3 studies, 826 participants; I² = 84%; Analysis 2.21).
Emery 1991 reported the number of people who were more dependent on walking aids at 17 months after surgery than before their injury. We found that mobility was better using a cemented HA (RR 0.53, 95% CI 0.30 to 0.93, favours cemented; 1 study, 39 participants; Analysis 2.22).
In addition, Brandfoot 2000 reported this outcome at 16 months, using responses extracted from the HHS, and Taylor 2012 reported this using the Timed Up and Go (TUG) test at 24 months. We did not pool data from these studies in the analyses because the data were reported without variances. See Appendix 4 for mean scores, as reported by study authors.

Late:

Parker 2010c reported data at five years, using a nine‐point mobility scale in which lower scores indicate better mobility. We found no evidence of a difference in mobility (MD ‐0.60, 95% CI ‐1.79 to 0.59, favours cemented; 1 study, 64 participants; Analysis 2.23).
Figved 2009 reported the number of people who were able to walk independently at five years. We found no evidence of a difference in mobility (RR 0.88, 95% CI 0.75 to 1.02, favours uncemented; 1 study, 79 participants; Analysis 2.24).

Mortality

Fifteen studies reported mortality. The uncemented stem designs in these studies were modern (Fernandez 2022; Figved 2009; Moerman 2017; Movrin 2020; Parker 2020; Talsnes 2013; Taylor 2012), first generation (Brandfoot 2000; Emery 1991; Harper 1994; Parker 2010c; Sadr 1977; Sonne‐Holm 1982; Vidovic 2013), or unknown (Santini 2005).

Early:

The estimate for mortality within four months of surgery includes clinically relevant harms and benefits (RR 0.95, 95% CI 0.80 to 1.13, favours cemented; 12 studies, 3136 participants; I² = 0%; low‐certainty evidence; Analysis 2.25). We downgraded the certainty of the evidence by one level for imprecision because the CI included both possible harms and benefits, and one level for study limitations because most studies in this analysis had unclear or high risks of bias. We generated a funnel plot (Figure 3), and we found evidence of small study effects which tend to favour cemented HAs (Harbord modified test, P value = 0.003).

Figure 3

Cemented hemiarthroplasty versus uncemented hemiarthroplasty. Funnel plot for early mortality (≤ 4 months), subgrouped by stem design

At 12 months:

We found that the risk of death at 12 months was reduced using cemented HA (RR 0.86, 95% CI 0.78 to 0.96, favours cemented; 15 studies, 3727 participants; I² = 0%; moderate‐certainty evidence; Analysis 2.26). This analysis included data reported at 16 months (Brandfoot 2000), 18 months (Emery 1991), and 24 months (Movrin 2020). We downgraded the certainty of the evidence by one level for study limitations because most studies in this analyses had unclear or high risks of bias. We generated a funnel plot (Figure 4), and we found no statistical evidence of small study size effects (Harbord modified test, P value = 0.169).

Figure 4

Cemented hemiarthroplasty versus uncemented hemiarthroplasty. Funnel plot for mortality at 12 months, subgrouped by stem design

Late:

Two studies reported data at a time point of five years (Figved 2009; Parker 2010c). We found no evidence of a difference in mortality according to the fixation technique (RR 1.01, 95% CI 0.89 to 1.25, favours uncemented; 2 studies, 620 participants; Analysis 2.27).

Unplanned return to theatre

Six studies reported unplanned return to theatre at the end of study follow‐up, which was at 12 months (DeAngelis 2012; Fernandez 2022; Figved 2009; Moerman 2017), 24 months (Taylor 2012), and 60 months (Parker 2010c). The effect estimate was imprecise, including benefits and harms (RR 0.70, 95% CI 0.45 to 1.10, favours cemented; 6 studies, 2336 participants; I² = 0%; low‐certainty evidence; Analysis 2.28). Some re‐operations were because of dislocation, loosening, acetabular wear, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, Girdlestone and open reduction and drainage of infection. We downgraded the certainty of the evidence by two levels for study limitations because most studies in the analysis had unclear risks of bias and all studies were at high risk of detection bias.

Other important outcomes

It was difficult to interpret pain outcomes because studies reported this outcome using different instruments and scales. We found no evidence of a difference in the number of people who were discharged to their own home according to the fixation technique. Similarly, we found no evidence of a difference in most adverse events unrelated to the implant or fracture, or both, according to whether or not cement was used to fix the HA (acute kidney injury, cerebrovascular accident, chest infection/pneumonia, myocardial infarction, urinary tract infection, deep vein thrombosis, and pulmonary infection). However, we noted that fewer people had a pulmonary embolism when the HA was fixed without cement. Cao 2017 reported no adverse events in either group of participants with extracapsular fractures (intraoperative fractures, loosening, deep infection, superficial infection, and dislocation). We report the summary effects of all these important outcomes and adverse effects in Table 9.

Open in table viewer

Table 9. HA (cemented vs uncemented): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain^a Experiencing no pain at ≤ 4 months (We inverted data in 2 studies in which data were reported as complaining of pain or experiencing mid‐thigh pain)	4	Harper 1994; Figved 2009; Moerman 2017; Sonne‐Holm 1982	500	RR 1.11, 95% CI 1.00 to 1.22 (favours uncemented); Analysis 2.29
Pain at ≤ 4 months Using VAS, and a 9‐point pain scale (lower values indicate less pain)	3	Movrin 2020; Parker 2010c; Parker 2020	802	Data not combined because of substantial statistical heterogeneity I² = 91%; Analysis 2.30
Pain^a Experiencing no pain at 12 months (We inverted data in 1 study in which data were reported as complaining of pain or experiencing mid‐thigh pain)	4	Emery 1991; Figved 2009; Moerman 2017; Sonne‐Holm 1982	376	RR 1.17, 95% CI 0.85 to 1.63 (favours uncemented); I² = 77%; Analysis 2.31
Pain at 12 months Using VAS, and a 9‐point pain scale (lower values indicate less pain)	4	Figved 2009; Movrin 2020; Parker 2010c; Parker 2020	726	SMD ‐0.06, 95% CI ‐0.33 to 0.21 (favours cemented); I² = 66%; Analysis 2.32
Pain at 12 months Mean reduction values (lower values indicate less pain)	1	Rehman 2014	110	MD ‐0.27, 95% CI ‐0.48 to ‐0.06 (favours cemented); Analysis 2.33
Pain at > 24 months. Reported by study authors at 5 years Using VAS (lower values indicate less pain)	1	Parker 2010c	58	MD ‐0.30, 95% CI ‐0.92 to 0.32 (favours cemented); Analysis 2.34
Pain Experiencing no pain at 5 years	1	Figved 2009	80	RR 1.00, 95% CI 0.77 to 1.30; Analysis 2.35
Length of hospital stay	9	Emery 1991; Figved 2009; Harper 1994; Moerman 2017; Parker 2010c; Parker 2020; Santini 2005; Taylor 2012; Vidovic 2013	1801	MD ‐0.40 days, 95% CI ‐1.03 to 0.23 (favours cemented); Analysis 2.36
Discharge destination Living in own home^a	6	DeAngelis 2012; Figved 2009; Parker 2010c; Santini 2005; Taylor 2012; Fernandez 2022	2331	RR 1.05, 95% CI 0.98 to 1.13 (favours uncemented); Analysis 2.37
Adverse events related to surgery
Intraoperative periprosthetic fracture	7	DeAngelis 2012; Figved 2009; Moerman 2017; Movrin 2020; Parker 2010c; Parker 2020; Taylor 2012	1669	RR 0.20, 95% CI 0.08 to 0.46 (favours cemented); Analysis 2.38
Postoperative periprosthetic fracture	6	Figved 2009; Moerman 2017; Movrin 2020; Santini 2005; Taylor 2012; Fernandez 2022	2819	RR 0.29, 95% CI 0.14 to 0.57 (favours cemented); Analysis 2.38
Loosening	4	Brandfoot 2000; Figved 2009; Moerman 2017; Sadr 1977	537	RR 0.52, 95% CI 0.14 to 1.89 (favours cemented); I² = 45%; Analysis 2.38
Deep infection	7	Figved 2009; Harper 1994; Moerman 2017; Movrin 2020; Parker 2010c; Santini 2005; Taylor 2012	1382	RR 1.56, 95% CI 0.72 to 3.38 (favours uncemented); Analysis 2.38
Superficial infection	7	DeAngelis 2012; Emery 1991; Figved 2009; Harper 1994; Moerman 2017; Parker 2010c; Parker 2020; Sonne‐Holm 1982; Taylor 2012; Fernandez 2022	1210	RR 1.23, 95% CI 0.73 to 2.06 (favours uncemented); Analysis 2.38
Dislocation	8	Figved 2009; Harper 1994; Moerman 2017; Movrin 2020; Parker 2010c; Parker 2020; Sadr 1977; Santini 2005; Taylor 2012; Fernandez 2022	3032	RR 1.08, 95% CI 0.61 to 1.91 (favours uncemented); Analysis 2.38
Adverse events unrelated to surgery
Acute kidney injury	4	Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2226	RR 1.23, 95% CI 0.76 to 2.00 (favours uncemented); Analysis 2.39
Blood transfusion	7	DeAngelis 2012; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Talsnes 2013; Fernandez 2022	2907	RR 1.00, 95% CI 0.83 to 1.20 (favours cemented); I² = 36%; Analysis 2.39
Cerebrovascular accident	5	DeAngelis 2012; Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2356	RR 0.93, 95% CI 0.41 to 2.10 (favours cemented); Analysis 2.39
Chest infection/pneumonia	8	DeAngelis 2012; Emery 1991; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Taylor 2012; Fernandez 2022	2789	RR 0.78, 95% CI 0.50 to 1.21 (favours cemented); Analysis 2.39
Myocardial infarction	7	DeAngelis 2012; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Santini 2005; Fernandez 2022	1457	RR 0.91, 95% CI 0.44 to 1.89 (favours cemented); Analysis 2.39
Urinary tract infection	5	Emery 1991; Moerman 2017; Santini 2005; Taylor 2012; Fernandez 2022	1745	RR 0.89, 95% CI 0.65 to 1.20 (favours cemented); Analysis 2.39
Venous thromboembolic phenomena (DVT)	7	Cao 2017; DeAngelis 2012; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2661	RR 1.28, 95% CI 0.56 to 2.90 (favours uncemented); Analysis 2.39
Venous thromboembolic phenomena (pulmonary embolism)	6	Emery 1991; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2499	RR 3.56, 95% CI 1.26 to 10.11 (favours uncemented); Analysis 2.39

^aOther data is reported in Appendix 5

CI: confidence interval
DVT: deep vein thrombosis
MD: mean difference
RR: risk ratio
VAS: visual analogue scale

Subgroup analysis

We did not conduct subgroup analysis to explore differences between studies according to our pre‐specified effect modifiers (age, gender, and fracture displacement) because these variables were insufficiently reported.

Visual inspection of the forest plots for each outcome where data were available in both first generation and modern uncemented HA subgroups showed that the overall trend for benefits associated with cemented HA were reduced in the modern uncemented subgroup. However, these subgroups were sparse and few studies were available across most of the comparisons. We conducted a formal subgroup analysis for mortality at 12 months because this analysis had sufficient studies. Santini 2005 did not report the exact type of uncemented HA, and we chose to include this study in the first generation subgroup. On visual inspection of the data, we noted that the reduction in mortality was increased amongst studies reporting a comparison with a modern uncemented design of HA rather than a first generation uncemented design (RR 0.80, 95% CI 0.68 to 0.95; 6 studies, 1466 participants, favours cemented modern HA; I² = 0%; Analysis 2.26; Figure 4). However, this was not supported by formal tests of interaction (P = 0.24).

Sensitivity analysis

Here, we report the results of sensitivity analyses only when we noted a difference in interpretation of the effect.

High or unclear risk of selection bias (for sequence generation)

Early functional status (≤ four months; continuous data): we excluded two studies from this analysis (Movrin 2020; Vidovic 2013). Although the effect continued to show no evidence of a difference between groups, we noted that the estimate favoured the alternative intervention (MD ‐1.20, 95% CI ‐6.66 to 4.26, favours uncemented; 1 study, 189 participants).
Functional status (at 12 months; continuous data): we excluded three studies from this analysis (Movrin 2020; Santini 2005; Vidovic 2013). Although the effect continued to show no evidence of a difference between groups, we noted that the estimate favoured the alternative intervention (MD ‐0.02, 95% CI ‐0.28 to 0.24, favours uncemented; 2 studies, 234 participants).
Early mobility (reported at ≤ four months; continuous data): we excluded two studies from this analysis (Moerman 2017; Parker 2010c). We found that the analysis no longer showed an improvement in mobility when the prosthesis was cemented (MD ‐0.40, 95% CI ‐0.81 to 0.01; 1 study, 329 participants).

High risk of attrition bias

In this sensitivity analysis, we considered attrition bias at the outcome level. We therefore conducted sensitivity analysis only on outcomes that included a study with high risk of attrition bias owing to losses for that specific outcome. We found no difference in the interpretation of the effect for all outcomes in this comparison.

Mixed HA and THA: cemented versus uncemented

This comparison includes data from two studies with 169 participants (Inngul 2015; Moroni 2002). In both studies, participants were randomised to a cemented or uncemented prosthesis, but the selection of a THA or HA was left to the treating surgeon and participant to select.

Here we report the effects for critical outcomes, and we summarise the effects of other important review outcomes in a table. These outcomes are reported without GRADE assessments.

Critical outcomes

ADL, delirium, and mobility

Neither study reported data for these outcomes.

Functional status

Both studies reported functional status measured using the HHS, with higher scores indicating better function.

At 12 months:

We calculated an effect estimate for Moroni 2002 but this estimate was very imprecise. We found no evidence of a difference in functional status at 24 months from surgery (MD ‐16.00, 95% CI ‐41.57 to 9.57, favours uncemented; 1 study, 28 participants; Analysis 3.1).

Inngul 2015 also reported data for this outcome at 4 months, 12 months and 4 years using the HHS, but we could not calculate effect estimates because the study authors did not clearly report the number of participants available in each group. See Appendix 4 for mean scores as reported by study authors.

HRQoL

Only Moroni 2002 reported this outcome, measured using SF‐36 at 24 months. This estimate was very imprecise. We found no evidence of a difference in HRQoL (MD ‐19.00, 95% CI ‐42.77 to 4.77, favours uncemented; 1 study, 28 participants; Analysis 3.2).

Mortality

Both studies reported mortality (Inngul 2015; Moroni 2002).

Early:

The effect estimate for mortality at four months was very imprecise, including clinically relevant harms and benefits (RR 4.42, 95% CI 0.51 to 38.55, favours uncemented; 1 study, 141 participants; Analysis 3.3).

At 12 months:

Similarly, the estimate for mortality at 12 months was imprecise (RR 2.02, 95% CI 0.81 to 5.07, favours uncemented; 2 studies, 169 participants; I² = 0%; Analysis 3.4). Moroni 2002 did not report mortality at 12 months, and this analysis includes data at 24 months from this study.

Late:

Data were available at four years in Inngul 2015, and the estimate was also imprecise (RR 0.88, 95% CI 0.50 to 1.56, favours cemented; 1 study, 141 participants; Analysis 3.5).

Unplanned return to theatre

Only Inngul 2015 reported unplanned return to theatre. The estimate was imprecise, showing no evidence of a difference at four years after surgery (RR 0.74, 95% CI 0.22 to 2.50, favours uncemented; 1 study, 141 participants; Analysis 3.8). Indications for re‐operation were dislocation and periprosthetic fracture, and revision included THA.

Other important outcomes

We found no evidence of a difference in some adverse events related to the implant or fracture, or both (superficial infection and dislocation). However, we noted fewer intraoperative periprosthetic fractures when cement was used in Inngul 2015 (0.06, 95% CI 0.00 to 0.98; 1 study, 141 participants; Analysis 3.9). We found no evidence of a difference in adverse events unrelated to the implant or fracture, or both (acute kidney injury, pneumonia, myocardial infarction, urinary tract infection). We report the summary effects for these adverse events in Table 10.

Open in table viewer

Table 10. THA (mixed HA and THA): cemented vs uncemented: effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Adverse events related to the implant or fracture, or both
Intraoperative periprosthetic fracture	1	Inngul 2015	141	RR 0.06, 95% CI 0.00 to 0.98 (favours cemented); Analysis 3.9
Superficial infection	1	Inngul 2015	141	RR 0.49, 95% CI 0.16 to 1.52 (favours cemented); Analysis 3.9
Dislocation	1	Moroni 2002	28	RR 0.87, 95% CI 0.14 to 5.32 (favours cemented); Analysis 3.9
Adverse events unrelated to implant or fracture, or both
Acute kidney injury	1	Inngul 2015	141	RR 0.37, 95% CI 0.02 to 8.87 (favours cemented); Analysis 3.10
Chest infection/pneumonia	1	Inngul 2015	141	RR 0.55, 95% CI 0.05 to 5.95 (favours cemented); Analysis 3.10
Myocardial infarction	1	Inngul 2015	141	RR 0.37, 95% CI 0.02 to 8.87 (favours cemented); Analysis 3.10
Urinary tract infection	1	Inngul 2015	141	RR 1.42, 95% CI 0.56 to 3.60 (favours uncemented); Analysis 3.10

CI: confidence interval
RR: risk ratio

2. Bipolar HA versus unipolar HA

This comparison includes 13 studies with 1499 participants (Abdelkhalek 2011; Calder 1995; Calder 1996; Cornell 1998; Davison 2001; Figved 2018; Hedbeck 2011; Jeffcote 2010; Kanto 2014; Malhotra 1995; Patel 2008; Raia 2003; Stoffel 2013). A summary of the types of implants and study characteristics is presented in Table 4. Whilst cemented prostheses were reported in most studies in this comparison, three studies reported outcomes with uncemented prostheses (Figved 2018; Malhotra 1995; Patel 2008), and in one study, mixed cemented and uncemented prostheses were included in both groups (Abdelkhalek 2011).

Here we report effects for critical outcomes. Where analyses included at least one study in each category, we subgrouped the analysis according to whether studies reported interventions with cemented or uncemented prostheses.

We used GRADE to assess the certainty of the evidence for the critical outcomes measured within four months of surgery (ADL, functional status, HRQoL, and mobility), within 4 months and at 12 months for mortality, and at the end of follow‐up for delirium and unplanned return to theatre. See summary of findings Table 2.

We summarise the effects of other important review outcomes in a table, which are not subgrouped by stem fixation. These outcomes are reported without GRADE assessments.

For outcomes measured with scales, we present range of scores and direction of effect for each scale in Appendix 3.

Critical outcomes

ADL

Two studies reported performance of ADL; in both studies, the prostheses were cemented (Hedbeck 2011; Raia 2003).

At 12 months:

Hedbeck 2011 used the Katz Index to identify participants who were independent (Katz 1963). We found no evidence of a difference in the number of people who were independent in ADL at 12 months (RR 1.06, 95% CI 0.85 to 1.33, favours bipolar; 1 study, 99 participants; Analysis 4.1).
In addition, Raia 2003 reported this outcome at 12 months using the ADL subset score of the Musculoskeletal Functional Assessment Instrument. We did not calculate an effect estimate because data were reported without distribution variables. See Appendix 5 for average scores as reported by study authors.

Delirium

Stoffel 2013 reported delirium following cemented HAs. We found no evidence of a difference in postoperative delirium (RR 0.48, 95% CI 0.09 to 2.58, favours bipolar; 1 study, 261 participants; very low‐certainty evidence; Analysis 4.2). We downgraded the certainty of the evidence by three levels ‐ two levels for imprecision because the evidence included very few participants, and one level for study limitations because the included study had high and unclear risks of bias.

Functional status

Eight studies reported functional status. Studies included cemented (Cornell 1998; Davison 2001; Hedbeck 2011; Raia 2003; Stoffel 2013), uncemented (Figved 2018; Malhotra 1995), and a mixture of cemented and uncemented HAs in both the bipolar and unipolar groups (Abdelkhalek 2011).

Early:

Hedbeck 2011 reported this outcome using the HHS at four months. We did not calculate an effect estimate for this study because data were reported without measures of variance. See Appendix 5 for mean scores as reported by study authors.

At 12 months:

Two studies reported this outcome, using the HHS (Stoffel 2013), and the Johansen hip score (Cornell 1998). In both scales, higher scores indicate better function. This analysis included data at 12 months (Stoffel 2013), and at 6 months (Cornell 1998). This estimate was imprecise, including clinically relevant benefits and harms; we found no evidence of a difference according to the articulation of the HA (SMD ‐0.04, 95% CI ‐0.27 to 0.19, favours unipolar; 2 studies, 299 participants; I² = 0%; Analysis 4.3).
Malhotra 1995 reported categorical data using the Devas 1983 system. Ranges of scores were reported as excellent, good, medium, or poor, and we combined data for scores which were excellent and good. We found no evidence of a difference according to the articulation of the HA (RR 1.17, 95% CI 0.95 to 1.43, favours bipolar; 1 study, 68 participants; Analysis 4.4).
In addition, four studies reported this outcome using the HHS at 12 months (Davison 2001; Figved 2018; Hedbeck 2011), and physical function scores of SF‐36 (Raia 2003). We did not calculate effect estimates for these studies because data were reported without means or without an appropriate measure of variance. See Appendix 5 for data as reported by study authors.

Late:

One study reported categorical data using the HHS (Abdelkhalek 2011). Ranges of scores were reported as excellent, good, medium, or poor, and we combined data for scores which were excellent and good. We found no evidence of a difference according to the articulation of the HA (RR 1.28, 95% CI 0.98 to 1.67, favours bipolar; 1 study, 50 participants; Analysis 4.5).
In addition, one study reported this outcome using the HHS at five years (Davison 2001). We did not calculate an effect estimate for this study because data were reported without an appropriate measure of variance. See Appendix 5 for mean scores as reported by study authors.

HRQoL

Three studies reported HRQoL. These studies included cemented (Hedbeck 2011; Raia 2003), and uncemented HAs (Figved 2018).

Early:

Hedbeck 2011 reported data using EQ‐5D up to four months since surgery; in this scale, higher scores indicate better quality of life. We found no evidence of a difference in HRQoL according to the articulation of the HA (MD 0.08, 95% CI ‐0.03 to 0.19, favours bipolar; 1 study, 115 participants; very low‐certainty evidence; Analysis 4.6). We downgraded the certainty of the evidence by three levels ‐ two levels for imprecision because data were available from only one small study, and one level for study limitations because this study had unclear risks of bias.

At 12 months:

One study reported data using EQ‐5D at 12 months (Hedbeck 2011). We found no evidence of a difference in quality of life at 12 months (MD 0.03, 95% CI ‐0.08 to 0.14, favours bipolar; 1 study, 99 participants; very low‐certainty evidence; Analysis 4.7).
In addition, Figved 2018 reported this outcome using EQ‐5D, and Raia 2003 reported this outcome at 12 months using SF‐36. We did not pool data from these studies because data were reported without an appropriate measure of variance. See Appendix 5 for average scores as reported by study authors.

Mobility

Five studies reported data for mobility within 12 months, and the HAs in all these studies were fixed with cement (Calder 1995; Calder 1996; Cornell 1998; Raia 2003; Stoffel 2013).

Cornell 1998 used TUG at six months, and we found no evidence of a difference in mobility according to articulation of the HA (MD 5.80, 95% CI ‐6.83 to 18.43, favours unipolar; 1 study, 48 participants; Analysis 4.8).
Stoffel 2013 used a six‐minute walk test at 12 months. We found the mobility was better when a unipolar HA was used (MD ‐45.00 metres, 95% CI ‐80.64 to ‐9.36, favours unipolar; 1 study, 186 participants; Analysis 4.9). The CI in this effect may suggest a clinically important improvement in mobility when a unipolar HA was used (based on a MCID of 59.4 metres in Overgaard 2017).
In addition, we did not calculate an effect estimate for three studies because these data were reported without distribution variables (Calder 1995; Calder 1996; Raia 2003). In Calder 1995 and Calder 1996, study authors reported mobility scores using a subscale of the Nottingham Health Profile. Raia 2003 reported average mobility scores using the Musculoskeletal Functional Assessment Instrument. See Appendix 5 for average scores as reported by study authors.

Mortality

Nine studies reported mortality. Studies included cemented (Calder 1996; Cornell 1998; Davison 2001; Hedbeck 2011; Jeffcote 2010; Kanto 2014; Raia 2003), and uncemented HAs (Figved 2018; Patel 2008).

Early:

The estimate for mortality within four months of surgery was very imprecise, including clinically relevant benefits and harms (RR 0.94, 95% CI 0.54 to 1.64, favours bipolar; 4 studies, 573 participants; I² = 3%; low‐certainty evidence; Analysis 4.10). We downgraded the certainty of the evidence by one level for imprecision because we noted a wide CI in the effect estimate, and one level because some of the included studies had unclear risks of bias.

At 12 months:

Similarly, the estimate for mortality at 12 months from surgery was very imprecise, including clinically relevant benefits and harms (RR 1.17, 95% CI 0.89 to 1.53, favours unipolar; 8 studies, 839 participants; I² = 0%; low‐certainty evidence; Analysis 4.11). This analysis included data reported at 6 months (Cornell 1998), 13 months (Patel 2008), and 24 months (Jeffcote 2010). We downgraded the certainty of the evidence by one level for imprecision because we noted a wide CI in the effect estimate, and one level because some of the included studies had unclear risks of bias.

Late:

Two studies also reported mortality after 24 months from surgery, at 36 months (Davison 2001), and 60 months (Kanto 2014). We found no evidence of a difference in mortality at this late time point according to the articulation of the HA (RR 0.94, 95% CI 0.72 to 1.23, favours bipolar; 2 studies, 362 participants; I² = 0%; Analysis 4.12).

Unplanned return to theatre

Four studies reported unplanned return to theatre. Studies included cemented (Davison 2001; Hedbeck 2011; Kanto 2014), and a mixture of cemented and uncemented stems in both the bipolar and unipolar groups (Abdelkhalek 2011).

We found no evidence of a difference in unplanned return to theatre (RR 1.08, 95% CI 0.44 to 2.64, favours unipolar; 4 studies, 532 participants; I² = 31%; very low‐certainty evidence; Analysis 4.13). Data were reported at the end of study follow‐up which was at 12 months (Hedbeck 2011), 24 months (Abdelkhalek 2011), 48 months (Davison 2001), and 60 months (Kanto 2014). Some Indications for re‐operations were dislocation, acetabular wear, pain, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, revised HA, open reduction and drainage of infection. We downgraded the certainty of the evidence by one level for imprecision because we noted a wide CI in the effect estimate, and two levels for study limitations because the studies were at high risk of detection bias and had high and unclear risks of bias in other domains.

Other important outcomes

We found no evidence of a difference in pain when reported using categorical data, although one small study that used a numerical rating score to measure this outcome found that pain was reduced when a bipolar articulation was used. We found no evidence of a difference according to the articulation of the HA for discharge destination. We also found no evidence of a difference in adverse events related to the implant or fracture, or both (periprosthetic fracture, deep infection, superficial infection, dislocation) (Figure 5), or in adverse events unrelated to the implant or fracture, or both (acute kidney injury, blood transfusion, cerebrovascular accident, chest infection/pneumonia, myocardial infarction, urinary tract infection, deep vein thrombosis, and pulmonary embolism). We report the summary effects of all these important outcomes and adverse effects in Table 11.

Figure 5

$Bipolar hemiarthroplasty versus unipolar hemiarthroplasty. Forest plot for adverse events related to the implant, fracture, or both$

Bipolar hemiarthroplasty versus unipolar hemiarthroplasty. Forest plot for adverse events related to the implant, fracture, or both

Open in table viewer

Table 11. HA (bipolar vs unipolar): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain (categorical data; no pain, or mild pain)	2	Abdelkhalek 2011; Calder 1996	300	RR 1.22, 95% CI 0.82 to 1.82; I² = 61% (favours bipolar); Analysis 4.14
Pain^a Using Numerical Rating Scale (lower scores indicate less pain)	1	Stoffel 2013	233	MD ‐0.60, 95% CI ‐1.07 to ‐0.13 (favours bipolar); Analysis 4.15
Length of hospital stay^a	1	Stoffel 2013	261	MD 0.20 days, 95% CI ‐0.95 to 1.35 (favours unipolar); Analysis 4.16
Discharge destination	2	Calder 1996; Kanto 2014	381	RR 0.95, 95% CI 0.84 to 1.08 (favours bipolar); Analysis 4.17
Adverse events related to surgery
Periprosthetic fracture	1	Hedbeck 2011	120	RR 7.00, 95% CI 0.37 to 132.66 (favours unipolar); Analysis 4.18
Deep infection	7	Calder 1996; Davison 2001; Hedbeck 2011; Jeffcote 2010; Kanto 2014; Malhotra 1995; Stoffel 2013	1122	RR 1.10, 95% CI 0.44 to 2.71 (favours unipolar); Analysis 4.18
Superficial infection	1	Stoffel 2013	261	RR 2.41, 95% CI 0.48 to 12.18 (favours unipolar); Analysis 4.18
Dislocation	9	Abdelkhalek 2011; Calder 1996; Cornell 1998; Davison 2001; Hedbeck 2011; Kanto 2014; Malhotra 1995; Raia 2003; Stoffel 2013	1274	RR 0.62, 95% CI 0.28 to 1.38 (favours bipolar); Analysis 4.18
Adverse events unrelated to surgery
Acute kidney injury	1	Stoffel 2013	261	RR 2.89, 95% CI 0.12 to 70.25 (favours unipolar); Analysis 4.19
Blood transfusion	1	Raia 2003	115	RR 0.91, 95% CI 0.51 to 1.62 (favours bipolar); Analysis 4.19
Cerebrovascular accident	2	Kanto 2014; Stoffel 2013	436	RR 1.57, 95% CI 0.20 to 12.69 (favours unipolar); Analysis 4.19
Chest infection/pneumonia	3	Hedbeck 2011; Kanto 2014; Stoffel 2013	556	RR 0.61, 95% CI 0.10 to 3.86 (favours bipolar); Analysis 4.19
Myocardial infarction	3	Hedbeck 2011; Kanto 2014; Stoffel 2013	556	RR 0.69, 95% CI 0.11 to 4.32 (favours bipolar); Analysis 4.19
Urinary tract infection	1	Stoffel 2013	261	RR 0.96, 95% CI 0.29 to 3.25 (favours bipolar); Analysis 4.19
Venous thromboembolic phenomena (DVT)	2	Hedbeck 2011; Stoffel 2013	381	RR 3.84, 95% CI 0.43 to 34.45 (favours unipolar); Analysis 4.19
Venous thromboembolic phenomena (pulmonary embolism)	1	Hedbeck 2011	120	RR 3.00, 95% CI 0.12 to 72.20 (favours bipolar); Analysis 4.19
^aAdditional data are reported in Appendix 4. We did not calculate effect estimates for the data in Appendix 4 because study authors did not report distribution variables that we required for analysis.

CI: confidence interval
DVT: deep vein thrombosis
MD: mean difference
RR: risk ratio

Sensitivity analysis

High or unclear risk of selection bias (for sequence generation)

We excluded studies with high or unclear risk of selection bias from the primary analyses. This did not alter our interpretation of the effect for any outcomes.

High risk of attrition bias

No studies in this comparison group were at high risk of attrition bias.

3. HAs versus other HAs

Here we present three comparisons that evaluate one design of HA with another design: short stem versus long stem; Thompson versus Exeter Trauma Stem; Moore versus Furlong. A summary of the implant and study characteristics is presented in Table 5.

For each of these comparisons, we report here the effects for critical outcomes and we summarise the effects of other important outcomes in a table. All outcomes in these three comparisons are reported without GRADE assessment. For outcomes measured with scales, we present range of scores and direction of effect for each scale in Appendix 3.

HA: short stem versus standard stem

This comparison includes data from only one study with 151 participants (Lim 2020).

Critical outcomes

ADL, delirium, functional status, HRQoL, and unplanned return to theatre

Lim 2020 did not report data for these outcomes.

Mobility

Lim 2020 measured this outcome using dichotomised Koval's categories (see Appendix 3). We found no evidence of a difference in mobility at two years according to whether a short or standard stem was used (RR 0.98, 95% CI 0.72 to 1.34, favours standard stem; 1 study, 75 participants; Analysis 5.1).

Mortality

Lim 2020 reported data for mortality at 24 months. The estimate was very imprecise; we found no evidence of a difference in mortality two years after surgery (RR 0.77, 95% CI 0.43 to 1.37, favours short stem; 1 study, 151 participants; Analysis 5.2).

Other important outcomes

We found no evidence of a difference in pain according to whether a short stem or standard stem was used. We found no evidence of a difference in adverse events related to the implant or fracture, or both, according to whether a short stem or a standard stem was used (postoperative periprosthetic fracture, loosening, superficial infection, and dislocation). We report the summary effects of all these important outcomes and adverse effects in Table 12.

Open in table viewer

Table 12. HA (short stem vs standard stem): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain (experiencing thigh pain; at 2 years)	1	Lim 2020	71	RR 0.87, 95% CI 0.13 to 5.83 (favours short stem) Analysis 5.3
Adverse events related to the implant or fracture, or both
Postoperative periprosthetic fracture	1	Lim 2020	151	RR 0.96, 95% CI 0.14 to 6.65 (favours short stem); Analysis 5.3
Loosening	1	Lim 2020	151	Not estimable. No events in either group
Superficial infection	1	Lim 2020	151	Not estimable. No events in either group
Dislocation	1	Lim 2020	112	RR 0.93, 95% CI 0.06 to 14.52 (favours short stem); Analysis 5.3

CI: confidence interval
RR: risk ratio

HA: Exeter Trauma Stem (ETS) versus Thompson

This comparison includes two studies with 1164 participants (Parker 2012; Sims 2018).

Critical outcomes

ADL and functional status

Neither study reported data for these outcomes.

Delirium

Parker 2012 reported delirium; the estimate was very imprecise such that no meaningful inference was possible (RR 5.00, 95% CI 0.24 to 102.85, favours Thompson; 1 study, 200 participants; Analysis 6.1).

HRQoL

Sims 2018 reported HRQoL at four months, measured using EQ‐5D, in which higher scores indicate better HRQoL. We found that HRQoL was slightly improved when an ETS was used (MD 0.06, 95% CI 0.00 to 0.11, favours ETS; 1 study, 618 participants; Analysis 6.2). We noted that the CI is compatible with no difference or a small clinically important benefit with an ETS, based on a MCID of 0.07 (Walters 2005).

Mobility

Sims 2018 reported mobility, using categorical data according to whether participants could walk outdoors with or without a walking stick. We combined data for those that were freely mobile or able to walk outdoors with one walking stick, and found no evidence of a difference according to whether an ETS or a Thompson HA was used (RR 1.14, 95% CI 0.83 to 1.57, favours ETS; 1 study, 494 participants; Analysis 6.3). We report data for other categories in Appendix 6.

In addition, Parker 2012 reported mean change in mobility at 3 months and 12 months after surgery. We did not calculate effect estimates for this study because the data were reported without an appropriate measure of variance. See Appendix 7 for mean scores as reported by study authors.

Mortality

Both studies reported mortality.

Early:

The estimate of this effect was imprecise, including clinically relevant benefits and harms. We found no evidence of a difference in mortality at up to four months from surgery (RR 1.20, 95% CI 0.76 to 1.88, favours Thompson; 2 studies, 1164 participants; I² = 45%; Analysis 6.4).

At 12 months:

We also found no evidence of a difference in mortality at 12 months after surgery (RR 1.44, 95% CI 0.94 to 2.21, favours Thompson; 1 study, 200 participants; Analysis 6.5).

Unplanned return to theatre

Both studies reported unplanned return to theatre. We found no evidence of a difference in unplanned return to theatre (RR 0.46, 95% CI 0.05 to 3.89, favours ETS; 2 studies, 1164 participants; I² = 45%; Analysis 6.6). Re‐operations were due to dislocation and acetabular wear, and resolved with revision of the HA.

Other important outcomes

We found no evidence of a difference in adverse events related to the implant or fracture, or both, according to whether an ETS or a Thompson HA was used (intraoperative periprosthetic fracture, deep or superficial infection, dislocation). We also found no evidence of a difference in adverse events unrelated to the implant or fracture, or both, according to whether a Thompson HA or ETS was used (acute kidney injury, blood transfusion, cerebrovascular accident, chest infection/pneumonia, myocardial infarction, DVT, or pulmonary embolism). We report the summary effects for these adverse events in Table 13. Additional outcome data for pain and length of stay is included in Appendix 7, since these data were reported without appropriate measures of variance such that we could not calculate effect estimates.

Open in table viewer

Table 13. HA: ETS vs Thompson: effects of adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Adverse events related to the implant or fracture, or both
Intraoperative periprosthetic fracture	1	Parker 2012	200	RR 1.00, 95% CI 0.21 to 4.84 (favours neither); Analysis 6.7
Deep infection	1	Parker 2012	200	Not estimable; zero events in both groups
Superficial infection	1	Parker 2012	200	RR 3.00, 95% CI 0.32 to 28.35 (favours Thompson); Analysis 6.7
Dislocation	1	Parker 2012	200	RR 0.20, 95% CI 0.01 to 4.11 (favours ETS); Analysis 6.7
Adverse events unrelated to implant or fracture, or both
Acute kidney injury	1	Parker 2012	200	RR 1.00, 95% CI 0.06 to 15.77 (favours neither); Analysis 6.8
Blood transfusion	1	Parker 2012	200	RR 1.00, 95% CI 0.54 to 1.84 (favours neither); Analysis 6.8
Cerebrovascular accident	1	Parker 2012	200	RR 2.00, 95% CI 0.18 to 21.71 (favours Thompson); Analysis 6.8
Chest infection/pneumonia	1	Parker 2012	200	RR 1.67, 95% CI 0.41 to 6.79 (favours Thompson); Analysis 6.8
Myocardial infarction	1	Parker 2012	200	RR 5.00, 95% CI 0.24 to 102.85 (favours Thompson); Analysis 6.8
Venous thromboembolic phenomena (DVT)	1	Parker 2012	200	RR 1.00, 95% CI 0.21 to 4.84 (favours neither); Analysis 6.8
Venous thromboembolic phenomena (pulmonary embolism)	1	Parker 2012	200	Not estimable; zero events in both groups

CI: confidence interval
DVT: deep vein thrombosis
ETS: Exeter trauma stem
HA: hemiarthroplasty
RR: risk ratio

Sensitivity analysis

We excluded Parker 2012 from the primary analysis of early mortality (at ≤ four months) and unplanned return to theatre because the study was at unclear risk of selection bias (for sequence generation). This did not alter our interpretation of the effect for these outcomes. Neither study in this comparison was at high risk of attrition bias.

HA: hydroxyapatite (HAC)‐coated Furlong versus Moore

This comparison includes one study with 82 participants and compares a first generation with a modern uncemented HA (Livesley 1993).

Critical outcomes

ADL, delirium, HRQoL, and mobility

Livesley 1993 did not report data for these outcomes.

Functional status

Livesley 1993 used a five‐point hip function assessment according to Benjamin 1990 to evaluate functional status at 12 months (higher scores indicate better function). We did not calculate effect estimates for this study because data were reported without an appropriate measure of variance. The study authors reported a mean of 33.0 for participants who had a Furlong prosthesis, and a mean of 27.3 for participants who had a Moore prosthesis.

Mortality

Early:

This effect estimate was imprecise. We found no evidence of a difference in mortality according to whether a Furlong or Moore HA was used (RR 0.35, 95% CI 0.07 to 1.82, favours Furlong; 1 study, 82 participants; Analysis 7.1).

At 12 months:

Similarly, we found an imprecise estimate at 12 months. There was no evidence of a difference in mortality according to the type of prosthesis (RR 0.81, 95% CI 0.46 to 1.43, favours Furlong; 1 study, 82 participants; Analysis 7.2).

Unplanned return to theatre

Livesley 1993 reported unplanned return to theatre. The estimate was very imprecise, precluding meaningful interpretation. We found no evidence of a difference in mortality according to the type of prosthesis (RR 1.42, 95% CI 0.13 to 15.00, favours Moore; 1 study, 82 participants; Analysis 7.3). Re‐operations were because of pain, periprosthetic fracture, or infection. The types of re‐operation were not reported.

Other important outcomes

We found no evidence of a difference in pain at rest, or in adverse events related to the implant or fracture according to the type of prosthesis (periprosthetic fracture, superficial infection, or dislocation). We report the summary effects for these adverse events in Table 14.

Open in table viewer

Table 14. HA: Furlong vs Austin‐Moore: effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain at rest (at 12 months)	1	Livesley 1993	82	RR 0.71, 95% CI 0.22 to 2.26 (favours Furlong); Analysis 7.4
Adverse events related to surgery
Periprosthetic fracture	1	Livesley 1993	82	RR 10.71, 95% CI 0.63 to 181.50 (favours Moore); Analysis 7.5
Superficial infection	1	Livesley 1993	82	RR 0.71, 95% CI 0.05 10.93 (favours Furlong); Analysis 7.5
Dislocation	1	Livesley 1993	82	RR 2.14, 95% CI 0.09 to 51.07 (favours Moore); Analysis 7.5

CI: confidence interval
RR: risk ratio

4. THA versus HA

This comparison includes 17 studies with 3232 participants (Baker 2006; Blomfeldt 2007; Cadossi 2013; Chammout 2019; Dorr 1986; HEALTH 2019; Iorio 2019; Keating 2006; Macaulay 2008; Mouzopoulos 2008; Parker 2019; Ravikumar 2000; Ren 2017; Sharma 2016; Sonaje 2017; Van den Bekerom 2010; Xu 2017). A summary of the implant and study characteristics is presented in Table 6. Whilst most designs of HA used in this comparison were modern, one study included a first generation uncemented HA (Ravikumar 2000), and Sharma 2016 did not specify whether a first generation or modern design was used.

Here we report effects for critical outcomes. Where analyses included at least one study in each category, we subgrouped the analysis according to whether studies used a first generation or modern HA stem design in one of the intervention groups.

We summarise the effects of other important review outcomes in a table, which are not subgrouped by stem design, and these outcomes are reported without GRADE assessments. For outcomes measured with scales, we present range of scores and direction of effect for each scale in Appendix 3.

Critical outcomes

ADL

Four studies reported performance of ADL (Blomfeldt 2007; Chammout 2019; Mouzopoulos 2008; Parker 2019).

Early:

We could only combine data from two of the studies. Blomfeldt 2007 used the Katz Index to identify participants that were independent (Katz 1963), and Chammout 2019 did not describe a measurement tool for this outcome. We found evidence that any difference in the number of people who were independent in ADL within four months of surgery is likely to be small (RR 1.03, 95% CI 0.91 to 1.18, favours THA; 2 studies, 225 participants; I² = 0%; very low‐certainty evidence; Analysis 8.1). We downgraded the certainty of the evidence by one level for imprecision because the evidence included few participants, and two levels for study limitations because one of the studies had unclear risks of selection bias, and we found during sensitivity analyses that this may influence the direction of the estimate.
Parker 2019 used a social mobility scale, in which lower scores indicate more independence. We found no evidence of a difference in mobility at 3 months (MD ‐0.10, 95% CI ‐0.46 to 0.26, favours THA; 1 study, 83 participants; Analysis 8.2).

At 12 months:

We also found no evidence of a difference in the number of people who were independent in ADL at 12 months in Blomfeldt 2007 and Chammout 2019 (RR 0.96, 95% CI 0.86 to 1.07, favours HA; 2 studies, 217 participants; I² = 0%; Analysis 8.3).
We also considered data from two studies that used continuous data, measured at 12 months (Mouzopoulos 2008; Parker 2019). Parker 2019 used a social mobility scale, in which lower scores indicate more independence. Mouzopoulos 2008 used the Barthel Index, in which higher scores indicate more independence; accordingly, we inverted the data from Mouzopoulos 2008 in this analysis. However, we did not pool these data owing to substantial statistical heterogeneity (I²= 80%). Data from individual studies are reported in Analysis 8.4.

Late:

Mouzopoulos 2008 also reported data at four years from surgery, and we found no evidence of a difference in performance of ADL at this later time point (MD 5.70, 95% CI 0.21 to 11.19, favours THA; 1 study, 43 participants; very low‐certainty evidence; Analysis 8.5).

Delirium

Two studies measured delirium at 12 months (Parker 2019; Van den Bekerom 2010). We found no evidence of a difference in delirium according to the type of arthroplasty (RR 1.41, 95% CI 0.60 to 3.33, favours HA; 2 studies, 357 participants; low‐certainty evidence; Analysis 8.6). We downgraded the certainty of the evidence by two levels ‐ one level for imprecision because we noted a wide CI in the effect, and one level for study limitations because of unclear risks of bias.

Functional status

Thirteen studies reported functional status, and all studies used modern stem designs in both intervention groups (Baker 2006; Blomfeldt 2007; Cadossi 2013; Chammout 2019; HEALTH 2019; Keating 2006; Macaulay 2008; Mouzopoulos 2008; Ren 2017; Sharma 2016; Sonaje 2017; Van den Bekerom 2010; Xu 2017).

Early:

Blomfeldt 2007 and Chammout 2019 reported mean data within four months of surgery using the HHS, and Keating 2006 reported mean scores using the Johansen hip score; in both scales, higher scores indicate better function. We found improved function within four months of surgery in people who received a THA (SMD 0.27, 95% CI 0.07 to 0.47, favours THA; 3 studies, 395 participants; very low‐certainty evidence; I² = 0%; Analysis 8.7). After converting this effect estimate to the HHS, there appeared to be no clinically important difference in functional status between THAs and HAs (MD 3.44, 95% CI 0.89 to 5.98); this was based on a MCID for HHS of 16 to 18 (Singh 2016).
In addition, Cadossi 2013 reported function using the HHS. We could not calculate effect estimates for this study because data were reported without an appropriate measure of variance. See Appendix 8 for mean scores as reported by study authors.
We downgraded the evidence by three levels to very low certainty ‐ one level for imprecision because the evidence included few participants, and two levels for study limitations because some studies had high and unclear risks of bias, and we found during sensitivity analysis that the direction of effect estimate was influenced by these studies.

At 12 months:

Six studies reported functional status using the HHS, in which higher scores indicate better function (Blomfeldt 2007; Chammout 2019; Macaulay 2008; Mouzopoulos 2008; Sonaje 2017; Xu 2017); one study reported this outcome using the Johansen hip score in which higher scores indicate better function (Keating 2006); and one study reported this outcome using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (HEALTH 2019). Because the WOMAC score has an opposite direction of effect (i.e. lower scores indicate better function), we inverted the data in HEALTH 2019. We found improved function at 12 months in people who had a THA (SMD 0.23, 95% CI 0.14 to 0.44, favours THA; 8 studies, 1273 participants; I² = 0%; Analysis 8.8; Figure 6). After converting this effect estimate to the HHS, there appeared to be no clinically important difference in functional status between THAs and HAs (MD 2.23, 95% CI 1.35 to 4.26); this was based on a MCID for HHS of 16 to 18 (Singh 2016).
Ren 2017 and Sonaje 2017 reported categorical data using the HHS; ranges of scores were reported as excellent, good, medium, or poor, and we combined data for scores which were excellent and good. The time point of measurement was not reported in Ren 2017, and was at 24 months in Sonaje 2017. We found evidence that any difference in excellent and good scores is likely to be small (RR 1.07, 95% CI 0.98 to 1.17, favours THA; 2 studies, 140 participants; I² = 0%; Analysis 8.9). We report data for other categories in Appendix 6.
In addition, three studies reported data at 12 months using the HHS (Cadossi 2013; Sharma 2016; Van den Bekerom 2010). We did not calculate effect estimates for these studies because data were reported without an appropriate measure of variance. See Appendix 8 for mean scores as reported by study authors.

Figure 6

Total hip arthroplasty versus hemiarthroplasty. Forest plot of functional status at 12 months

Late:

Three studies reported mean scores using the HHS at more than 24 months since surgery (Blomfeldt 2007; Mouzopoulos 2008; Xu 2017), and one study used the Oxford Hip Score (Baker 2006). Time points of measurement were at four years (Blomfeldt 2007; Mouzopoulos 2008), five years (Xu 2017), and nine years (Baker 2006). We found that hip function was improved when a THA was used (SMD 0.65, 95% CI 0.23 to 1.08, favours THA; 4 studies, 224 participants; I² = 56%; Analysis 8.10). We noted that this effect did not suggest a clinically important improvement in hip function (based on a MCID of 16 to 18 points; Singh 2016).
In addition, two studies reported late data at three years using the HHS (Cadossi 2013), and at five years using the HHS (Van den Bekerom 2010). We did not calculate effect estimates for these studies because data were reported without an appropriate measure of variance. See Appendix 8 for mean scores as reported by study authors.

HRQoL

Five studies reported HRQoL (Baker 2006; Chammout 2019; HEALTH 2019; Keating 2006; Macaulay 2008).

Early:

Two studies reported EQ‐5D at four months after surgery (Chammout 2019; Keating 2006); in this scale, higher scores indicate improved quality of life. We found no evidence of a difference in HRQoL at four months after surgery according to the type of arthroplasty (MD 0.03, 95% CI ‐0.06 to 0.12, favours THA; 2 studies, 279 participants; I² = 51%; very low‐certainty evidence; Analysis 8.11). We downgraded the certainty of the evidence by three levels ‐ two levels for imprecision because the evidence was compatible with no difference and a clinically meaningful difference (based on a MCID for EQ‐5D of 0.07; Walters 2005), and one level for study limitations because studies had high and unclear risks of bias.

At 12 months:

Four studies reported this outcome at 12 months using EQ‐5D (Chammout 2019; HEALTH 2019; Keating 2006), or SF‐36 (Macaulay 2008); in both scales, higher scores indicate improved quality of life. We found that HRQoL at 12 months was improved when a THA was used (SMD 0.19, 95% CI 0.07 to 0.31, favours THA; 4 studies, 1158 participants; I² = 0%; moderate‐certainty evidence; Analysis 8.12). After converting this effect estimate to the EQ‐5D scale, it is likely that the evidence is most compatible with no clinically important difference in HRQoL between THAs and HAs (0.05, 95% CI 0.02 to 0.08): this was based on a MCID for EQ‐5D of 0.07 (Walters 2005).

Late:

In addition, one study also reported HRQoL using SF‐36 at nine years (Baker 2006). This effect was imprecise; we found no evidence of a difference in health‐related quality of life (MD 5.90, 95% CI ‐1.99 to 13.79, favours THA; 1 study, 34 participants; Analysis 8.13).

Mobility

Five studies reported mobility (Baker 2006; HEALTH 2019; Macaulay 2008; Parker 2019; Ravikumar 2000).

Early:

Parker 2019 used a nine‐point mobility scale, with lower scores indicating better mobility. We found no evidence of a difference in mobility at three months after surgery (MD ‐0.40, 95% CI ‐0.96 to 0.16, favours THA; 1 study, 83 participants; low‐certainty evidence; Analysis 8.14). We downgraded the evidence by one level for imprecision because the evidence included few participants, and one level for study limitations because the study included unclear risks of bias.
Dorr 1986 reported mobility using a six‐point scale to describe ambulation. We did not calculate effect estimates for this study because data were reported without an appropriate measure of variance. See Appendix 8 for mean scores as reported by study authors.

At 12 months:

We combined data from two studies which used the Timed Up and Go (TUG) test to measure mobility (Macaulay 2008; HEALTH 2019); lower values (a shorter length of time) indicate better mobility. We found no evidence of a difference in mobility (MD ‐2.74, 95% CI ‐6.82 to 1.35, favours THA; 2 studies, 575 participants; I² = 9%; Analysis 8.15).
Parker 2019 used a nine‐point mobility scale, with lower scores indicating better mobility. We found no evidence of a difference in mobility according to the type of arthroplasty (MD 0.40, 95% CI ‐0.32 to 1.12, favours HA; 1 study, 78 participants; Analysis 8.16).
Macaulay 2008 and Ravikumar 2000 reported the number of people who were able to ambulate independently at 12 months. We also found no evidence of a difference according to the type of arthroplasty with this mobility measure (RR 0.96, 95% CI 0.71 to 1.31, favours HA; 2 studies, 175 participants; I² = 49%; Analysis 8.17).
In addition, one study reported mobility using a six‐point scale to describe ambulation (Dorr 1986). We did not calculate effect estimates for this study because data were reported without an appropriate measure of variance. See Appendix 8 for mean scores as reported by study authors.

Late:

Ravikumar 2000 also reported the number of people able to ambulate independently at the end of the study follow‐up, which was at 13 years. We found no evidence of a difference in mobility according to the type of arthroplasty (RR 1.27, 95% CI 0.71 to 2.29, favours THA; 1 study, 32 participants; Analysis 8.18).

Mortality

Fourteen studies reported mortality (Baker 2006; Blomfeldt 2007; Cadossi 2013; Chammout 2019; HEALTH 2019; Iorio 2019; Keating 2006; Macaulay 2008; Mouzopoulos 2008; Parker 2019; Ravikumar 2000; Sharma 2016; Van den Bekerom 2010; Xu 2017).

Early:

The effect estimate for mortality within four months of surgery was very imprecise, including clinically relevant benefits and harms. We found no evidence of a difference (RR 0.77, 95% CI 0.42 to 1.42, favours THA; 6 studies, 725 participants; I² = 0%; very low‐certainty evidence; Analysis 8.19). We downgraded the certainty of the evidence by two levels for imprecision because the wide CI included relevant benefits and harms, and one level for study limitations because included studies had high and unclear risks of bias.

At 12 months:

We found no evidence of a difference in mortality at 12 months from surgery according to the type of arthroplasty (RR 1.00, 95% CI 0.83 to 1.22, favours THA; 11 studies, 2667 participants; I² = 0%; moderate‐certainty evidence; Analysis 8.20). We downgraded the certainty of the evidence by one level because included studies had high and unclear risks of bias. We generated a funnel plot which showed no evidence of publication bias from visual inspection; we also found no statistical evidence of small study size effects (Harbord modified test, P value = 0.966).

Late:

Seven studies reported a late follow‐up after surgery. These data were reported at 36 months (Cadossi 2013), 39 months (Baker 2006), 48 months (Blomfeldt 2007; Mouzopoulos 2008), 60 months (Van den Bekerom 2010; Xu 2017), and 13 years (Ravikumar 2000). We found no evidence of a difference at this late follow‐up according to the type of arthroplasty (RR 1.00, 95% CI 0.81 to 1.23, favours HA; 7 studies, 891 participants; I² = 46%; Analysis 8.21).

Unplanned return to theatre

Ten studies reported unplanned return to theatre, which was reported at 12 months (Iorio 2019, Parker 2019), 24 months (Chammout 2019; HEALTH 2019, Keating 2006), 39 months (Baker 2006), 48 months (Dorr 1986; Mouzopoulos 2008), 60 months (Van den Bekerom 2010), and 13 years (Ravikumar 2000). We found no evidence of a difference in unplanned return to theatre according to the type of arthroplasty (RR 0.63, 95% CI 0.37 to 1.07, favours THA; 10 studies, 2594 participants; I² = 40%; low‐certainty evidence; Analysis 8.22). Some re‐operations were because of dislocation, acetabular wear, pain, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, revised HA, open reduction, and drainage of infection. We downgraded the certainty of the evidence by one level for imprecision because the wide CI was consistent with both benefit and harms, and one level for study limitations because the evidence included studies with high and unclear risks of bias which included high risks of detection bias.

Other important outcomes

We found substantial levels of statistical heterogeneity in data for some pain outcomes, and did pool data in these instances. We found no evidence of a difference in discharge destination according to the type of arthroplasty. We also found no evidence of a difference in adverse events related to the implant or fracture, or both (periprosthetic fracture, loosening, deep infection, superficial infection, dislocation) or in adverse events unrelated to the implant or fracture, or both (acute kidney injury, cerebrovascular accident, chest infection/pneumonia, myocardial infarction, urinary tract infection, deep vein thrombosis, and pulmonary embolism). We found that fewer participants had a blood transfusion when a HA was used; however, this analysis was from only two small studies. We report the summary effects of all these important outcomes and adverse effects in Table 15.

Open in table viewer

Table 15. THA vs HA: effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain^a (reported at ≤ 4 months) Using Hip Rating Questionnaire or HHS (higher scores indicate less pain), and VAS and 8‐point pain scale (lower scores indicate less pain; data inverted in meta‐analysis)	5	Blomfeldt 2007; Cadossi 2013; Chammout 2019; Keating 2006; Parker 2019	572	SMD 0.10, 95% CI ‐0.10 to 0.30 (favours THA); Analysis 8.27
Pain^a (at 12 months) Using VAS, 8‐point pain scale or WOMAC (lower scores indicate less pain); and Hip Rating Questionnaire, WOMAC^b or HHS (higher scores indicate less pain; data inverted in meta‐analysis) Follow‐up: 12 months and 24 months	7	Blomfeldt 2007; Cadossi 2013; Chammout 2019; HEALTH 2019; Keating 2006; Macaulay 2008; Parker 2019; Sonaje 2017	1359	SMD ‐0.19, 95% CI ‐0.44 to 0.06 (favours THA); I² = 73%; Analysis 8.24
Pain (> 24 months) Using HHS (higher scores indicate less pain) Follow‐up: 48 months	2	Blomfeldt 2007; Cadossi 2013	83	We did not combine data because of substantial statistical heterogeneity (I² = 96%); Analysis 8.25
Pain (> 24 months) Using categorical data; we report data for those experiencing no pain^c Follow‐up: 13 years	1	Ravikumar 2000	135	RR 1.47, 95% CI 1.07 to 2.00 (favours THA); Analysis 8.26
Length of hospital stay^a	4	Keating 2006; Macaulay 2008; Mouzopoulos 2008; Xu 2017	382	MD 0.72 days, 95% CI ‐0.21 to 1.64 (favours HA); Analysis 8.23
Discharge destination (own home)	2	HEALTH 2019; Keating 2006	1612	RR 0.97, 95% CI 0.87 to 1.08 (favours HA); Analysis 8.28
Discharge destination (geriatric ward)	1	Chammout 2019	120	RR 0.88, 95% CI 0.34 to 2.26 (favours HA); Analysis 8.29
Adverse events related to the implant or fracture, or both
Postoperative periprosthetic fracture	3	HEALTH 2019; Sonaje 2017; Xu 2017	1557	RR 1.08, 95% CI 0.70 to 1.66 (favours HA); Analysis 8.30
Prosthetic loosening	4	Blomfeldt 2007; HEALTH 2019; Van den Bekerom 2010; Xu 2017	1889	RR 0.64, 95% CI 0.17 to 2.41 (favours THA); Analysis 8.30
Deep infection	8	Chammout 2019; Dorr 1986; HEALTH 2019; Parker 2019; Ravikumar 2000; Sharma 2016; Xu 2017; Van den Bekerom 2010	2343	RR 0.87, 95% CI 0.50 to 1.54 (favours THA); Analysis 8.30
Superficial infection	10	Baker 2006; Blomfeldt 2007; Chammout 2019; Dorr 1986; HEALTH 2019; Keating 2006; Macaulay 2008; Parker 2019; Sharma 2016; Van den Bekerom 2010	2495	RR 1.25, 95% CI 0.67 to 2.30 (favours HA); Analysis 8.30
Dislocation	12	Baker 2006; Blomfeldt 2007; Chammout 2019;Dorr 1986; HEALTH 2019; Iorio 2019; Keating 2006; Macaulay 2008; Ravikumar 2000; Sharma 2016; Van den Bekerom 2010; Xu 2017	2719	RR 1.96, 95% CI 1.17 to 3.27 (favours HA); Analysis 8.30
Adverse events unrelated to the implant or fracture, or both
Acute kidney injury	2	Chammout 2019; HEALTH 2019	1561	RR 1.09, 95% CI 0.62 to 1.92 (favours HA); Analysis 8.31
Blood transfusion	2	Keating 2006; Parker 2019	285	RR 2.14, 95% CI 1.27 to 3.61 (favours HA); Analysis 8.31
Cerebrovascular accident	4	Chammout 2019; Keating 2006; Parker 2019; Van den Bekerom 2010	657	RR 1.63, 95% CI 0.63 to 4.21 (favours HA); Analysis 8.31
Chest infection/pneumonia (reported at > 4 months)	5	Baker 2006; Blomfeldt 2007; Chammout 2019; Macaulay 2008; Van den Bekerom 2010	613	RR 0.87, 95% CI 0.38 to 2.00 (favours THA); Analysis 8.31
Myocardial infarction	4	Blomfeldt 2007; Chammout 2019; Keating 2006; Macaulay 2008	460	RR 1.48, 95% CI 0.48, 4.58 (favours HA); Analysis 8.31
Urinary tract infection	1	Macaulay 2008	40	RR 0.19, 95% CI 0.01 to 3.46 (favours THA); Analysis 8.31
Venous thromboembolic phenomena (DVT)	4	Baker 2006; Blomfeldt 2007; Keating 2006; Parker 2019	486	RR 4.25, 95% CI 0.86 to 21.06 (favours HA); Analysis 8.31
Venous thromboembolic phenomena (pulmonary embolism)	5	Baker 2006; Chammout 2019; Keating 2006; Macaulay 2008; Van den Bekerom 2010	673	RR 0.49, 95% CI 0.14 to 1.63 (favours THA); Analysis 8.31
^aAdditional data are reported in Appendix 8. We did not calculate effect estimates for the data in Appendix 8 because study authors did not report distribution variables that we required for analysis. ^bTwo studies reported data from different versions of the WOMAC scale, with opposite directions of effect. We inverted the data from one of these studies so that the direction was consistent across the analysis. ^cData for additional categories are reported in Appendix 5.

CI: confidence interval
DVT: deep vein thrombosis
HA: hemiarthroplasty
MD: mean difference
RR: risk ratio
SMD: standardised mean difference
THA: total hip arthroplasty
VAS: visual analogue scale
WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index

Sensitivity analysis

We performed sensitivity analysis on critical outcomes in which data were available from more than one study for risk of bias judgements (sequence generation and attrition bias). We did not perform sensitivity analysis on mixed populations because most studies reported insufficient information for us to judge whether participants' characteristics in the included studies were mixed. We did not perform sensitivity analysis according to whether interventions are no longer in current use since this was not relevant.

Here, we report the findings of sensitivity analyses only for those outcomes in which we noted an effect which differed in interpretation to the primary analysis.

High or unclear risk of selection bias (for sequence generation)

Early ADL (≤ 4 months; categorical data): we excluded Blomfeldt 2007 from the primary analysis. Only one study remained in analysis. Although the estimate continued to show no evidence of a difference in performance of ADL, we noted the direction favoured the alternative intervention (RR 1.00, 95% CI 0.78 to 1.29, favours HA; 1 study, 111 participants).
Early functional status (≤ 4 months; continuous data): we excluded Blomfeldt 2007 and Keating 2006 from the primary analysis. Only one study remained in analysis; this estimate no longer indicated a benefit in favour of HA (MD 1.00, 95% CI ‐4.03 to 6.03; 1 study, 111 participants).
Late functional status (at > 24 months): we excluded two studies from the primary analysis (Blomfeldt 2007; Mouzopoulos 2008), and found that the effect estimate no longer demonstrated an improvement in hip function when THA was used (MD 4.83, 95% CI 0.48 to 9.18; 1 study, 64 participants).
Early HRQoL (≤ 4 months): we excluded Keating 2006 from the primary analysis. Only one study remained in analysis. Although the estimate continued to show no evidence of a difference in early HRQoL, we noted the direction favoured the alternative intervention (MD ‐0.02, 95% CI ‐0.11 to 0.07, favours HA; 1 study, 111 participants).

High risk of attrition bias

HRQoL (at 12 months): we excluded HEALTH 2019 from the primary analysis, including only studies at low risk of attrition bias for this outcome. We found that the effect estimate no longer showed evidence of a difference between interventions (SMD 0.17, 95% CI ‐0.05 to 0.40; 4 studies, 314 participants; I² = 0%).

5. Single versus multiple articulations of THA

This comparison included two studies with 83 participants in which a standard cup (single articulation) was compared to a dual‐mobility cup (Griffin 2016; Rashed 2020). A summary of the implant and study characteristics is presented in Table 7. For outcomes measured with scales, we present range of scores and direction of effect for each scale in Appendix 3.

Critical outcomes

ADL, delirium, mobility, and unplanned return to theatre

Neither study reported data for these outcomes.

Functional status

This outcome was measured using the Oxford Hip Score in Griffin 2016, and the HHS in Rashed 2020. In both scales, higher scores indicate better function.

Early:

We found no evidence of a difference in functional status within four months of surgery according to the articulation type (SMD ‐0.33, 95% CI ‐0.78 to 0.12, favours dual‐mobility; 2 studies, 78 participants; I² = 0%; Analysis 9.1).

At 12 months:

When measured at 12 months, we found that functional status was improved when a dual‐mobility cup was used (SMD ‐0.60, 95% CI ‐1.05 to ‐0.15, favours dual‐mobility; 2 studies, 79 participants; I² = 0%; Analysis 9.2).

HRQoL

Only one study reported HRQoL (Griffin 2016). This was measured using EQ‐5D, with a range of scores from 0 to 1 (higher scores indicate better quality of life).

Early:

We found no evidence of a difference in HRQoL at four months after surgery according to the articulation (MD 0.24, 95% CI ‐0.21 to 0.69, favours single articulation; 1 study, 16 participants; Analysis 9.3).

At 12 months:

We found improved HRQoL at 12 months after surgery when a standard cup was used (MD 0.30, 95% CI 0.08 to 0.52, favours single articulation; 1 study, 19 participants; Analysis 9.3).

Mortality

Both studies reported mortality (Griffin 2016; Rashed 2020). We found no evidence of a difference in mortality at 12 months after surgery according to whether a dual‐mobility cup or a standard cup was used with the THA (RR 0.62, 95% CI 0.08 to 4.77, favours single articulation; 2 studies, 82 participants; I² = 0%; Analysis 9.4).

Other important outcomes

We found no evidence of a difference in adverse events related to the implant or fracture, or both (deep infection, superficial infection, and dislocation); we noted zero events for dislocation from two small studies. We also found no evidence of a difference in adverse events unrelated to the implant or fracture, or both (DVT). We report the summary statistics for these adverse events in Table 16.

Open in table viewer

Table 16. THA (dual‐mobility cup vs standard cup): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Adverse events related to implant or fracture, or both
Deep infection	1	Rashed 2020	62	RR 1.00, 95% CI 0.07 to 15.28 (favours neither); Analysis 9.5
Superficial infection	1	Rashed 2020	62	RR 3.00, 95% CI 0.33 to 27.29 (favours DM); Analysis 9.5
Dislocation	2	Griffin 2016; Rashed 2020	82	Not estimable; zero events in both groups
Adverse events unrelated to implant or fracture, or both
Venous thromboembolic phenomena	1	Rashed 2020	62	RR 0.33, 95% CI 0.01 to 7.88 (favours single); Analysis 9.6

CI: confidence interval
DM: dual‐mobility
RR: risk ratio

6. Short stem versus standard stem of THA

This comparison includes only one study with 161 participants, comparing a short stem THA with a standard stem (Kim 2012). A summary of the implant and study characteristics is presented in Table 7. For outcomes measured with scales, we present range of scores and direction of effect for each scale in Appendix 3.

Critical outcomes

ADL, delirium, HRQoL, and unplanned return to theatre

Kim 2012 did not report data for these outcomes.

Functional status

Kim 2012 reported functional status measured using the HHS. We found no evidence of a difference in functional status at 24 months after surgery according to whether a short or standard stem was used in the THA, when measured with the HHS (MD ‐0.40, 95% CI ‐3.19 to 2.39, favours standard stem; 1 study, 140 participants; Analysis 10.1).

Mobility

Kim 2012 reported mobility using categorical data according to distance walked (walks > six blocks with or without aids, walks < six blocks, walks indoors only). We found no evidence of a difference in being able to walk more than six blocks with or without aids at 24 months according to whether a short or standard stem was used in the THA (RR 1.10, 95% CI 0.84 to 1.44, favours short stem; 1 study, 424 participants; Analysis 10.2). We report data for other categories in Appendix 6.

Mortality

We found no evidence of a difference in mortality at 12 months from surgery according to whether a short or standard stem was used in the THA (RR 1.20, 95% CI 0.38 to 3.78, favours standard stem; 1 study, 161 participants; Analysis 10.3).

Other important outcomes

We found no evidence of a difference in pain according to whether a short or standard stem was used. We found no evidence of a difference in some adverse events related to the implant or fracture, or both (superficial infection and dislocation), and for adverse events unrelated to the implant or fracture, or both (acute kidney injury, pneumonia, urinary tract infection). We noted fewer intraoperative periprosthetic fractures when a short stem was used, but, as for all adverse events, data were available from only one small study (Kim 2012). We report the summary effects of important outcomes and adverse events in Table 17.

Open in table viewer

Table 17. THA (short stem vs standard stem): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain Number of people experiencing thigh pain at 24 months	1	Kim 2012	140	RR 0.04, 95% CI 0.00 to 0.72 (favours short stem); Analysis 10.4
Adverse events related to implant or fracture, or both
Intraoperative periprosthetic fracture	1	Kim 2012	140	RR 0.13, 95% CI 0.02 to 0.97 (favours short stem); Analysis 10.5
Superficial infection	1	Kim 2012	140	RR 1.00, 95% CI 0.06 to 15.67 (favours neither); Analysis 10.5
Dislocation	1	Kim 2012	140	RR 0.25, 95% CI 0.03 to 2.18 (favours short stem); Analysis 10.5
Adverse events unrelated to implant or fracture, or both
Acute kidney injury	1	Kim 2012	140	RR 0.50, 95% CI 0.05 to 5.39 (favours short stem); Analysis 10.6
Chest infection/pneumonia	1	Kim 2012	140	RR 0.67, 95% CI 0.11 to 3.87 (favours short stem); Analysis 10.6
Urinary tract infection	1	Kim 2012	140	RR 0.47, 95% CI 0.20 to 1.07 (favours short stem); Analysis 10.6

CI: confidence interval
RR: risk ratio

Discussion

Summary of main results

We included 58 studies (50 RCTs, eight quasi‐RCTs) with 10,654 participants with 10,662 hip fractures. All hip fractures were intracapsular, except in one study that included only extracapsular fractures. We also identified seven ongoing studies with an estimated recruitment of 7199 participants.

We found evidence for 10 different comparisons of types of arthroplasties. We report below the main findings of three of these comparisons, representing the most substantial bodies of evidence in the review.

Cemented versus uncemented HA (17 studies, 3644 participants)

Eight studies compared cemented prostheses with first‐generation uncemented prostheses, and nine studies with modern uncemented prostheses. Moderate‐certainty evidence indicated no clinically important difference between interventions in performance of ADL and independent mobility at four months. The estimates for treatment effects in delirium, the risk of mortality within four months of surgery, and unplanned return to theatre were imprecise, of low certainty, and compatible with clinically relevant benefits and harms. There were, however, statistically significant benefits with cemented prostheses in HRQoL at 4 months, and mortality at 12 months, with moderate‐certainty evidence. The magnitude of these effects were compatible with small to large benefits. The evidence for function was of very low certainty, and although the estimate included benefits and harms, these were not clinically important. Subgroup analysis by uncemented prosthesis design suggested that the mortality benefit from cemented prostheses cannot be explained by higher mortality reported in the uncemented group from studies including first‐generation prostheses.

There was no difference in the overall risk of adverse events. However, within this overall risk profile, we found evidence that the risk of intra‐ and postoperative periprosthetic fracture was lower with cemented HA, but the risk of pulmonary embolic events was greater.

We analysed the data for extracapsular fractures separately, and found very low‐certainty evidence of an improvement in functional status within four months of surgery. This difference may be clinically important.

Bipolar HA versus unipolar HA (13 studies, 1499 participants)

Prostheses were fixed with cement in nine studies, and without cement in three studies. No studies reported performance of ADL or functional status within four months of surgery. For the outcomes of delirium, HRQoL within four months of surgery, and unplanned return to theatre, the evidence was of very low certainty, and plausibly included clinically relevant benefits and harms. For mortality at both 4 and 12 months from surgery, the evidence was of low certainty, and plausibly included clinically relevant benefits and harms.

We found no difference in the overall risk of adverse events.

THA versus HA (17 studies, 3232 participants)

We found very low‐certainty evidence in performance of ADL and HRQoL; the findings were compatible both with no effect and a clinically relevant improvement. Similarly, the moderate‐certainty evidence for mortality at 12 months plausibly included clinically relevant benefits and harms. These findings were the same for delirium, mortality at four months, and unplanned return to theatre, but were supported by low‐certainty evidence. For functional status, we noted that an improvement which favoured THA was not clinically important, and that this evidence was of very low certainty.

We found no difference in the overall risk of adverse events.

Other comparisons, which had fewer participants contributing to the available evidence and for which the estimates were generally too imprecise to yield meaningful inferences, were between: cemented and uncemented THAs; a combination of THAs and HAs which were cemented or uncemented; short or standard stem HAs; Exeter Trauma Stem or Thompson HAs; Furlong or Austin‐Moore HAs; single‐ or dual‐mobility articulations of THA; and short or standard stem THAs.

Overall completeness and applicability of evidence

We included 58 studies with 10,654 participants with a hip fracture. Most of our evidence is applicable only to people with intracapsular fractures as only one of the studies included participants with extracapsular fractures. Most extracapsular fractures are treated primarily with fixation rather than arthroplasty. Where reported, we noted a range of mean ages from 63 years to 87 years, and 73% of participants were female. We expected that most studies would include some participants with cognitive impairment, although approximately one‐third of studies excluded participants with cognitive impairment. Studies did not consistently report American Society of Anesthesiologists (ASA) status scores to indicate participants' fitness for surgery. In general, we assess that the review includes participants who are largely representative of the general hip fracture population undergoing arthroplasty surgery.

The studies reported outcomes following interventions that are all still in use worldwide. We recognise that there is variation in practice in different countries. The provision of cemented or modern uncemented HA and THA treatments may be particularly variable across different resource settings. We assess that the findings of this review are therefore applicable only to the countries in which studies were conducted, of which two‐thirds were in European and western countries.

The included studies were conducted between 1977 and 2020. There have been very substantial changes in co‐interventions in hip fracture care over this period of time. This may mean that, in older studies, the absolute effects are not directly applicable to contemporary care, but we found no evidence that the relative effects varied across time. Therefore, we assess that the historical literature is relevant and appropriate for pooling with more contemporary studies.

We identified studies that evaluated most of our clinically relevant, prespecified comparisons. The majority of studies provided evidence for one of three major groups of comparisons: cemented versus uncemented HAs; unipolar versus bipolar HAs; and THA versus HA. Even within these comparisons, with relatively more included studies, we found that many did not report fully outcomes such as performance of ADL or HRQoL. These are key components of the core outcome set for hip fracture, and yet our ability to draw inferences on the effect of interventions on these outcomes was limited. However, mortality was generally well‐reported, an outcome that is valued by individuals and clinicians in assessing intervention effects.

We were unable to fully perform our prespecified subgroup analyses to explore the impact of specific participant characteristics on the outcomes, such as the effect of age or cognitive impairment, since study characteristics were inconsistently reported within and between studies.

We prioritised short‐term outcomes in this review. We attempted to explore the longer‐term outcomes of the interventions, adding a long‐term measure of outcomes after 24 months from surgery. Longer‐term outcome could help to determine cost‐benefit decisions around intervention choices. Although some studies did present longer‐term data, these findings were often less precise due to attrition from death in this older, frail population.

Quality of the evidence

We used GRADE to formally assess the certainty of the evidence for the critical outcomes for the three main comparisons. The certainty of the evidence ranged from moderate to very low certainty. This was often due to imprecision in the estimate and the risk of bias in the included studies.

We judged several studies to have unclear risk of selection bias because they did not provide information about the allocation methods, or high risk of selection bias because they used quasi‐randomised methods to allocate participants to groups. We used sensitivity analysis to explore this, and found that re‐analysing the data without these studies sometimes influenced the effect: either importantly changing the size of the effect by including or excluding clinically relevant effects, or even changing the direction of the effect. All outcomes in the analyses of our main comparisons included studies with unclear or high risks of selection bias, and we therefore downgraded the certainty of the evidence for all outcomes in our main comparison groups owing to study limitations. We also downgraded the evidence for unplanned return to theatre because studies were at a high risk of detection bias for this outcome.

As well as the risks of bias, the majority of the studies had few participants, reported imprecise estimates, and were likely to be at high risk of a type II error (when a researcher may conclude that there is not a signficant effect when actually there is). The potential benefit of meta‐analysis to overcome this limitation was confounded by the reporting of widely different sets of outcomes across the included studies. Approximately two‐thirds of the studies predated the publication of the hip core outcome set which guided the selection of the critical outcomes in this review (Haywood 2014).

We did not downgrade for indirectness as the study populations and types of interventions were consistent with our intended criteria. We did not downgrade for inconsistency. We evaluated the risk of publication bias in only two analyses (in which we had more than 10 studies), and found no reason to downgrade for this potential limitation.

Potential biases in the review process

The review authors conducted a thorough search and independently assessed study eligibility, extracted data, and assessed risk of bias in the included studies before reaching consensus together or with one other review author.

During the review process, we made changes to the methods, which we describe in Differences between protocol and review. The most significant change was to collect data at three time points rather than two time points. This reflected the wider than expected variation in the outcome time points in the included studies. We aggregated outcome data for the 12‐months time point across a window between 4 and 24 months. Due to the high rates of attrition, we recognise that estimates based on later time points may systematically tend towards no effect. However, we believed it was important to report available data, but recognise that the decision may have influenced the pooled effect estimates.

Although data were sometimes more frequently reported after four months from surgery (typically at 12 months), we prioritised early outcomes in the summary of findings tables. The consequence of this decision is that some critical outcomes for the bipolar versus unipolar comparison have no data in the relevant table. However, this is consistent with our protocol, based on a core outcome set for hip fracture, which prioritises early outcomes over late recovery (Haywood 2014). We reached the decision to report mortality at two time points ‐ within four months of surgery and at 12 months after surgery ‐ following discussion with the Cochrane Bone, Joint, and Muscle Trauma Group. Mortality at 12 months still remains a more common time point reported by study authors, reflecting the expectations of organisations that fund research and journal editors.

Each of the interventions included in the review is complex: they are a combination of different design components which are not mutually exclusive. They are described more fully in Table 2. We made prespecified decisions in our protocol in stratifying our comparisons. This had most effect in the comparison of cemented and uncemented prostheses, where we had to divide the pooled analyses into those including studies of HA, THA, and a mixed intervention. This may have reduced the precision of some of the effect estimates by reducing the available studies in any one pooled analysis. However, the effects were largely concordant across the comparison, and we assess this to be unlikely to have substantially changed the inferences from the available data.

We did not explore adverse events related to implants beyond those described in the protocol for this review. We listed all outcomes reported by each study in the Characteristics of included studies, and these lists include additional adverse events for which we did not report data. Data for these additional adverse events for studies previously included are available in a previous version of this review (Parker 2010c). We attempted to collect information about the reasons for unplanned return to theatre, or the types of re‐operation, but found that this information was not clearly reported in many of the studies. This limited our ability to comment further on these events.

Newer studies were typically reported more completely. However, the majority of the available data in this review are derived from the historical literature. Where possible, we have presented the data in chronological order to try to indicate visually if effect estimates have varied systematically with time. We recognise that there may be an interaction, too, with the changes in co‐interventions with time.

We used GRADE only to assess the certainty of the evidence for the critical outcomes in this review that are included in our summary of findings tables. Therefore, we did not report any judgements of certainty for the remaining review outcomes. We highlighted this distinction when introducing the results for each comparison group. Given the risks of bias in all studies, as well as the imprecision in many of the findings, we anticipate that the certainty of most of these remaining review outcomes is likely to range from low to very low.

Agreements and disagreements with other studies or reviews

Although we found no recent comprehensive systematic reviews that evaluate all types of HA and THA within a single review, we found reviews comparing fixation techniques, articulations, and stem designs similar to the comparisons included in this Cochrane Review: cemented versus uncemented HAs (Azegami 2011; Imam 2019a), unipolar with bipolar HAs (Imam 2019b), and HA with THA (Hopley 2010; Lewis 2019; Liu 2020; Metcalfe 2019).

Azegami 2011 included eight RCTs, and reported the potential for reduced pain and improved mobility in cemented HAs compared to uncemented HAs. A more recent review of nine RCTS found no significant differences in pain or other complications, although the review authors observed that cemented HAs may lead to fewer intraoperative fractures (Imam 2019a). The improved mobility and reduction in intraoperative fractures is compatible with our findings. The reporting of pain in our included studies was highly variable, precluding effective pooling of studies, so that in this Cochrane Review, the reduction in pain in the cemented group was not evident. Whilst our data are compatible with this finding, they are also compatible with an alternative hypothesis that modern uncemented prostheses may yield reduced pain.

Imam 2019b included 13 RCTs and 17 observational studies, and found no significant difference in function and mortality between bipolar and unipolar HAs. Although review authors concluded that bipolar HAs lead to lower rates of re‐operation, their analysis included observational studies. An analysis with only RCTs was consistent with our findings that the re‐operation risk is similar with both interventions.

A larger number of reviews comparing HA to THA have been completed in recent years (Hopley 2010; Lewis 2019; Liu 2020; Metcalfe 2019). Results vary across the reviews, with reduced risk of re‐operation and improved function being reported for THA in three reviews (Hopley 2010; Lewis 2019; Liu 2020). Metcalfe 2019 combined a meta‐analysis of five RCTs with data from a comprehensive national cohort of hip fractures of 143,000 individuals, and reported no difference in re‐operation rates or function, which reflects the findings in this review.

This review included two large multicentre studies. The findings of Fernandez 2022 for cemented compared to uncemented HAs, and of HEALTH 2019 for THA compared to HA, provide substantial data which are consistent with the findings in this review.

Figure 1

PRISMA flow diagram

Ir a la figura de la revisiónAbrir en una pestaña nueva

Figure 2

Ir a la figura de la revisiónAbrir en una pestaña nueva

Figure 3

Cemented hemiarthroplasty versus uncemented hemiarthroplasty. Funnel plot for early mortality (≤ 4 months), subgrouped by stem design

Ir a la figura de la revisiónAbrir en una pestaña nueva

Figure 4

Cemented hemiarthroplasty versus uncemented hemiarthroplasty. Funnel plot for mortality at 12 months, subgrouped by stem design

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Bipolar hemiarthroplasty versus unipolar hemiarthroplasty. Forest plot for adverse events related to the implant, fracture, or both$

Figure 5

Bipolar hemiarthroplasty versus unipolar hemiarthroplasty. Forest plot for adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Figure 6

Total hip arthroplasty versus hemiarthroplasty. Forest plot of functional status at 12 months

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 1.1

Comparison 1: THA: cemented vs uncemented, Outcome 1: ADL (measurement tool not defined)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 1.2

Comparison 1: THA: cemented vs uncemented, Outcome 2: Functional status (using HHS, range for scores from 0 to 100; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 1.3

Comparison 1: THA: cemented vs uncemented, Outcome 3: HRQoL (using EQ‐5D, range of scores from o to 1; higher scores indicate better quality of life)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 1.4

Comparison 1: THA: cemented vs uncemented, Outcome 4: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 1.5

Comparison 1: THA: cemented vs uncemented, Outcome 5: Unplanned return to theatre (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 1.6

Comparison 1: THA: cemented vs uncemented, Outcome 6: Pain (using PNRS, range of scores from 0 to 11: lower values indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 1: THA: cemented vs uncemented, Outcome 7: Adverse events related to the implant, fracture, or both$

Analysis 1.7

Comparison 1: THA: cemented vs uncemented, Outcome 7: Adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.1

Comparison 2: HA: cemented vs uncemented, Outcome 1: Early ADL (≤ 4 months, continuous data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.2

Comparison 2: HA: cemented vs uncemented, Outcome 2: Early ADL (≤ 4 months, categorical data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.3

Comparison 2: HA: cemented vs uncemented, Outcome 3: ADL (12 months, continuous data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.4

Comparison 2: HA: cemented vs uncemented, Outcome 4: ADL (12 months, categorical data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.5

Comparison 2: HA: cemented vs uncemented, Outcome 5: Late ADL (> 24 months; categorical data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.6

Comparison 2: HA: cemented vs uncemented, Outcome 6: Delirium (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.7

Comparison 2: HA: cemented vs uncemented, Outcome 7: Early functional status (≤ 4 months, continuous data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.8

Comparison 2: HA: cemented vs uncemented, Outcome 8: Early functional status (≤ 4 months; categorical data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 2: HA: cemented vs uncemented, Outcome 9: Early functional status: extracapsular fractures (≤ 4 months. HHS; higher scores indicate better function)$

Analysis 2.9

Comparison 2: HA: cemented vs uncemented, Outcome 9: Early functional status: extracapsular fractures (≤ 4 months. HHS; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.10

Comparison 2: HA: cemented vs uncemented, Outcome 10: Functional status (12 months; continuous data using HHS, OHS and VELCA; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.11

Comparison 2: HA: cemented vs uncemented, Outcome 11: Functional status (12 months, categorical data using HHS)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 2: HA: cemented vs uncemented, Outcome 12: Functional status: extracapsular fractures (12 months. HHS; higher scores indicate improved function)$

Analysis 2.12

Comparison 2: HA: cemented vs uncemented, Outcome 12: Functional status: extracapsular fractures (12 months. HHS; higher scores indicate improved function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.13

Comparison 2: HA: cemented vs uncemented, Outcome 13: Late functional status (> 24 months using HHS; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.14

Comparison 2: HA: cemented vs uncemented, Outcome 14: Early HRQoL (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.15

Comparison 2: HA: cemented vs uncemented, Outcome 15: HRQoL (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.16

Comparison 2: HA: cemented vs uncemented, Outcome 16: Late HRQoL (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.17

Comparison 2: HA: cemented vs uncemented, Outcome 17: Early mobility (≤ 4 months, independent mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.18

Comparison 2: HA: cemented vs uncemented, Outcome 18: Early mobility (≤ 4 months, continuous data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.19

Comparison 2: HA: cemented vs uncemented, Outcome 19: Early mobility (mean reduction values at ≤ 4 months; higher scores indicate better mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.20

Comparison 2: HA: cemented vs uncemented, Outcome 20: Mobility (12 months, continuous data using different mobility scales; lower scores indicate better mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.21

Comparison 2: HA: cemented vs uncemented, Outcome 21: Mobility (12 months, independent mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.22

Comparison 2: HA: cemented vs uncemented, Outcome 22: Mobility (12 months, dependent on walking aid)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.23

Comparison 2: HA: cemented vs uncemented, Outcome 23: Late mobility (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.24

Comparison 2: HA: cemented vs uncemented, Outcome 24: Late mobility (> 24 months; independent mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.25

Comparison 2: HA: cemented vs uncemented, Outcome 25: Early mortality (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.26

Comparison 2: HA: cemented vs uncemented, Outcome 26: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.27

Comparison 2: HA: cemented vs uncemented, Outcome 27: Late mortality (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.28

Comparison 2: HA: cemented vs uncemented, Outcome 28: Unplanned return to theatre (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.29

Comparison 2: HA: cemented vs uncemented, Outcome 29: Early pain (≤ 4 months, experiencing no pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.30

Comparison 2: HA: cemented vs uncemented, Outcome 30: Early pain (≤ 4 months; mean scores, lower scores indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.31

Comparison 2: HA: cemented vs uncemented, Outcome 31: Pain (12 months, experiencing no pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.32

Comparison 2: HA: cemented vs uncemented, Outcome 32: Pain (12 months, using continuous data; lower values indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.33

Comparison 2: HA: cemented vs uncemented, Outcome 33: Pain (12 months; mean reduction values: lower scores indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.34

Comparison 2: HA: cemented vs uncemented, Outcome 34: Late pain (> 24 months, using mean scores; lower scores indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.35

Comparison 2: HA: cemented vs uncemented, Outcome 35: Late pain (> 24 months; experiencing no pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.36

Comparison 2: HA: cemented vs uncemented, Outcome 36: Length of hospital stay (days)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 2.37

Comparison 2: HA: cemented vs uncemented, Outcome 37: Discharge destination (own home)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 2: HA: cemented vs uncemented, Outcome 38: Adverse events related to the implant, fracture, or both$

Analysis 2.38

Comparison 2: HA: cemented vs uncemented, Outcome 38: Adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 2: HA: cemented vs uncemented, Outcome 39: Adverse events unrelated to the implant, fracture, or both$

Analysis 2.39

Comparison 2: HA: cemented vs uncemented, Outcome 39: Adverse events unrelated to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.1

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 1: Functional status (12 months, using HHS, range of scores from 0 to 100; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.2

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 2: HRQoL (12 months, using SF‐36, range of scores from 0 to 100; higher scores indicate better quality of life)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.3

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 3: Early mortality (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.4

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 4: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.5

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 5: Late mortality (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.6

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 6: Pain (12 months, using HHS pain scales; higher values indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.7

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 7: Pain (> 24 months, using HHS pain scales; higher values indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 3.8

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 8: Unplanned return to theatre (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 9: Adverse events related to the implant, fracture, or both$

Analysis 3.9

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 9: Adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 10: Adverse events unrelated to implant, fracture, or both$

Analysis 3.10

Comparison 3: Mixed HA and THA: cemented vs uncemented, Outcome 10: Adverse events unrelated to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.1

Comparison 4: Bipolar HA vs unipolar HA, Outcome 1: ADL (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.2

Comparison 4: Bipolar HA vs unipolar HA, Outcome 2: Delirium/confusion

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.3

Comparison 4: Bipolar HA vs unipolar HA, Outcome 3: Functional status (12 months; using different measurement tools; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.4

Comparison 4: Bipolar HA vs unipolar HA, Outcome 4: Functional status (12 months. HHS; excellent and good)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.5

Comparison 4: Bipolar HA vs unipolar HA, Outcome 5: Functional status (> 24 months. HHS; excellent or good)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.6

Comparison 4: Bipolar HA vs unipolar HA, Outcome 6: Early HRQoL (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.7

Comparison 4: Bipolar HA vs unipolar HA, Outcome 7: HRQoL (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.8

Comparison 4: Bipolar HA vs unipolar HA, Outcome 8: Mobility (Get up and Go Test; in seconds)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.9

Comparison 4: Bipolar HA vs unipolar HA, Outcome 9: Mobility (6 minute walk test; in metres)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.10

Comparison 4: Bipolar HA vs unipolar HA, Outcome 10: Early mortality (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.11

Comparison 4: Bipolar HA vs unipolar HA, Outcome 11: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.12

Comparison 4: Bipolar HA vs unipolar HA, Outcome 12: Late mortality (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.13

Comparison 4: Bipolar HA vs unipolar HA, Outcome 13: Unplanned return to theatre (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.14

Comparison 4: Bipolar HA vs unipolar HA, Outcome 14: Pain (categorical data; no pain, or mild pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.15

Comparison 4: Bipolar HA vs unipolar HA, Outcome 15: Pain (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.16

Comparison 4: Bipolar HA vs unipolar HA, Outcome 16: Length of hospital stay (days)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 4.17

Comparison 4: Bipolar HA vs unipolar HA, Outcome 17: Discharge destination: return to preoperative residence

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 4: Bipolar HA vs unipolar HA, Outcome 18: Adverse events related to implant, fracture, or both$

Analysis 4.18

Comparison 4: Bipolar HA vs unipolar HA, Outcome 18: Adverse events related to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 4: Bipolar HA vs unipolar HA, Outcome 19: Adverse event unrelated to implant, fracture, or both$

Analysis 4.19

Comparison 4: Bipolar HA vs unipolar HA, Outcome 19: Adverse event unrelated to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 5.1

Comparison 5: HA: short stem vs standard stem, Outcome 1: Mobility (24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 5.2

Comparison 5: HA: short stem vs standard stem, Outcome 2: Mortality (24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 5.3

Comparison 5: HA: short stem vs standard stem, Outcome 3: Pain (24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 5: HA: short stem vs standard stem, Outcome 4: Adverse events related to implant, fracture, or both$

Analysis 5.4

Comparison 5: HA: short stem vs standard stem, Outcome 4: Adverse events related to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 6.1

Comparison 6: HA: ETS vs Thompson, Outcome 1: Delirium

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 6.2

Comparison 6: HA: ETS vs Thompson, Outcome 2: Early HRQoL (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 6.3

Comparison 6: HA: ETS vs Thompson, Outcome 3: Early mobility (freely mobile without aids, or able to walk outdoors with one aid)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 6.4

Comparison 6: HA: ETS vs Thompson, Outcome 4: Early mortality (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 6.5

Comparison 6: HA: ETS vs Thompson, Outcome 5: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 6.6

Comparison 6: HA: ETS vs Thompson, Outcome 6: Unplanned return to theatre (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 6: HA: ETS vs Thompson, Outcome 7: Adverse events related to implant, fracture, or both$

Analysis 6.7

Comparison 6: HA: ETS vs Thompson, Outcome 7: Adverse events related to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 6: HA: ETS vs Thompson, Outcome 8: Adverse events unrelated to implant, fracture, or both$

Analysis 6.8

Comparison 6: HA: ETS vs Thompson, Outcome 8: Adverse events unrelated to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 7.1

Comparison 7: HA: Furlong vs Moore, Outcome 1: Early mortality (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 7.2

Comparison 7: HA: Furlong vs Moore, Outcome 2: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 7.3

Comparison 7: HA: Furlong vs Moore, Outcome 3: Unplanned return to theatre (at end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 7.4

Comparison 7: HA: Furlong vs Moore, Outcome 4: Pain at rest

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 7: HA: Furlong vs Moore, Outcome 5: Adverse events related to the implant, fracture, or both$

Analysis 7.5

Comparison 7: HA: Furlong vs Moore, Outcome 5: Adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.1

Comparison 8: THA vs HA, Outcome 1: Early ADL (≤ 4 months, using categorical data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.2

Comparison 8: THA vs HA, Outcome 2: Early ADL (≤ 4 months; using social mobility scale (lower scores indicate better mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.3

Comparison 8: THA vs HA, Outcome 3: ADL (12 months, using categorical data)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.4

Comparison 8: THA vs HA, Outcome 4: ADL (12 months; using different measurement tools; lower scores indicate more independence))

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.5

Comparison 8: THA vs HA, Outcome 5: Late ADL (> 24 months; using Barthel Index, range of scores from 0 to 100; higher scores indicate more independence)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.6

Comparison 8: THA vs HA, Outcome 6: Delirium

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.7

Comparison 8: THA vs HA, Outcome 7: Early functional status (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.8

Comparison 8: THA vs HA, Outcome 8: Functional status (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.9

Comparison 8: THA vs HA, Outcome 9: Functional status (HHS; excellent or good)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.10

Comparison 8: THA vs HA, Outcome 10: Late functional status (> 24 months; using OHS and HHS; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.11

Comparison 8: THA vs HA, Outcome 11: Early HRQoL (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.12

Comparison 8: THA vs HA, Outcome 12: HRQoL (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.13

Comparison 8: THA vs HA, Outcome 13: HRQoL (> 24 months. Using SF‐36; higher scores indicate better quality of life)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.14

Comparison 8: THA vs HA, Outcome 14: Early mobility (≤ 4 months; lower scores indicate better mobility

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.15

Comparison 8: THA vs HA, Outcome 15: Mobility (12 months, using TUG; lower values indicate better mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.16

Comparison 8: THA vs HA, Outcome 16: Mobility (12 months, using 9‐point mobility scale; lower scores indicate better mobility)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.17

Comparison 8: THA vs HA, Outcome 17: Mobility (12 months; able to ambulate independently)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.18

Comparison 8: THA vs HA, Outcome 18: Late mobility (> 24 months; able to ambulate independently)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.19

Comparison 8: THA vs HA, Outcome 19: Early mortality (≤ 4 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.20

Comparison 8: THA vs HA, Outcome 20: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.21

Comparison 8: THA vs HA, Outcome 21: Late mortality (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.22

Comparison 8: THA vs HA, Outcome 22: Unplanned return to theatre (end of follow‐up)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.23

Comparison 8: THA vs HA, Outcome 23: Length of hospital stay (days)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.24

Comparison 8: THA vs HA, Outcome 24: Pain (12 months: data not combined; lower scores indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.25

Comparison 8: THA vs HA, Outcome 25: Late pain (> 24 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.26

Comparison 8: THA vs HA, Outcome 26: Pain (> 24 months: categorical data: no pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.27

Comparison 8: THA vs HA, Outcome 27: Early pain (≤ 4 months: higher scores indicate less pain)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.28

Comparison 8: THA vs HA, Outcome 28: Discharge destination (own home)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 8.29

Comparison 8: THA vs HA, Outcome 29: Discharge destination (geriatric ward)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 8: THA vs HA, Outcome 30: Adverse events related to implant, fracture, or both$

Analysis 8.30

Comparison 8: THA vs HA, Outcome 30: Adverse events related to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 8: THA vs HA, Outcome 31: Adverse events unrelated to implant, fracture, or both$

Analysis 8.31

Comparison 8: THA vs HA, Outcome 31: Adverse events unrelated to implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 9.1

Comparison 9: THA: single articulation vs dual‐mobility, Outcome 1: Early functional status (≤ 4 months, using different scales; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 9.2

Comparison 9: THA: single articulation vs dual‐mobility, Outcome 2: Functional status (12 months, using OHS and HHS; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 9.3

Comparison 9: THA: single articulation vs dual‐mobility, Outcome 3: HRQoL (using EQ‐5D, range of scores from 0 to 1; higher scores indicate better quality of life)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 9.4

Comparison 9: THA: single articulation vs dual‐mobility, Outcome 4: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 9: THA: single articulation vs dual‐mobility, Outcome 5: Adverse events related to the implant, fracture, or both$

Analysis 9.5

Comparison 9: THA: single articulation vs dual‐mobility, Outcome 5: Adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 9: THA: single articulation vs dual‐mobility, Outcome 6: Adverse events unrelated to the implant, fracture, or both$

Analysis 9.6

Comparison 9: THA: single articulation vs dual‐mobility, Outcome 6: Adverse events unrelated to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 10.1

Comparison 10: THA: short stem vs standard stem, Outcome 1: Functional status (at 24 months; using HHS, range of scores from 0 to 100; higher scores indicate better function)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 10.2

Comparison 10: THA: short stem vs standard stem, Outcome 2: Mobility

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 10.3

Comparison 10: THA: short stem vs standard stem, Outcome 3: Mortality (12 months)

Ir a la figura de la revisiónAbrir en una pestaña nueva

Analysis 10.4

Comparison 10: THA: short stem vs standard stem, Outcome 4: Pain

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 10: THA: short stem vs standard stem, Outcome 5: Adverse events related to the implant, fracture, or both$

Analysis 10.5

Comparison 10: THA: short stem vs standard stem, Outcome 5: Adverse events related to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

$Comparison 10: THA: short stem vs standard stem, Outcome 6: Adverse events unrelated to the implant, fracture, or both$

Analysis 10.6

Comparison 10: THA: short stem vs standard stem, Outcome 6: Adverse events unrelated to the implant, fracture, or both

Ir a la figura de la revisiónAbrir en una pestaña nueva

Summary of findings 1. Cemented versus uncemented hemiarthroplasty for hip fracture in adults

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Cemented versus uncemented hemiarthroplasty for hip fracture in adults
Patient or population: adults with displaced and undisplaced hip fractures; included studies were for intracapsular fractures, except for one study of extracapsular fractures Setting: hospitals; included studies were conducted in China, Croatia, Denmark, Italy, New Zealand, Norway, Pakistan, Slovenia, Sweden, the UK and USA Intervention: HA fixed with cement (included studies which used unipolar or bipolar articulations) Comparison: HA fixed without cement (included studies which used unipolar or bipolar articulations. Designs of HA in 6 studies were first‐generation, and in 2 studies were unknown. We categorised them as first‐generation.)
Outcomes	Risk with uncemented HA	Risk with cemented HA	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Activities in daily living, early (within 4 months): using GARS (range from 18 to 72), a social dependency scale (range of scores 1 to 9); lower values in these scales indicate more independence. Also using OARS‐IADL (range from 0 to 14) and a 5‐point Likert scale derived from EQ‐5D; higher values in these scales indicate more independence Follow‐up: time points in the included studies were at 3 months and 4 months	The mean GARS score in the uncemented group was 45.7. The mean social mobility scale score in the uncemented group was 4.6. The mean OARS‐IADL score in the uncemented group was 3.7. The mean Likert score in the uncemented group was 3.15.	SMD 0.03 lower (0.21 lower to 0.16 higher)	‐	1275 (4 studies)	⊕⊕⊕⊝ moderate^a	This effect did not indicate a clinically important difference, based on a 'rule of thumb' of: 0.2 for a small difference, 0.5 for a medium difference, and 0.8 for a large difference.
Delirium (end of follow‐up) Follow‐up: time points in the included studies were at 12 months and 5 years	Study population		RR 1.06 (0.55 to 2.06)	800 (2 studies)	⊕⊕⊝⊝ low^c
	40 per 1000^b	42 per 1000 (22 to 82)	RR 1.06 (0.55 to 2.06)	800 (2 studies)	⊕⊕⊝⊝ low^c
Functional status, early (within 4 months): using HHS (range from 0 to 100); higher values indicate better function Follow‐up: time points in the included studies were at 6 weeks and 3 months	The mean HHS scores in the uncemented groups ranged from 62.53 to 72.1.	MD 3.38 higher (0.05 higher to 6.70 higher)	‐	416 (3 studies)	⊕⊝⊝⊝ very low^d	This effect did not indicate a clinically important improvement (based on a MCID of 15.9 to 18 points). In addition, data were available in 1 study with extracapsular fractures which showed improvement with cemented HAs (MD 14.70, 95% CI 11.78 to 17.62; 85 participants). We noted that the CI in this effect may indicate a clinically important improvement with cemented HAs in extracapsular fractures (based on a MCID of 15.9 to 18 points).
HRQoL, early (within 4 months): using EQ‐5D (range 0 to 1), and SF‐12 (range 0 to 100); higher values indicate better quality of life. Follow‐up: time points in the included studies were at 3 months and 4 months	The mean EQ‐5D score in the uncemented group ranged from 0.31 to0.58. The mean SF‐12 score in the uncemented group was 33.8.	SMD 0.20 higher (0.02 higher to 0.10 higher)	‐	1122 (3 studies)	⊕⊕⊕⊝ moderate^a	The difference between fixation techniques was compatible with no effect or a clinically important benefit of cemented HAs based on a MCID for EQ‐5D of 0.07.
Mobility, early (within 4 months): able to walk outdoors using no more than 1 walking aid. Follow‐up: time points in the included studies were at 3 months and 4 months	Study population		RR 1.04 (0.95 to 1.14)	980 (3 studies)	⊕⊕⊕⊝ moderate^a
	354 per 1000^b	369 per 1000 (227 to 404)	RR 1.04 (0.95 to 1.14)	980 (3 studies)	⊕⊕⊕⊝ moderate^a
Mortality, early (within 4 months) Follow‐up: time points in the included studies were at hospital discharge, 7 days, 6 weeks, 3 months and 4 months	Study population		RR 95 (0.80 to 1.13)	3136 (12 studies)	⊕⊕⊝⊝ low^e
	143 per 1000^b	136 per 1000 (114 to 162)	RR 95 (0.80 to 1.13)	3136 (12 studies)	⊕⊕⊝⊝ low^e
Mortality at 12 months Follow‐up: time points in the included studies were at 12 months, 16 months, 18 months, and 24 months	Study population		RR 0.86 (0.78 to 0.96)	3727 (15 studies)	⊕⊕⊕⊝ moderate^a
	283 per 1000^b	243 per 1000 (221 to 272)	RR 0.86 (0.78 to 0.96)	3727 (15 studies)	⊕⊕⊕⊝ moderate^a
Unplanned return to theatre (end of follow‐up)^f Follow‐up: time points in the included studies were at 12 months, 2 years and 5 years	Study population		RR 0.70 (0.45 to 1.10)	2336 (6 studies)	⊕⊕⊝⊝ low^g
	39 per 1000^b	27 per 1000 (17 to 43)	RR 0.70 (0.45 to 1.10)	2336 (6 studies)	⊕⊕⊝⊝ low^g
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; EQ‐5D: EuroQoL 5 Dimensions instrument; GARS: Groningen Activity Restriction Scale;HA: hemiarthroplasty; HRQoL: health‐related quality of life; HHS: Harris Hip Score; MCID: minimal clinically important difference; MD: mean difference; OARS‐IADL: Older Americans Resources Scale of Instrumental Activities of Daily Living; RR: risk ratio; SF‐12: Short‐form 12; SMD: standardised mean difference; THA: total hip arthroplasty
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.
^aWe downgraded by one level for study limitations because included studies had some high/unclear risks of bias. ^bDerived from the pooled estimate of the uncemented HA group ^cWe downgraded by two levels: one level for imprecision because we noted a wide CI in the estimate, and one level for study limitations because the studies had unclear risks of bias. ^dWe downgraded by three levels: one level for imprecision because we noted a wide CI in the estimate, and two levels for study limitations because some studies had unclear risks of bias, and we found during sensitivity analyses that the estimate was influenced by these studies. ^eDowngraded by two levels: one level for imprecision because the CI included possible benefits and possible harms, and one level for study limitations because the studies had unclear risks of bias. ^fSome re‐operations were because of infection, acetabular wear, dislocation, periprosthetic fracture or loosening. We noted that types of re‐operation included replacement with THA, Girdlestone and drainage of infection. ^gWe downgraded by two levels for study limitations because some studies had unclear risks of bias and all studies were at high risk of detection bias.

Summary of findings 1. Cemented versus uncemented hemiarthroplasty for hip fracture in adults

Ir a la tabla de la revisión

Summary of findings 2. Bipolar hemiarthroplasty compared with unipolar hemiarthroplasty for hip fracture in adults

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Bipolar hemiarthroplasty compared with unipolar hemiarthroplastyfor hip fracture in adults
Patient or population: adults with displaced and undisplaced hip fractures Setting: hospitals; included studies were conducted in Australia, Egypt, Finland, India, Norway, Sweden, the UK and USA Intervention: bipolar HA. These were fixed with cement in 9 studies, without cement in 3 studies, and at the discretion of the surgeon in 1 study. Comparison: unipolar HA. These were fixed with cement in 9 studies, without cement in 3 studies, and at the discretion of the surgeon in 1 study.
Outcomes	Risk with unipolar HA	Risk with Bipolar HA	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Activities of daily living, early (within 4 months)	‐	‐	‐	‐	‐	No studies reported this outcome
Early delirium Follow‐up: postoperative period	Study population		RR 0.48 (0.09 to 2.58)	261 (1 study)	⊕⊝⊝⊝ very low^b
Early delirium Follow‐up: postoperative period	31 per 1000^a	15 per 1000 (3 to 81)	RR 0.48 (0.09 to 2.58)	261 (1 study)	⊕⊝⊝⊝ very low^b
Functional status, early (within 4 months)	‐	‐	‐	‐	‐	No studies reported this outcome
HRQoL, early (within 4 months): using EQ‐5D (range 0 to 1); higher values indicate better quality of life Follow‐up: 4 months	The mean EQ‐5D score in the unipolar group was 0.54	MD 0.08 higher (0.03 lower to 0.19 higher)	‐	115 (1 study)	⊕⊝⊝⊝ very low^b
Mobility, early (within 4 months)		‐		‐		No studies reported this outcome
Mortality, early (within 4 months) Follow‐up: time points in the included studies were during hospital stay, at 3 months and at 4 months	Study population		RR 0.94 (0.54 to 1.64)	573 (4 studies)	⊕⊕⊝⊝ low^c
	105 per 1000^a	99 per 1000 (57 to 173)	RR 0.94 (0.54 to 1.64)	573 (4 studies)	⊕⊕⊝⊝ low^c
Mortality at 12 months Follow‐up: time points in the included studies were at 6 months, 12 months, 13 months, and 24 months	Study population		RR 1.17 (0.89 to 1.53)	839 (8 studies)	⊕⊕⊝⊝ low^c
	184 per 1000^a	216 per 1000 (164 to 282)	RR 1.17 (0.89 to 1.53)	839 (8 studies)	⊕⊕⊝⊝ low^c
Unplanned return to theatre (end of follow‐up)^d Follow‐up: time points in the included studies were at 12 months, 24 months, 48 months, and 60 months	Study population		RR 1.08 (0.44 to 2.64)	532 (4 studies)	⊕⊝⊝⊝ very low^e
	57 per 1000^a	62 per 1000 (25 to 151)	RR 1.08 (0.44 to 2.64)	532 (4 studies)	⊕⊝⊝⊝ very low^e
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; EQ‐5D: EuroQoL 5 Dimensions instrument; HA: hemiarthroplasty; HRQoL: health‐related quality of life; MD: mean difference; RR: risk ratio; THA: total hip arthroplasty
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.
^aDerived from the unipolar HA group if results from a single study, or otherwise, from the pooled estimate of the unipolar group ^bWe downgraded by three levels: two levels for imprecision because the evidence included very few participants, and one level for study limitations because the included study had high and/or unclear risks of bias. ^cWe downgraded by two levels: one level for imprecision because we noted a wide CI in the effect estimate, and one level for study limitations because some of the included studies had unclear risks of bias. ^dSome re‐operations were because of dislocation, acetabular wear, pain, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, revised HA, open reduction and drainage of infection. ^eWe downgraded by three levels: one level for imprecision, and two levels for study limitations because studies had high and unclear risks of bias, which included high risks of detection bias.

Summary of findings 2. Bipolar hemiarthroplasty compared with unipolar hemiarthroplasty for hip fracture in adults

Ir a la tabla de la revisión

Summary of findings 3. Total hip arthroplasty compared with hemiarthroplasty for hip fracture in adults

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Total hip arthroplasty compared with hemiarthroplasty for hip fracture in adults
Patient or population: adults with displaced and undisplaced hip fractures Setting: hospitals; included studies were conducted in Canada, China, Greece, Finland, India, Italy, the Netherlands, New Zealand, Norway, South Africa, Spain, Sweden, the UK and USA Intervention: THA Comparison: HA (in 1 of the included studies, this was a first‐generation design of HA)
Outcomes	Risk with HA	Risk with THA	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
Activities of daily living, early (within 4 months): using Katz Index and an undefined measurement tool to identify people who were independent Follow‐up: time points in the included studies were at 3 months and 4 months	Study population		RR 1.03 (0.91 to 1.18)	225 (2 studies)	⊕⊝⊝⊝ very low^b
	764 per 1000^a	787 per 1000 (695 to 901)	RR 1.03 (0.91 to 1.18)	225 (2 studies)	⊕⊝⊝⊝ very low^b
Delirium (end of follow‐up) Follow‐up: time point in the included studies was 12 months	Study population		RR 1.41 (0.60 to 3.33)	357 (2 studies)	⊕⊕⊝⊝ low^c
	47 per 1000^a	67 per 1000 (28 to 158)	RR 1.41 (0.60 to 3.33)	357 (2 studies)	⊕⊕⊝⊝ low^c
Functional status, early (within 4 months): using HHS (range from 0 to 100) and Johansen hip score (range from 0 to 100); higher scores indicate better function Follow‐up: time points in the included studies were at 3 months and 4 months	The mean HHS scores in HA groups ranged from 69 to 77.5. The mean Johansen hip score in the HA group was 71.4.	SMD 0.27 higher (0.07 higher to 0.47 higher)	‐	395 (3 studies)	⊕⊝⊝⊝ very low^d	There appeared to be no clinically important difference in this effect, based on a MCID for HHS of 16 to 18
HRQoL, early (within 4 months): using EQ‐5D (range from 0 to 1); higher scores indicate better quality of life Follow‐up: time points in the included studies were at 3 months and 4 months	The mean EQ‐5D scores in the HA groups ranged from 0.61 to 0.67.	MD 0.03 higher (0.06 lower to 0.12 higher)	‐	279 (2 studies)	⊕⊝⊝⊝ very low^e	Compatible with no effect or a clinically important benefit of THA, based on a MCID for EQ‐5D of 0.07
Mobility, early (within 4 months): using a 9‐point mobility scale; lower scores indicate better mobility Follow‐up: time point in the included study was 3 months	The mean mobility score in the HA group was 3.8	MD 0.40 lower (0.96 lower to 0.16 higher)	‐	83 (1 study)	⊕⊕⊝⊝ low^f
Mortality, early (within 4 months) Follow‐up: time points in the included studies were at 1 week, 1 month, 2 months, and 4 months	Study population		RR 0.77 (0.42 to 1.42)	725 (6 studies)	⊕⊝⊝⊝ very low^g
	62 per 1000^a	48 per 1000 (26 to 89)	RR 0.77 (0.42 to 1.42)	725 (6 studies)	⊕⊝⊝⊝ very low^g
Mortality at 12 months Follow‐up: time points in the included studies were at 12 months and 24 months	Study population		RR 1.00 (0.82 to 1.34)	2667 (11 studies)	⊕⊕⊕⊝ moderate^h
	135 per 1000^a	135 per 1000 (112 to 165)	RR 1.00 (0.82 to 1.34)	2667 (11 studies)	⊕⊕⊕⊝ moderate^h
Unplanned return to theatre (end of follow‐up)ⁱ Follow‐up: time points in the included studies were at 12 months, 24 months, 48 months, 60 months, and 13 years	Study population		RR 0.68 (0.41 to 1.15)	2476 (9 studies)	⊕⊕⊝⊝ low^j
	84 per 1000^a	57 per 1000 (35 to 97)	RR 0.68 (0.41 to 1.15)	2476 (9 studies)	⊕⊕⊝⊝ low^j
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; EQ‐5D: EuroQoL 5 Dimensions instrument; HA: hemiarthroplasty; HHS: Harris Hip Score; MCID: minimal clinically important difference; RR: risk ratio; THA: total hip arthroplasty
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.
^aDerived from the pooled estimate of the HA group ^bWe downgraded by three levels: one level for imprecision because the evidence included very few participants, and two levels for study limitations because one of the studies had unclear risk of selection bias and we found during sensitivity analyses that this may influence the estimate. ^cWe downgraded by two levels: one level for imprecision because we noted a wide CI in the effect, and one level for study limitations because of unclear risks of bias. ^dWe downgraded by three levels: one level for imprecision because the evidence included few participants, and two levels for study limitations because some studies had high and unclear risks of bias and we found during sensitivity analysis that the direction of effect was influenced by these studies. ^eWe downgraded by three levels: two levels for imprecision because the evidence was compatible with no difference and a clinically meaningful difference (based on a MCID for EQ‐5D of 0.07), and one level for study limitations because studies had high and unclear risks of bias. ^fWe downgraded by two levels: one level for imprecision because the evidence included few participants, and one level for study limitations because the study included unclear risks of bias. ^gWe downgraded by three levels: two levels for imprecision because the evidence was consistent with both benefits and harms, and one level for study limitations because some included studies had high and unclear risks of bias. ^hWe downgraded by one level for study limitations because included studies were at high or unclear risks of bias. ⁱSome re‐operations were because of dislocation, acetabular wear, pain, periprosthetic fracture or infection. We noted that types of re‐operation included replacement with THA, open reduction, and internal fixation. ^jWe downgraded by two levels: one level for imprecision because the evidence was consistent with both benefits and harms, and one level for study limitations because included studies had high and unclear risks of bias which included high risks of detection bias.

Summary of findings 3. Total hip arthroplasty compared with hemiarthroplasty for hip fracture in adults

Ir a la tabla de la revisión

Table 1. Trochanteric region fractures: type and surgical management (revised AO/OTA classification, January 2018)

Type	Features	Stability	Description
Simple, pertrochanteric fractures (A1)	Isolated pertrochanteric fracture 2‐part fracture Lateral wall intact	Stable	The fracture line can begin anywhere on the greater trochanter and end either above or below the lesser trochanter. The medial cortex is interrupted in only 1 place.
Multifragmentary pertrochanteric fractures (A2)	With 1 or more intermediate fragments Lateral wall may be incompetent	Unstable	The fracture line can start laterally anywhere on the greater trochanter and runs towards the medial cortex which is typically broken in 2 places. This can result in the detachment of a third fragment which may include the lesser trochanter.
Intertrochanteric fractures (A3)	Simple oblique fracture Simple transverse fracture Wedge or multifragmentary fracture	Unstable	The fracture line passes between the 2 trochanters, above the lesser trochanter medially and below the crest of the vastus lateralis laterally.
AO/OTA: Arbeitsgemeinschaft für Osteosynthesefragen (German for "Association for the Study of Internal Fixation") / Orthopaedic Trauma Association

Table 1. Trochanteric region fractures: type and surgical management (revised AO/OTA classification, January 2018)

Ir a la tabla de la revisión

Table 2. Proposed grouping of different types of arthroplasty for hip fracture in adults

Implant category	Variable (articulation/fixation technique)	Implant subcategory	Examples^a	Description
Total hip arthroplasty	Articulation	Femoral head and acetabular bearing surface materials	Metal‐on‐polyethylene (MoP) Ceramic‐on‐polyethylene (CoP) Ceramic‐on‐ceramic (CoC) Metal‐on‐metal (MoM) Polyethylene material Highly cross‐linked (HCL) Not HCL	Bearing surfaces may be grouped into hard (ceramic and metal) and soft (polyethylene variants). Arthroplasties exist with many of the possible combinations of these bearing surfaces.
		Femoral head size	Large head ≥ 36 mm Standard small head < 36 mm	Over the development of hip arthroplasty, different sizes of femoral head have been used, from 22 mm to very large diameters approximating that of the native femoral head. The size of the head represents a compromise between stability and linear and volumetric wear at the articulation. The optimum size varies by indication and bearing materials. 36 mm is considered as a cut‐off between standard and large sizes.
		Acetabular cup mobility	Single Dual	A standard THA has a single articulating surface between the femoral head and acetabulum bearing surface. Alternative designs incorporate a further articulation within the structure of the femoral head.
	Fixation technique	Cemented	Exeter Hip System CPT Hip System	Both components are cemented with polymethylmethacrylate bone cement that is inserted at the time of surgery. It sets hard and acts a grout between the prosthesis and the bone.
		Modern uncemented	Corail Hip System Avenir Hip System Taperloc Hip System	Neither component is cemented but rely on osseous integration forming a direct mechanical linkage between the bone and the implant. The femoral prosthesis may be coated with a substance such as hydroxyapatite which promotes bone growth into the prosthesis. Alternatively, the surface of the prosthesis may be macroscopically and microscopically roughened so that bone grows onto the surface of the implant. The acetabular component may be prepared similarly and may or may not be augmented with screws fixed into the pelvis.
		Hybrid	Combinations	The femoral stem is cemented and the acetabular cup is uncemented.
		Reverse hybrid	Combinations	The acetabular cup is cemented and the femoral stem is uncemented.
Hemiarthroplasty	Articulation	Unipolar	Thompson Austin‐Moore Exeter Trauma Stem Exeter Unitrax Endo Femoral Head CPT Zimmer Unitrax	A single articulation between the femoral head and the native acetabulum. The femoral component can be a single ‘monoblock’ of alloy or be modular, assembled from component parts during surgery.
	Articulation	Bipolar	CPT modular bipolar Exeter modular bipolar Bateman Monk Centrax	The object of the second joint is to reduce acetabular wear. This type of prosthesis has a spherical inner metal head with a size between 22 to 36 mm in diameter. This fits into a polyethylene shell, which in turn is enclosed by a metal cap. There are a number of different types of prostheses with different stem designs.
	Fixation technique	First‐generation uncemented	Thompson Austin Moore	These prostheses were designed before the development of polymethylmethacrylate bone cement and were therefore originally inserted as a ‘press fit’. Long‐term stability through osseus integration was not part of the design concept.
		Cemented	Thompson Exeter Trauma Stem Exeter Hip System CPT Hip System	The femoral stem is cemented with polymethylmethacrylate bone cement that is inserted at the time of surgery. It sets hard and acts a grout between the prosthesis and the bone.
		Modern uncemented	Corail Furlong Avenir	The femoral stem relies on osseous integration forming a direct mechanical linkage between the bone and the implant. A prosthesis may be coated with a substance such as hydroxyapatite, which promotes bone growth into the prosthesis. Alternatively, the surface of the prosthesis may be macroscopically and microscopically roughened so that bone grows onto the surface of the implant.
^aThis list is not exhaustive. Abbreviations: CoC: Ceramic‐on‐ceramic CoP: Ceramic‐on‐polyethylene CPT: collarless polished tapered HCL: Highly cross‐linked MoM: Metal‐on‐metal MoP: Metal‐on‐polyethylene THA: total hip arthroplasty

Table 2. Proposed grouping of different types of arthroplasty for hip fracture in adults

Ir a la tabla de la revisión

Table 3. Implant and study characteristics. Prostheses implanted with cement versus without cement

Study ID	Type of cemented implant Type of uncemented implant	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Brandfoot 2000	1. Cemented, Thompson, unipolar 2. Uncemented Thompson, unipolar	RCT (91)	98	Mortality (16 months, 91)
Cao 2017	1. Cemented, stem type and uni/bipolar NR 2. Uncemented, stem type and uni/bipolar NR	RCT (85)	NR	Function (3 and 6 months, 85)
Chammout 2017	1. Cemented, modular CPT, 32 mm head, cemented cup 2. Uncemented, Bi‐Metric stem, 32 mm head, cemented cup	RCT (69)	100	ADL (3 months, 65; 24 months, 59) Function (24 months, 65) HRQoL (3 months, 64; 12 months, 62) Mortality (12 months, 69) Unplanned return to theatre (24 months, 69)
DeAngelis 2012	1. Cemented, VerSys stem, unipolar 2. Uncemented, beaded stem, unipolar	RCT (130)	100	Unplanned return to theatre (12 months, 130)
Emery 1991	1. Cemented, Thompson, bipolar 2. Uncemented, Moore, bipolar	RCT (53)	100	Mobility (3 months, 39) Mortality (3 and 17/18 months, 53)
Figved 2009	1. Cemented, Spectron, bipolar 2. Uncemented, Corail, bipolar	RCT (230 fractures, 223 participants)	100	ADL (3 months, 190; 12 months, 168) Function (3 months, 189; 12 months, 167) HRQoL (3 months, 143; 12 months, 113) Mobility (3 months, 190; 12 months, 168) Mortality (3 and 12 months, 213) Unplanned return to theatre (12 months, 217)
Harper 1994	1. Cemented, Thompson, unipolar 2. Uncemented, Thompson, unipolar	RCT (137)	100	Mortality (3 and 12 months, 137)
Inngul 2015	1. Cemented, Exeter stem, unipolar or 32mm, cemented cross‐linked polyethylene cup 2. Uncemented, HAC Bimetric stem, unipolar or 32 mm, cemented cross‐linked polyethylene cup	RCT (141)	100	Mortality (4 and 12 months, 141) Unplanned return to theatre (12 months, 141)
Moerman 2017	1. Cemented, Muller, bi/unipolar NR 2. Uncemented, DB10, bi/unipolar NR	RCT (201)	100	ADL (3 months, 114; 12 months, 96) HRQoL (3 months, 102; 12 months, 90) Mobility (3 months, 88; 12 months, 74) Mortality (12 months, 201) Unplanned return to theatre (12 months, 201)
Moroni 2002	1. Cemented, AHS prosthesis, unipolar or THA 2. Uncemented (HAC), Furlong, unipolar or THA	RCT (28)	NR	Function (24 months, 28) HRQoL (24 months, 28) Mortality (24 months, 28)
Movrin 2020	1. Cemented, Muller, bi/unipolar NR 2. Uncemented, DB10, bi/unipolar NR	RCT (158)	100	Function (3 month, 148; 24 months, 94) Mortality (7 days and 24 months, 158)
Parker 2010c	1. Cemented, Thompson, unipolar 2. Uncemented, Moore, unipolar	RCT (400)	100	Delirium (60 months, 400) Mobility (3 months, 327; 60 months, 64) Mortality (12 and 60 months, 400) Unplanned return to theatre (60 months, 400)
Parker 2020	1. Cemented, Exeter Trauma or CPT, unipolar 2. Uncemented, Furlong, unipolar	RCT (400)	100	ADL (4 months, 329; 12 months 283) Delirium (12 months, 400) Mobility (3 months, 329; 12 months, 282) Mortality (3 and 12 months, 400)
Rehman 2014	1. Cemented, Thompson, unipolar 2. Uncemented, Moore, unipolar	RCT (110)	100	Mobility (3 months, 110)
Sadr 1977	1. Cemented, Thompson, unipolar 2. Uncemented, Thompson, unipolar	RCT (40)	100	Function (17 months, 25) Mortality (6 weeks and 12 months, 40)
Santini 2005	1. Cemented, stem type NR, unipolar 2. Uncemented, stem type NR, unipolar	RCT (106)	NR	ADL (12 months, 106) Function (12 months, 106) Mobility (unknown time point, 106) Mortality (at hospital discharge and 12 months, 106)
Sonne‐Holm 1982	1. Cemented, Moore, unipolar 2. Uncemented, Moore, unipolar	RCT (112)	NR	Function (3 and 12 months, 75) Mobility (3 and 12 months, 75) Mortality (6 weeks, 112)
Talsnes 2013	1. Cemented, Landos Titan, bipolar 2. Uncemented, Landos Corail, bipolar	RCT (334)	100	Mortality (12 months, 334)
Taylor 2012	1. Cemented, Exeter, unipolar 2. Uncemented, Zweymuller Alloclassic, unipolar	RCT (160)	100	Mortality (6 weeks and 12 months, 160) Unplanned return to theatre (24 months, 160)
Vidovic 2013	1. Cemented, modular, unipolar 2. Uncemented, Moore, unipolar	RCT (79)	100	Function (3 months, 79; 12 months, 60) Mortality (12 months, 79)
Fernandez 2022	1.Cemented HA, stem and head at surgeon's preference 2.Uncemented HA, stem and head at surgeon's preference	RCT (1225)	99	ADL (4 months, 715; 12 months, 580) HRQoL (4 months, 877; 12 months, 876) Mobility (4 months, 715; 12 months, 583) HRQoL (4 months, 877; 12 months, 876) Unplanned return to theatre (12 months, 1225) Mortality (12 months, 1225)
ADL: activities of daily living AHS: manufacturer's name for implant CPT: collarless, polished, double‐taper design concept DB: manufacturer's name for implant HAC: hydroxyapatite‐coated HRQoL: health‐related quality of life N: total number randomised n: number analysed NR: not reported RCT: randomised controlled trial

Table 3. Implant and study characteristics. Prostheses implanted with cement versus without cement

Ir a la tabla de la revisión

Table 4. Implant and study characteristics. Bipolar HA versus unipolar HA

Study ID	Type of HA bipolar Type of HA unipolar	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Abdelkhalek 2011	1. Mixed cemented/uncemented, bipolar; 2. Mixed cemented/uncemented, unipolar	Quasi RCT (50)	100	Function (4.4 years, 50) Unplanned return to theatre (24 months, 50)
Calder 1995	1. Monk, cemented, bipolar 2. Thompson, cemented, unipolar	RCT (73)	100	Pain (6 months, 73) Mobility (6 months, 73)
Calder 1996	1. Monk, cemented, bipolar 2. Thompson, cemented, unipolar	RCT (250)	100	Mortality (4 and 12 months, 250)
Cornell 1998	1. Cemented modular, bipolar 2. Cemented modular, unipolar	RCT (48)	100	Function (6 months, 48) Mobility (6 months, 48) Mortality (6 months, 48)
Davison 2001	1. Cemented, Monk, bipolar 2. Cemented, Thompson, unipolar	RCT (187)	100	Mortality (12 and 36 months, 187) Unplanned return to theatre (36 months, 187)
Figved 2018	1. Cemented, 28 mm cobalt chromium head and a SelfCentering Bipolar (DePuy) 2. Cemented, Modular Cathcart Unipolar (DePuy)	RCT (28)	100	Function (48 months, 19) HRQoL (12 months, 25; 48 months, 19) Mortality (3 and 12 months, 28)
Hedbeck 2011	1. Cemented, UHR (Stryker), from 42 to 72 mm, bipolar 2. Cemented, Exeter modular, unipolar	RCT (120)	100	ADL (12 months, 99) HRQoL (4 months, 115; 12 months, 99) Mortality (4 and 12 months, 120) Unplanned return to theatre (12 months, 120)
Jeffcote 2010	1. Cemented, Centrax, bipolar 2. Cemented, Unitrax, unipolar	RCT (51)	100	Mortality (24 months, 51)
Kanto 2014	1. Cemented, Vario cup, bipolar 2. Cemented, Lubinus, unipolar	RCT (175)	100	Mortality (during hospital stay and 5 years, 175) Unplanned return to theatre (5 years, 175)
Malhotra 1995	1. Uncemented, Bateman type, bipolar 2. Uncemented; Austin‐Moore; unipolar	RCT (68)	NR	Function (NR, 66)
Patel 2008	1. Uncemented, medical internation stem, bipolar 2: Uncemented, Thompson; unipolar	RCT (40)	100	Mortality (13 months, 40)
Raia 2003	1. Centrax, appropriate‐sized cemented Premise stem, bipolar 2. Unitrax; appropriate‐sized cemented Premise stem, unipolar	RCT (115)	100	Mortality (12 months, 115)
Stoffel 2013	1. Cemented, collarless polished stem, bipolar 2. Cemented, collarless polished stem, unipolar	RCT (294)	100	Delirium (12 months, 261) Function (12 months, 251) Mobility (12 months, 186)
ADL: activities of daily living HA: hemiarthroplasty HRQoL: health‐related quality life N: total number randomised n: number analysed NR: not reported RCT: randomised controlled trial UHR: universal head system (manufacturer's name)

Table 4. Implant and study characteristics. Bipolar HA versus unipolar HA

Ir a la tabla de la revisión

Table 5. Implant and study characteristics. HAs versus other HAs

Study ID	Type of HA in each intervention group	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Lim 2020	1. Short stem, Bencox M stem, proximal Ti‐plasma spray microporous coating, uncemented, bipolar 2. Standard stem, Bencox ID stem, proximal Ti‐plasma spray microporous coating, uncemented, standard stem, bipolar	RCT (151)	100	ADL (24 months, 75) Mortality (24 months, 151)
Livesley 1993	1. HAC bipolar 2. Uncemented; press‐fit Moore‐bipolar	Quasi‐RCT (82)	100	Mortality (1 and 12 months, 82) Unplanned return to theatre (12 months, 82)
Parker 2012	1. Uncemented, Exeter, unipolar 2. Cemented, Thompson, unipolar	RCT (200)	100	Delirium (12 months, 200) Mortality (3 and 12 months, 200) Unplanned return to theatre (12 months, 200)
Sims 2018	1. Uncemented, Exeter, unipolar 2. Cemented, Thompson, unipolar	RCT (964)	100	HRQoL (4 months, 618) Mobility (4 months, 494) Mortality (4 months, 964) Unplanned return to theatre (12 months, 964)
ADL: activities of daily living HA: hemiarthroplasty HAC: hydroxyapatite‐coated HRQoL: health‐related quality of life N: total number randomised n: number analysed RCT: randomised controlled trial

Table 5. Implant and study characteristics. HAs versus other HAs

Ir a la tabla de la revisión

Table 6. Implant and study characteristics. THA versus HA

Study ID	Type of THA Type of HA	Study design (N)	Displaced fractures, %	Critical review outcomes (time point, n)
Baker 2006	1. 28 mm femoral head articulating with an all‐polyethylene Zimmer cemented acetabular cup 2. Endo Femoral Head (Zimmer); cemented; unipolar	RCT (81)	100	Mortality (39 months, 81)
Blomfeldt 2007	1. Modular Exeter femoral component; 28 mm head; OGEE cemented acetabular component 2. Bipolar; modular Exeter, 28 mm head	RCT (120)	100	ADL (4 months, 114; 12 months, 111) Delirium (4 months, 116) Function (48 months, 83) Mortality (4, 12 and 48 months, 120)
Cadossi 2013	1. Uncemented Conus stem and a large‐diameter femoral head 2. Uncemented, bipolar	RCT (96)	100	Mortality (12 and 36 months, 96)
Chammout 2019	1. Cemented 32 mm cobalt‐chromium head; cemented highly cross‐linked polyethylene acetabular component 2. Cemented, unipolar	RCT (120)	100	ADL (3 months, 111; 24 months, 99) Delirium (3 months, 111) Function (24 months, 103) HRQoL (3 months, 111; 12 months, 106) Mortality (24 months, 120) Unplanned return to theatre (24 months, 120)
Dorr 1986	1. 28 mm head size was used 2. Cemented (n = 37) or uncemented (n = 13), bipolar	RCT (89)	100	Unplanned return to theatre (48 months, 89)
HEALTH 2019	1. Surgeon's preference 2. Surgeon's preference	RCT (1495)	100	Function (24 months, 669) HRQoL (24 months, 844) Mobility (24 months, 535) Mortality (24 months, 1441) Unplanned return to theatre (24 months, 1441)
Iorio 2019	1. Dual mobility cup with cementless femoral stem 2. Cementless femoral stem with bipolar head	RCT (60)	100	Mortality (1 and 12 months, 60) Unplanned return to theatre (12 months, 60)
Keating 2006	1. NR 2. Bipolar, cemented	RCT (180)	100	Delirium (24 months, 168) Function (24 months, 168) HRQoL (4 and 12 months, 168) Mortality (24 months, 180) Unplanned return to theatre (12 months, 180)
Macaulay 2008	1. Surgeon's preference 2. Surgeon's preference	RCT (41)	100	Function (24 months, 40) HRQoL (12 months, 40) Mobility (12 and 24 months, 40) Mortality (24 months, 40)
Mouzopoulos 2008	1. Plus (DePuy) 2. Merete	RCT (86)	100	ADL (48 months, 43) Function (48 months, 43) Mortality (12 and 48 months, 86) Unplanned return to theatre (48 months, 49)
Parker 2019	1. CPCS stem (n=29), CPT Zimmer (n=23) 2. Monoblock Exeter Trauma Stem (n=22), CPT bipolar (n=4), CPT modular (n=27)	RCT (105)	100	ADL (12 months, 78) Delirium (12 months, 105) Mobility (12 months, 78) Mortality (4 and 12 months, 105) Unplanned return to theatre (12 months, 105)
Ravikumar 2000	1. Cemented with Howse II 2. Uncemented Austin‐Moore	RCT (180)	100	Mobility (13 years, 32) Mortality (4 and 12 months and 13 years, 180) Unplanned return to theatre (12 months, 180)
Ren 2017	1. Surgeon's preference 2. Cemented	RCT (100)	NR	Function (NR, 100)
Sharma 2016	1. NR 2. NR	RCT (80)	100	Mortality (1 week, 80)
Sonaje 2017	1. NR 2. NR	Quasi‐RCT (42)	100	Function (24 months, 40)
Van den Bekerom 2010	1. Cemented; 32 mm diameter modular head 2. Cemented, bipolar	RCT (281)	100	Mortality (12 and 60 months, 252) Unplanned return to theatre (60 months, 252)
Xu 2017	1. Uncemented prosthesis 2. Bipolar; uncemented	RCT (76)	NR	Function (60 months, 76) Mortality (60 months, 76)
ADL: activity of daily living CPCS: collarless, polished, cemented stem CPT: collarless, polished, double‐taper design concept HA: hemiarthroplasty N: total number randomised n: number analysed NR: not reported OGEE: manufacturer's name for implant RCT: randomised controlled trial THA: total hemiarthroplasty

Table 6. Implant and study characteristics. THA versus HA

Ir a la tabla de la revisión

Table 7. Implant and study characteristics. THAs versus other THAs

Study ID	Type of THA	Study design (N)	Displaced fractures %	Critical review outcomes (time point, n)
Griffin 2016	1. Single articulation: surgeon's preference 2. Dual mobility: surgeon's preference for prosthesis, Novae DM acetabular component; uncemented	RCT (21)	100	Function (12 months, 19) HRQoL (12 months, 19) Mortality (12 months, 21)
Rashed 2020	1. Single articulation: cemented 32 mm head 2. Dual mobility: cemented dual‐mobility cup (Ecofit 2M)	RCT (108)	100	Function (12 months, 60) Mortality (12 months, 62)
Kim 2012	1. Short stem: short, anatomical metaphyseal‐fitting cementless femoral component, 36 mm modular head, cementless acetabular component 2. Standard stem: anatomical medullary locking fully porous coated cementless femoral component, 36 mm Biolox delta ceramic modular head	RCT (161)	100	Function (24 months, 140) Mobility (24 months, 142) Mortality (12 months, 162)
DM: dual‐mobility HRQoL: health‐related quality of life N: total number randomised n: number analysed RCT: randomised controlled trial THA: total hip arthroplasty

Table 7. Implant and study characteristics. THAs versus other THAs

Ir a la tabla de la revisión

Table 8. THA (cemented vs uncemented): effects of other important outcomes and adverse events

Other important outcomes
	Number of studies	Studies	Participants	Effect estimate
Pain at ≤ 4 months Using Pain Numerical Rating Score (range of scores from 0 to 11; lower scores indicate less pain)	1	Chammout 2017	64	MD ‐0.90, 95% CI ‐1.82 to 0.02 (favours cemented); Analysis 1.6
Pain at 12 months Using Pain Numerical Rating Score (range of scores from 0 to 11; lower scores indicate less pain)	1	Chammout 2017	63	MD 1.00, 95% CI 0.03 to 1.97 (favours uncemented); Analysis 1.6
Adverse events related to implant or fracture, or both
Intraoperative periprosthetic fracture	1	Chammout 2017	69	RR 0.14, 95% CI 0.01 to 2.59 (favours cemented); Analysis 1.7
Postoperative periprosthetic fracture	1	Chammout 2017	69	RR 0.97, 95% CI 0.06 to 14.91 (favours cemented); Analysis 1.7
Loosening	1	Chammout 2017	69	RR 0.32, 95% CI 0.01 to 7.69 (favours cemented); Analysis 1.7
Superficial infection	1	Chammout 2017	69	RR 0.32, 95% CI 0.01 to 7.69 (favours cemented); Analysis 1.7
Dislocation	1	Chammout 2017	69	RR 0.32, 95% CI 0.04 to 2.96 (favours cemented); Analysis 1.7
CI: confidence interval MD: mean difference RR: risk ratio

Table 8. THA (cemented vs uncemented): effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 9. HA (cemented vs uncemented): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain^a Experiencing no pain at ≤ 4 months (We inverted data in 2 studies in which data were reported as complaining of pain or experiencing mid‐thigh pain)	4	Harper 1994; Figved 2009; Moerman 2017; Sonne‐Holm 1982	500	RR 1.11, 95% CI 1.00 to 1.22 (favours uncemented); Analysis 2.29
Pain at ≤ 4 months Using VAS, and a 9‐point pain scale (lower values indicate less pain)	3	Movrin 2020; Parker 2010c; Parker 2020	802	Data not combined because of substantial statistical heterogeneity I² = 91%; Analysis 2.30
Pain^a Experiencing no pain at 12 months (We inverted data in 1 study in which data were reported as complaining of pain or experiencing mid‐thigh pain)	4	Emery 1991; Figved 2009; Moerman 2017; Sonne‐Holm 1982	376	RR 1.17, 95% CI 0.85 to 1.63 (favours uncemented); I² = 77%; Analysis 2.31
Pain at 12 months Using VAS, and a 9‐point pain scale (lower values indicate less pain)	4	Figved 2009; Movrin 2020; Parker 2010c; Parker 2020	726	SMD ‐0.06, 95% CI ‐0.33 to 0.21 (favours cemented); I² = 66%; Analysis 2.32
Pain at 12 months Mean reduction values (lower values indicate less pain)	1	Rehman 2014	110	MD ‐0.27, 95% CI ‐0.48 to ‐0.06 (favours cemented); Analysis 2.33
Pain at > 24 months. Reported by study authors at 5 years Using VAS (lower values indicate less pain)	1	Parker 2010c	58	MD ‐0.30, 95% CI ‐0.92 to 0.32 (favours cemented); Analysis 2.34
Pain Experiencing no pain at 5 years	1	Figved 2009	80	RR 1.00, 95% CI 0.77 to 1.30; Analysis 2.35
Length of hospital stay	9	Emery 1991; Figved 2009; Harper 1994; Moerman 2017; Parker 2010c; Parker 2020; Santini 2005; Taylor 2012; Vidovic 2013	1801	MD ‐0.40 days, 95% CI ‐1.03 to 0.23 (favours cemented); Analysis 2.36
Discharge destination Living in own home^a	6	DeAngelis 2012; Figved 2009; Parker 2010c; Santini 2005; Taylor 2012; Fernandez 2022	2331	RR 1.05, 95% CI 0.98 to 1.13 (favours uncemented); Analysis 2.37
Adverse events related to surgery
Intraoperative periprosthetic fracture	7	DeAngelis 2012; Figved 2009; Moerman 2017; Movrin 2020; Parker 2010c; Parker 2020; Taylor 2012	1669	RR 0.20, 95% CI 0.08 to 0.46 (favours cemented); Analysis 2.38
Postoperative periprosthetic fracture	6	Figved 2009; Moerman 2017; Movrin 2020; Santini 2005; Taylor 2012; Fernandez 2022	2819	RR 0.29, 95% CI 0.14 to 0.57 (favours cemented); Analysis 2.38
Loosening	4	Brandfoot 2000; Figved 2009; Moerman 2017; Sadr 1977	537	RR 0.52, 95% CI 0.14 to 1.89 (favours cemented); I² = 45%; Analysis 2.38
Deep infection	7	Figved 2009; Harper 1994; Moerman 2017; Movrin 2020; Parker 2010c; Santini 2005; Taylor 2012	1382	RR 1.56, 95% CI 0.72 to 3.38 (favours uncemented); Analysis 2.38
Superficial infection	7	DeAngelis 2012; Emery 1991; Figved 2009; Harper 1994; Moerman 2017; Parker 2010c; Parker 2020; Sonne‐Holm 1982; Taylor 2012; Fernandez 2022	1210	RR 1.23, 95% CI 0.73 to 2.06 (favours uncemented); Analysis 2.38
Dislocation	8	Figved 2009; Harper 1994; Moerman 2017; Movrin 2020; Parker 2010c; Parker 2020; Sadr 1977; Santini 2005; Taylor 2012; Fernandez 2022	3032	RR 1.08, 95% CI 0.61 to 1.91 (favours uncemented); Analysis 2.38
Adverse events unrelated to surgery
Acute kidney injury	4	Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2226	RR 1.23, 95% CI 0.76 to 2.00 (favours uncemented); Analysis 2.39
Blood transfusion	7	DeAngelis 2012; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Talsnes 2013; Fernandez 2022	2907	RR 1.00, 95% CI 0.83 to 1.20 (favours cemented); I² = 36%; Analysis 2.39
Cerebrovascular accident	5	DeAngelis 2012; Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2356	RR 0.93, 95% CI 0.41 to 2.10 (favours cemented); Analysis 2.39
Chest infection/pneumonia	8	DeAngelis 2012; Emery 1991; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Taylor 2012; Fernandez 2022	2789	RR 0.78, 95% CI 0.50 to 1.21 (favours cemented); Analysis 2.39
Myocardial infarction	7	DeAngelis 2012; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Santini 2005; Fernandez 2022	1457	RR 0.91, 95% CI 0.44 to 1.89 (favours cemented); Analysis 2.39
Urinary tract infection	5	Emery 1991; Moerman 2017; Santini 2005; Taylor 2012; Fernandez 2022	1745	RR 0.89, 95% CI 0.65 to 1.20 (favours cemented); Analysis 2.39
Venous thromboembolic phenomena (DVT)	7	Cao 2017; DeAngelis 2012; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2661	RR 1.28, 95% CI 0.56 to 2.90 (favours uncemented); Analysis 2.39
Venous thromboembolic phenomena (pulmonary embolism)	6	Emery 1991; Figved 2009; Moerman 2017; Parker 2010c; Parker 2020; Fernandez 2022	2499	RR 3.56, 95% CI 1.26 to 10.11 (favours uncemented); Analysis 2.39
^aOther data is reported in Appendix 5 CI: confidence interval DVT: deep vein thrombosis MD: mean difference RR: risk ratio VAS: visual analogue scale

Table 9. HA (cemented vs uncemented): effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 10. THA (mixed HA and THA): cemented vs uncemented: effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Adverse events related to the implant or fracture, or both
Intraoperative periprosthetic fracture	1	Inngul 2015	141	RR 0.06, 95% CI 0.00 to 0.98 (favours cemented); Analysis 3.9
Superficial infection	1	Inngul 2015	141	RR 0.49, 95% CI 0.16 to 1.52 (favours cemented); Analysis 3.9
Dislocation	1	Moroni 2002	28	RR 0.87, 95% CI 0.14 to 5.32 (favours cemented); Analysis 3.9
Adverse events unrelated to implant or fracture, or both
Acute kidney injury	1	Inngul 2015	141	RR 0.37, 95% CI 0.02 to 8.87 (favours cemented); Analysis 3.10
Chest infection/pneumonia	1	Inngul 2015	141	RR 0.55, 95% CI 0.05 to 5.95 (favours cemented); Analysis 3.10
Myocardial infarction	1	Inngul 2015	141	RR 0.37, 95% CI 0.02 to 8.87 (favours cemented); Analysis 3.10
Urinary tract infection	1	Inngul 2015	141	RR 1.42, 95% CI 0.56 to 3.60 (favours uncemented); Analysis 3.10
CI: confidence interval RR: risk ratio

Table 10. THA (mixed HA and THA): cemented vs uncemented: effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 11. HA (bipolar vs unipolar): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain (categorical data; no pain, or mild pain)	2	Abdelkhalek 2011; Calder 1996	300	RR 1.22, 95% CI 0.82 to 1.82; I² = 61% (favours bipolar); Analysis 4.14
Pain^a Using Numerical Rating Scale (lower scores indicate less pain)	1	Stoffel 2013	233	MD ‐0.60, 95% CI ‐1.07 to ‐0.13 (favours bipolar); Analysis 4.15
Length of hospital stay^a	1	Stoffel 2013	261	MD 0.20 days, 95% CI ‐0.95 to 1.35 (favours unipolar); Analysis 4.16
Discharge destination	2	Calder 1996; Kanto 2014	381	RR 0.95, 95% CI 0.84 to 1.08 (favours bipolar); Analysis 4.17
Adverse events related to surgery
Periprosthetic fracture	1	Hedbeck 2011	120	RR 7.00, 95% CI 0.37 to 132.66 (favours unipolar); Analysis 4.18
Deep infection	7	Calder 1996; Davison 2001; Hedbeck 2011; Jeffcote 2010; Kanto 2014; Malhotra 1995; Stoffel 2013	1122	RR 1.10, 95% CI 0.44 to 2.71 (favours unipolar); Analysis 4.18
Superficial infection	1	Stoffel 2013	261	RR 2.41, 95% CI 0.48 to 12.18 (favours unipolar); Analysis 4.18
Dislocation	9	Abdelkhalek 2011; Calder 1996; Cornell 1998; Davison 2001; Hedbeck 2011; Kanto 2014; Malhotra 1995; Raia 2003; Stoffel 2013	1274	RR 0.62, 95% CI 0.28 to 1.38 (favours bipolar); Analysis 4.18
Adverse events unrelated to surgery
Acute kidney injury	1	Stoffel 2013	261	RR 2.89, 95% CI 0.12 to 70.25 (favours unipolar); Analysis 4.19
Blood transfusion	1	Raia 2003	115	RR 0.91, 95% CI 0.51 to 1.62 (favours bipolar); Analysis 4.19
Cerebrovascular accident	2	Kanto 2014; Stoffel 2013	436	RR 1.57, 95% CI 0.20 to 12.69 (favours unipolar); Analysis 4.19
Chest infection/pneumonia	3	Hedbeck 2011; Kanto 2014; Stoffel 2013	556	RR 0.61, 95% CI 0.10 to 3.86 (favours bipolar); Analysis 4.19
Myocardial infarction	3	Hedbeck 2011; Kanto 2014; Stoffel 2013	556	RR 0.69, 95% CI 0.11 to 4.32 (favours bipolar); Analysis 4.19
Urinary tract infection	1	Stoffel 2013	261	RR 0.96, 95% CI 0.29 to 3.25 (favours bipolar); Analysis 4.19
Venous thromboembolic phenomena (DVT)	2	Hedbeck 2011; Stoffel 2013	381	RR 3.84, 95% CI 0.43 to 34.45 (favours unipolar); Analysis 4.19
Venous thromboembolic phenomena (pulmonary embolism)	1	Hedbeck 2011	120	RR 3.00, 95% CI 0.12 to 72.20 (favours bipolar); Analysis 4.19
^aAdditional data are reported in Appendix 4. We did not calculate effect estimates for the data in Appendix 4 because study authors did not report distribution variables that we required for analysis.
CI: confidence interval DVT: deep vein thrombosis MD: mean difference RR: risk ratio

Table 11. HA (bipolar vs unipolar): effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 12. HA (short stem vs standard stem): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain (experiencing thigh pain; at 2 years)	1	Lim 2020	71	RR 0.87, 95% CI 0.13 to 5.83 (favours short stem) Analysis 5.3
Adverse events related to the implant or fracture, or both
Postoperative periprosthetic fracture	1	Lim 2020	151	RR 0.96, 95% CI 0.14 to 6.65 (favours short stem); Analysis 5.3
Loosening	1	Lim 2020	151	Not estimable. No events in either group
Superficial infection	1	Lim 2020	151	Not estimable. No events in either group
Dislocation	1	Lim 2020	112	RR 0.93, 95% CI 0.06 to 14.52 (favours short stem); Analysis 5.3
CI: confidence interval RR: risk ratio

Table 12. HA (short stem vs standard stem): effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 13. HA: ETS vs Thompson: effects of adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Adverse events related to the implant or fracture, or both
Intraoperative periprosthetic fracture	1	Parker 2012	200	RR 1.00, 95% CI 0.21 to 4.84 (favours neither); Analysis 6.7
Deep infection	1	Parker 2012	200	Not estimable; zero events in both groups
Superficial infection	1	Parker 2012	200	RR 3.00, 95% CI 0.32 to 28.35 (favours Thompson); Analysis 6.7
Dislocation	1	Parker 2012	200	RR 0.20, 95% CI 0.01 to 4.11 (favours ETS); Analysis 6.7
Adverse events unrelated to implant or fracture, or both
Acute kidney injury	1	Parker 2012	200	RR 1.00, 95% CI 0.06 to 15.77 (favours neither); Analysis 6.8
Blood transfusion	1	Parker 2012	200	RR 1.00, 95% CI 0.54 to 1.84 (favours neither); Analysis 6.8
Cerebrovascular accident	1	Parker 2012	200	RR 2.00, 95% CI 0.18 to 21.71 (favours Thompson); Analysis 6.8
Chest infection/pneumonia	1	Parker 2012	200	RR 1.67, 95% CI 0.41 to 6.79 (favours Thompson); Analysis 6.8
Myocardial infarction	1	Parker 2012	200	RR 5.00, 95% CI 0.24 to 102.85 (favours Thompson); Analysis 6.8
Venous thromboembolic phenomena (DVT)	1	Parker 2012	200	RR 1.00, 95% CI 0.21 to 4.84 (favours neither); Analysis 6.8
Venous thromboembolic phenomena (pulmonary embolism)	1	Parker 2012	200	Not estimable; zero events in both groups
CI: confidence interval DVT: deep vein thrombosis ETS: Exeter trauma stem HA: hemiarthroplasty RR: risk ratio

Table 13. HA: ETS vs Thompson: effects of adverse events

Ir a la tabla de la revisión

Table 14. HA: Furlong vs Austin‐Moore: effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain at rest (at 12 months)	1	Livesley 1993	82	RR 0.71, 95% CI 0.22 to 2.26 (favours Furlong); Analysis 7.4
Adverse events related to surgery
Periprosthetic fracture	1	Livesley 1993	82	RR 10.71, 95% CI 0.63 to 181.50 (favours Moore); Analysis 7.5
Superficial infection	1	Livesley 1993	82	RR 0.71, 95% CI 0.05 10.93 (favours Furlong); Analysis 7.5
Dislocation	1	Livesley 1993	82	RR 2.14, 95% CI 0.09 to 51.07 (favours Moore); Analysis 7.5
CI: confidence interval RR: risk ratio

Table 14. HA: Furlong vs Austin‐Moore: effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 15. THA vs HA: effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain^a (reported at ≤ 4 months) Using Hip Rating Questionnaire or HHS (higher scores indicate less pain), and VAS and 8‐point pain scale (lower scores indicate less pain; data inverted in meta‐analysis)	5	Blomfeldt 2007; Cadossi 2013; Chammout 2019; Keating 2006; Parker 2019	572	SMD 0.10, 95% CI ‐0.10 to 0.30 (favours THA); Analysis 8.27
Pain^a (at 12 months) Using VAS, 8‐point pain scale or WOMAC (lower scores indicate less pain); and Hip Rating Questionnaire, WOMAC^b or HHS (higher scores indicate less pain; data inverted in meta‐analysis) Follow‐up: 12 months and 24 months	7	Blomfeldt 2007; Cadossi 2013; Chammout 2019; HEALTH 2019; Keating 2006; Macaulay 2008; Parker 2019; Sonaje 2017	1359	SMD ‐0.19, 95% CI ‐0.44 to 0.06 (favours THA); I² = 73%; Analysis 8.24
Pain (> 24 months) Using HHS (higher scores indicate less pain) Follow‐up: 48 months	2	Blomfeldt 2007; Cadossi 2013	83	We did not combine data because of substantial statistical heterogeneity (I² = 96%); Analysis 8.25
Pain (> 24 months) Using categorical data; we report data for those experiencing no pain^c Follow‐up: 13 years	1	Ravikumar 2000	135	RR 1.47, 95% CI 1.07 to 2.00 (favours THA); Analysis 8.26
Length of hospital stay^a	4	Keating 2006; Macaulay 2008; Mouzopoulos 2008; Xu 2017	382	MD 0.72 days, 95% CI ‐0.21 to 1.64 (favours HA); Analysis 8.23
Discharge destination (own home)	2	HEALTH 2019; Keating 2006	1612	RR 0.97, 95% CI 0.87 to 1.08 (favours HA); Analysis 8.28
Discharge destination (geriatric ward)	1	Chammout 2019	120	RR 0.88, 95% CI 0.34 to 2.26 (favours HA); Analysis 8.29
Adverse events related to the implant or fracture, or both
Postoperative periprosthetic fracture	3	HEALTH 2019; Sonaje 2017; Xu 2017	1557	RR 1.08, 95% CI 0.70 to 1.66 (favours HA); Analysis 8.30
Prosthetic loosening	4	Blomfeldt 2007; HEALTH 2019; Van den Bekerom 2010; Xu 2017	1889	RR 0.64, 95% CI 0.17 to 2.41 (favours THA); Analysis 8.30
Deep infection	8	Chammout 2019; Dorr 1986; HEALTH 2019; Parker 2019; Ravikumar 2000; Sharma 2016; Xu 2017; Van den Bekerom 2010	2343	RR 0.87, 95% CI 0.50 to 1.54 (favours THA); Analysis 8.30
Superficial infection	10	Baker 2006; Blomfeldt 2007; Chammout 2019; Dorr 1986; HEALTH 2019; Keating 2006; Macaulay 2008; Parker 2019; Sharma 2016; Van den Bekerom 2010	2495	RR 1.25, 95% CI 0.67 to 2.30 (favours HA); Analysis 8.30
Dislocation	12	Baker 2006; Blomfeldt 2007; Chammout 2019;Dorr 1986; HEALTH 2019; Iorio 2019; Keating 2006; Macaulay 2008; Ravikumar 2000; Sharma 2016; Van den Bekerom 2010; Xu 2017	2719	RR 1.96, 95% CI 1.17 to 3.27 (favours HA); Analysis 8.30
Adverse events unrelated to the implant or fracture, or both
Acute kidney injury	2	Chammout 2019; HEALTH 2019	1561	RR 1.09, 95% CI 0.62 to 1.92 (favours HA); Analysis 8.31
Blood transfusion	2	Keating 2006; Parker 2019	285	RR 2.14, 95% CI 1.27 to 3.61 (favours HA); Analysis 8.31
Cerebrovascular accident	4	Chammout 2019; Keating 2006; Parker 2019; Van den Bekerom 2010	657	RR 1.63, 95% CI 0.63 to 4.21 (favours HA); Analysis 8.31
Chest infection/pneumonia (reported at > 4 months)	5	Baker 2006; Blomfeldt 2007; Chammout 2019; Macaulay 2008; Van den Bekerom 2010	613	RR 0.87, 95% CI 0.38 to 2.00 (favours THA); Analysis 8.31
Myocardial infarction	4	Blomfeldt 2007; Chammout 2019; Keating 2006; Macaulay 2008	460	RR 1.48, 95% CI 0.48, 4.58 (favours HA); Analysis 8.31
Urinary tract infection	1	Macaulay 2008	40	RR 0.19, 95% CI 0.01 to 3.46 (favours THA); Analysis 8.31
Venous thromboembolic phenomena (DVT)	4	Baker 2006; Blomfeldt 2007; Keating 2006; Parker 2019	486	RR 4.25, 95% CI 0.86 to 21.06 (favours HA); Analysis 8.31
Venous thromboembolic phenomena (pulmonary embolism)	5	Baker 2006; Chammout 2019; Keating 2006; Macaulay 2008; Van den Bekerom 2010	673	RR 0.49, 95% CI 0.14 to 1.63 (favours THA); Analysis 8.31
^aAdditional data are reported in Appendix 8. We did not calculate effect estimates for the data in Appendix 8 because study authors did not report distribution variables that we required for analysis. ^bTwo studies reported data from different versions of the WOMAC scale, with opposite directions of effect. We inverted the data from one of these studies so that the direction was consistent across the analysis. ^cData for additional categories are reported in Appendix 5.
CI: confidence interval DVT: deep vein thrombosis HA: hemiarthroplasty MD: mean difference RR: risk ratio SMD: standardised mean difference THA: total hip arthroplasty VAS: visual analogue scale WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index

Table 15. THA vs HA: effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 16. THA (dual‐mobility cup vs standard cup): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Adverse events related to implant or fracture, or both
Deep infection	1	Rashed 2020	62	RR 1.00, 95% CI 0.07 to 15.28 (favours neither); Analysis 9.5
Superficial infection	1	Rashed 2020	62	RR 3.00, 95% CI 0.33 to 27.29 (favours DM); Analysis 9.5
Dislocation	2	Griffin 2016; Rashed 2020	82	Not estimable; zero events in both groups
Adverse events unrelated to implant or fracture, or both
Venous thromboembolic phenomena	1	Rashed 2020	62	RR 0.33, 95% CI 0.01 to 7.88 (favours single); Analysis 9.6
CI: confidence interval DM: dual‐mobility RR: risk ratio

Table 16. THA (dual‐mobility cup vs standard cup): effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Table 17. THA (short stem vs standard stem): effects of other important outcomes and adverse events

Outcome	Number of studies	Studies	Participants	Effect estimate
Other important outcomes
Pain Number of people experiencing thigh pain at 24 months	1	Kim 2012	140	RR 0.04, 95% CI 0.00 to 0.72 (favours short stem); Analysis 10.4
Adverse events related to implant or fracture, or both
Intraoperative periprosthetic fracture	1	Kim 2012	140	RR 0.13, 95% CI 0.02 to 0.97 (favours short stem); Analysis 10.5
Superficial infection	1	Kim 2012	140	RR 1.00, 95% CI 0.06 to 15.67 (favours neither); Analysis 10.5
Dislocation	1	Kim 2012	140	RR 0.25, 95% CI 0.03 to 2.18 (favours short stem); Analysis 10.5
Adverse events unrelated to implant or fracture, or both
Acute kidney injury	1	Kim 2012	140	RR 0.50, 95% CI 0.05 to 5.39 (favours short stem); Analysis 10.6
Chest infection/pneumonia	1	Kim 2012	140	RR 0.67, 95% CI 0.11 to 3.87 (favours short stem); Analysis 10.6
Urinary tract infection	1	Kim 2012	140	RR 0.47, 95% CI 0.20 to 1.07 (favours short stem); Analysis 10.6
CI: confidence interval RR: risk ratio

Table 17. THA (short stem vs standard stem): effects of other important outcomes and adverse events

Ir a la tabla de la revisión

Comparison 1. THA: cemented vs uncemented

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 ADL (measurement tool not defined) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

1.1.1 Early (≤ 4 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.1.2 At 12 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.2 Functional status (using HHS, range for scores from 0 to 100; higher scores indicate better function) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

1.2.1 Early (reported at ≤ 4 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.2.2 At 12 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.3 HRQoL (using EQ‐5D, range of scores from o to 1; higher scores indicate better quality of life) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

1.3.1 Early (≤ 4 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.3.2 At 12 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.4 Mortality (12 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

1.5 Unplanned return to theatre (end of follow‐up) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

1.6 Pain (using PNRS, range of scores from 0 to 11: lower values indicate less pain) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

1.6.1 Early (reported at ≤ 4 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.6.2 Reported at 12 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.7 Adverse events related to the implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

1.7.1 Intraoperative periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
1.7.2 Postoperative periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
1.7.3 Loosening	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
1.7.4 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
1.7.5 Dislocation	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Comparison 1. THA: cemented vs uncemented

Ir a la tabla de la revisión

Comparison 2. HA: cemented vs uncemented

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Early ADL (≤ 4 months, continuous data) Show forest plot	4	1275	Std. Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.21, 0.16]

2.2 Early ADL (≤ 4 months, categorical data) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

2.3 ADL (12 months, continuous data) Show forest plot	5	1173	Std. Mean Difference (IV, Random, 95% CI)	‐0.09 [‐0.21, 0.02]

2.3.1 First generation uncemented stem	1	106	Std. Mean Difference (IV, Random, 95% CI)	‐0.18 [‐0.56, 0.20]
2.3.2 Modern stem	4	1067	Std. Mean Difference (IV, Random, 95% CI)	‐0.08 [‐0.20, 0.04]
2.4 ADL (12 months, categorical data) Show forest plot	1		Risk Ratio (IV, Fixed, 95% CI)	Totals not selected

2.5 Late ADL (> 24 months; categorical data) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

2.6 Delirium (end of follow‐up) Show forest plot	2	800	Risk Ratio (M‐H, Random, 95% CI)	1.06 [0.55, 2.06]

2.6.1 First generation uncemented stem	1	400	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.14, 7.03]
2.6.2 Modern stem	1	400	Risk Ratio (M‐H, Random, 95% CI)	1.07 [0.53, 2.16]
2.7 Early functional status (≤ 4 months, continuous data) Show forest plot	3	416	Mean Difference (IV, Random, 95% CI)	3.38 [0.05, 6.70]

2.7.1 First generation uncemented stem	1	79	Mean Difference (IV, Random, 95% CI)	4.21 [1.77, 6.65]
2.7.2 Modern stem	2	337	Mean Difference (IV, Random, 95% CI)	2.43 [‐4.42, 9.29]
2.8 Early functional status (≤ 4 months; categorical data) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

2.9 Early functional status: extracapsular fractures (≤ 4 months. HHS; higher scores indicate better function) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.10 Functional status (12 months; continuous data using HHS, OHS and VELCA; higher scores indicate better function) Show forest plot	5	494	Std. Mean Difference (IV, Random, 95% CI)	0.13 [‐0.09, 0.35]

2.10.1 First generation uncemented stem	2	166	Std. Mean Difference (IV, Random, 95% CI)	0.32 [‐0.30, 0.94]
2.10.2 Modern stem	3	328	Std. Mean Difference (IV, Random, 95% CI)	0.04 [‐0.18, 0.25]
2.11 Functional status (12 months, categorical data using HHS) Show forest plot	2	100	Risk Ratio (M‐H, Random, 95% CI)	1.15 [0.91, 1.45]

2.12 Functional status: extracapsular fractures (12 months. HHS; higher scores indicate improved function) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.13 Late functional status (> 24 months using HHS; higher scores indicate better function) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.14 Early HRQoL (≤ 4 months) Show forest plot	3	1122	Std. Mean Difference (IV, Random, 95% CI)	0.20 [0.07, 0.34]

2.15 HRQoL (12 months) Show forest plot	3	1079	Std. Mean Difference (IV, Random, 95% CI)	0.12 [‐0.00, 0.24]

2.16 Late HRQoL (> 24 months) Show forest plot	1	71	Mean Difference (IV, Fixed, 95% CI)	‐0.09 [‐0.23, 0.05]

2.17 Early mobility (≤ 4 months, independent mobility) Show forest plot	3	980	Risk Ratio (M‐H, Random, 95% CI)	1.04 [0.95, 1.14]

2.17.1 First generation uncemented stem	1	75	Risk Ratio (M‐H, Random, 95% CI)	1.58 [0.84, 2.95]
2.17.2 Modern stem	2	905	Risk Ratio (M‐H, Random, 95% CI)	1.03 [0.95, 1.12]
2.18 Early mobility (≤ 4 months, continuous data) Show forest plot	3	766	Std. Mean Difference (IV, Random, 95% CI)	‐0.26 [‐0.40, ‐0.12]

2.18.1 First generation uncemented stem	1	327	Std. Mean Difference (IV, Random, 95% CI)	‐0.36 [‐0.58, ‐0.14]
2.18.2 Modern stem	2	439	Std. Mean Difference (IV, Random, 95% CI)	‐0.18 [‐0.37, 0.00]
2.19 Early mobility (mean reduction values at ≤ 4 months; higher scores indicate better mobility) Show forest plot	1	110	Mean Difference (IV, Fixed, 95% CI)	‐0.40 [‐0.68, ‐0.12]

2.20 Mobility (12 months, continuous data using different mobility scales; lower scores indicate better mobility) Show forest plot	4	762	Std. Mean Difference (IV, Random, 95% CI)	‐0.24 [‐0.42, ‐0.06]

2.20.1 First generation uncemented stem	2	386	Std. Mean Difference (IV, Random, 95% CI)	‐0.12 [‐0.54, 0.30]
2.20.2 Modern stem	2	376	Std. Mean Difference (IV, Random, 95% CI)	‐0.32 [‐0.53, ‐0.12]
2.21 Mobility (12 months, independent mobility) Show forest plot	3	826	Risk Ratio (M‐H, Random, 95% CI)	0.98 [0.70, 1.37]

2.21.1 First generation uncemented stem	1	75	Risk Ratio (M‐H, Random, 95% CI)	1.22 [0.81, 1.82]
2.21.2 Modern	2	751	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.52, 1.55]
2.22 Mobility (12 months, dependent on walking aid) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

2.23 Late mobility (> 24 months) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.24 Late mobility (> 24 months; independent mobility) Show forest plot	1	79	Risk Ratio (M‐H, Fixed, 95% CI)	0.88 [0.75, 1.02]

2.25 Early mortality (≤ 4 months) Show forest plot	12	3136	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.80, 1.13]

2.25.1 First generation uncemented stem	7	980	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.79, 1.54]
2.25.2 Modern stem	5	2156	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.73, 1.10]
2.26 Mortality (12 months) Show forest plot	15	3727	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.78, 0.96]

2.26.1 First generation uncemented stem	8	1036	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.78, 1.18]
2.26.2 Modern stem	7	2691	Risk Ratio (M‐H, Random, 95% CI)	0.83 [0.73, 0.94]
2.27 Late mortality (> 24 months) Show forest plot	2	620	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.89, 1.15]

2.28 Unplanned return to theatre (end of follow‐up) Show forest plot	6	2336	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.45, 1.10]

2.28.1 First generation uncemented stem	1	400	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.30, 1.26]
2.28.2 Modern stem	5	1936	Risk Ratio (M‐H, Random, 95% CI)	0.77 [0.44, 1.35]
2.29 Early pain (≤ 4 months, experiencing no pain) Show forest plot	4	500	Risk Ratio (M‐H, Random, 95% CI)	1.11 [1.00, 1.22]

2.29.1 First generation uncemented stem	2	194	Risk Ratio (M‐H, Random, 95% CI)	1.26 [0.81, 1.97]
2.29.2 Modern stem	2	306	Risk Ratio (M‐H, Random, 95% CI)	1.05 [0.88, 1.24]
2.30 Early pain (≤ 4 months; mean scores, lower scores indicate less pain) Show forest plot	4	1507	Mean Difference (IV, Fixed, 95% CI)	0.01 [‐0.09, 0.11]

2.30.1 First generation uncemented stem	1	320	Mean Difference (IV, Fixed, 95% CI)	‐0.60 [‐0.87, ‐0.33]
2.30.2 Modern stem	3	1187	Mean Difference (IV, Fixed, 95% CI)	0.10 [‐0.01, 0.20]
2.31 Pain (12 months, experiencing no pain) Show forest plot	4	376	Risk Ratio (M‐H, Random, 95% CI)	1.17 [0.85, 1.63]

2.31.1 First generation uncemented stem	2	114	Risk Ratio (M‐H, Random, 95% CI)	2.09 [0.97, 4.48]
2.31.2 Modern stem	2	262	Risk Ratio (M‐H, Random, 95% CI)	0.93 [0.82, 1.06]
2.32 Pain (12 months, using continuous data; lower values indicate less pain) Show forest plot	5	1305	Std. Mean Difference (IV, Random, 95% CI)	‐0.02 [‐0.21, 0.18]

2.32.1 First generation uncemented stem	1	272	Std. Mean Difference (IV, Random, 95% CI)	‐0.32 [‐0.56, ‐0.08]
2.32.2 Modern stem	4	1033	Std. Mean Difference (IV, Random, 95% CI)	0.09 [‐0.03, 0.21]
2.33 Pain (12 months; mean reduction values: lower scores indicate less pain) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.34 Late pain (> 24 months, using mean scores; lower scores indicate less pain) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.35 Late pain (> 24 months; experiencing no pain) Show forest plot	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	1.00 [0.77, 1.30]

2.36 Length of hospital stay (days) Show forest plot	9	1741	Mean Difference (IV, Random, 95% CI)	‐0.40 [‐1.03, 0.23]

2.36.1 First generation uncemented stem	5	765	Mean Difference (IV, Random, 95% CI)	‐0.64 [‐1.84, 0.55]
2.36.2 Modern stem	4	976	Mean Difference (IV, Random, 95% CI)	‐0.43 [‐1.73, 0.88]
2.37 Discharge destination (own home) Show forest plot	6	2231	Risk Ratio (M‐H, Random, 95% CI)	1.05 [0.98, 1.13]

2.37.1 First generation uncemented stem	2	501	Risk Ratio (M‐H, Random, 95% CI)	0.98 [0.71, 1.34]
2.37.2 Modern stem	4	1730	Risk Ratio (M‐H, Random, 95% CI)	1.07 [0.95, 1.20]
2.38 Adverse events related to the implant, fracture, or both Show forest plot	14		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

2.38.1 Intraoperative periprosthetic fracture	7	1669	Risk Ratio (M‐H, Random, 95% CI)	0.20 [0.08, 0.46]
2.38.2 Postoperative periprosthetic fracture	8	2819	Risk Ratio (M‐H, Random, 95% CI)	0.29 [0.14, 0.57]
2.38.3 Loosening	4	537	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.14, 1.89]
2.38.4 Deep infection	7	1382	Risk Ratio (M‐H, Random, 95% CI)	1.56 [0.72, 3.38]
2.38.5 Superficial infection	10	3038	Risk Ratio (M‐H, Random, 95% CI)	1.23 [0.73, 2.06]
2.38.6 Dislocation	10	3032	Risk Ratio (M‐H, Random, 95% CI)	1.08 [0.61, 1.91]
2.39 Adverse events unrelated to the implant, fracture, or both Show forest plot	11		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

2.39.1 Acute kidney injury	4	2226	Risk Ratio (M‐H, Random, 95% CI)	1.23 [0.76, 2.00]
2.39.2 Blood transfusion	7	2907	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.83, 1.20]
2.39.3 Cerebrovascular accident	5	2356	Risk Ratio (M‐H, Random, 95% CI)	0.93 [0.41, 2.10]
2.39.4 Pneumonia/chest infection	8	2789	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.50, 1.21]
2.39.5 Myocardial infarction	7	2682	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.44, 1.89]
2.39.6 Urinary tract infection	5	1745	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.65, 1.20]
2.39.7 Venous thromboembolic phenomena (DVT)	7	2661	Risk Ratio (M‐H, Random, 95% CI)	1.28 [0.56, 2.90]
2.39.8 Venous thromboembolic phenomena (pulmonary embolism)	6	2499	Risk Ratio (M‐H, Random, 95% CI)	3.56 [1.26, 10.11]

Comparison 2. HA: cemented vs uncemented

Ir a la tabla de la revisión

Comparison 3. Mixed HA and THA: cemented vs uncemented

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Functional status (12 months, using HHS, range of scores from 0 to 100; higher scores indicate better function) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

3.2 HRQoL (12 months, using SF‐36, range of scores from 0 to 100; higher scores indicate better quality of life) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

3.3 Early mortality (≤ 4 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

3.4 Mortality (12 months) Show forest plot	2	169	Risk Ratio (M‐H, Random, 95% CI)	2.02 [0.81, 5.07]

3.5 Late mortality (> 24 months) Show forest plot	1	141	Risk Ratio (M‐H, Fixed, 95% CI)	0.88 [0.50, 1.56]

3.6 Pain (12 months, using HHS pain scales; higher values indicate less pain) Show forest plot	1	106	Mean Difference (IV, Fixed, 95% CI)	2.60 [‐0.87, 6.07]

3.7 Pain (> 24 months, using HHS pain scales; higher values indicate less pain) Show forest plot	1	86	Mean Difference (IV, Fixed, 95% CI)	3.60 [‐0.01, 7.21]

3.8 Unplanned return to theatre (end of follow‐up) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

3.9 Adverse events related to the implant, fracture, or both Show forest plot	2		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

3.9.1 Intraoperative periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
3.9.2 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
3.9.3 Dislocation	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
3.10 Adverse events unrelated to implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3.10.1 Acute kidney infection	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3.10.2 Chest infection/pneumonia	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3.10.3 Myocardial infarction	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3.10.4 Urinary tract infection	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Comparison 3. Mixed HA and THA: cemented vs uncemented

Ir a la tabla de la revisión

Comparison 4. Bipolar HA vs unipolar HA

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 ADL (12 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.2 Delirium/confusion Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.3 Functional status (12 months; using different measurement tools; higher scores indicate better function) Show forest plot	2	299	Std. Mean Difference (IV, Random, 95% CI)	‐0.04 [‐0.27, 0.19]

4.4 Functional status (12 months. HHS; excellent and good) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.5 Functional status (> 24 months. HHS; excellent or good) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.6 Early HRQoL (≤ 4 months) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.7 HRQoL (12 months) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.8 Mobility (Get up and Go Test; in seconds) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.9 Mobility (6 minute walk test; in metres) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.10 Early mortality (≤ 4 months) Show forest plot	4	573	Risk Ratio (M‐H, Random, 95% CI)	0.94 [0.54, 1.64]

4.11 Mortality (12 months) Show forest plot	8	839	Risk Ratio (M‐H, Random, 95% CI)	1.17 [0.89, 1.53]

4.11.1 Cemented	7	799	Risk Ratio (M‐H, Random, 95% CI)	1.18 [0.90, 1.55]
4.11.2 Uncemented	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.01, 7.72]
4.12 Late mortality (> 24 months) Show forest plot	2	362	Risk Ratio (M‐H, Random, 95% CI)	0.94 [0.72, 1.23]

4.13 Unplanned return to theatre (end of follow‐up) Show forest plot	4	532	Risk Ratio (M‐H, Random, 95% CI)	1.08 [0.44, 2.64]

4.13.1 Cemented	3	482	Risk Ratio (M‐H, Random, 95% CI)	1.71 [0.73, 3.99]
4.13.2 Cemented and uncemented	1	50	Risk Ratio (M‐H, Random, 95% CI)	0.29 [0.07, 1.24]
4.14 Pain (categorical data; no pain, or mild pain) Show forest plot	2	300	Risk Ratio (M‐H, Random, 95% CI)	1.22 [0.82, 1.82]

4.15 Pain (12 months) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.16 Length of hospital stay (days) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.17 Discharge destination: return to preoperative residence Show forest plot	2	381	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.84, 1.08]

4.18 Adverse events related to implant, fracture, or both Show forest plot	10		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

4.18.1 Periprosthetic fracture	1	120	Risk Ratio (M‐H, Random, 95% CI)	7.00 [0.37, 132.66]
4.18.2 Superficial infection	1	261	Risk Ratio (M‐H, Random, 95% CI)	2.41 [0.48, 12.18]
4.18.3 Deep infection	7	1122	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.44, 2.71]
4.18.4 Dislocation	9	1274	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.28, 1.38]
4.19 Adverse event unrelated to implant, fracture, or both Show forest plot	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

4.19.1 Acute kidney injury	1	261	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 70.25]
4.19.2 Blood transfusion	1	115	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.51, 1.62]
4.19.3 Cerebrovascular accident	2	436	Risk Ratio (M‐H, Random, 95% CI)	1.57 [0.20, 12.69]
4.19.4 Pneumonia/chest infection	3	556	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.10, 3.86]
4.19.5 Myocardial infarction	3	556	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.11, 4.32]
4.19.6 Urinary tract infection	1	261	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.29, 3.25]
4.19.7 Venous thromboembolic phenomena (DVT)	2	381	Risk Ratio (M‐H, Random, 95% CI)	3.84 [0.43, 34.45]
4.19.8 Venous thromboembolic phenomena (pulmonary embolism)	1	120	Risk Ratio (M‐H, Random, 95% CI)	3.00 [0.12, 72.20]

Comparison 4. Bipolar HA vs unipolar HA

Ir a la tabla de la revisión

Comparison 5. HA: short stem vs standard stem

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Mobility (24 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

5.2 Mortality (24 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

5.3 Pain (24 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

5.4 Adverse events related to implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

5.4.1 Postoperative periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
5.4.2 Loosening	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
5.4.3 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
5.4.4 Dislocation	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Comparison 5. HA: short stem vs standard stem

Ir a la tabla de la revisión

Comparison 6. HA: ETS vs Thompson

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Delirium Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

6.2 Early HRQoL (≤ 4 months) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

6.3 Early mobility (freely mobile without aids, or able to walk outdoors with one aid) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

6.4 Early mortality (≤ 4 months) Show forest plot	2	1164	Risk Ratio (M‐H, Random, 95% CI)	1.20 [0.76, 1.88]

6.5 Mortality (12 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

6.6 Unplanned return to theatre (end of follow‐up) Show forest plot	2	1164	Risk Ratio (M‐H, Random, 95% CI)	0.46 [0.05, 3.89]

6.7 Adverse events related to implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

6.7.1 Intraoperative periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.7.2 Deep infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.7.3 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.7.4 Dislocation	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8 Adverse events unrelated to implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

6.8.1 Acute kidney injury	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8.2 Blood transfusion	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8.3 Cerebrovascular accident	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8.4 Chest infection/pneumonia	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8.5 Myocardial infarction	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8.6 Venous thromboembolic phenomena (DVT)	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.8.7 Venous thromboembolic phenomena (pulmonary embolism)	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Comparison 6. HA: ETS vs Thompson

Ir a la tabla de la revisión

Comparison 7. HA: Furlong vs Moore

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Early mortality (≤ 4 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

7.2 Mortality (12 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

7.3 Unplanned return to theatre (at end of follow‐up) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

7.4 Pain at rest Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

7.5 Adverse events related to the implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

7.5.1 Periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
7.5.2 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
7.5.3 Dislocation	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Comparison 7. HA: Furlong vs Moore

Ir a la tabla de la revisión

Comparison 8. THA vs HA

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Early ADL (≤ 4 months, using categorical data) Show forest plot	2	225	Risk Ratio (M‐H, Random, 95% CI)	1.03 [0.91, 1.18]

8.2 Early ADL (≤ 4 months; using social mobility scale (lower scores indicate better mobility) Show forest plot	1	83	Mean Difference (IV, Fixed, 95% CI)	‐0.10 [‐0.46, 0.26]

8.3 ADL (12 months, using categorical data) Show forest plot	2	217	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.86, 1.07]

8.4 ADL (12 months; using different measurement tools; lower scores indicate more independence)) Show forest plot	2		Std. Mean Difference (IV, Random, 95% CI)	Totals not selected

8.5 Late ADL (> 24 months; using Barthel Index, range of scores from 0 to 100; higher scores indicate more independence) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

8.6 Delirium Show forest plot	2	357	Risk Ratio (M‐H, Random, 95% CI)	1.41 [0.60, 3.33]

8.7 Early functional status (≤ 4 months) Show forest plot	3	395	Std. Mean Difference (IV, Random, 95% CI)	0.27 [0.07, 0.47]

8.8 Functional status (12 months) Show forest plot	8	1273	Std. Mean Difference (IV, Random, 95% CI)	0.29 [0.14, 0.44]

8.9 Functional status (HHS; excellent or good) Show forest plot	2	140	Risk Ratio (M‐H, Random, 95% CI)	1.07 [0.98, 1.17]

8.10 Late functional status (> 24 months; using OHS and HHS; higher scores indicate better function) Show forest plot	4	224	Std. Mean Difference (IV, Random, 95% CI)	0.65 [0.23, 1.08]

8.11 Early HRQoL (≤ 4 months) Show forest plot	2	279	Mean Difference (IV, Random, 95% CI)	0.03 [‐0.06, 0.12]

8.12 HRQoL (12 months) Show forest plot	4	1158	Std. Mean Difference (IV, Random, 95% CI)	0.19 [0.07, 0.31]

8.13 HRQoL (> 24 months. Using SF‐36; higher scores indicate better quality of life) Show forest plot	1	34	Mean Difference (IV, Fixed, 95% CI)	5.90 [‐1.99, 13.79]

8.14 Early mobility (≤ 4 months; lower scores indicate better mobility Show forest plot	1	83	Mean Difference (IV, Fixed, 95% CI)	‐0.40 [‐0.96, 0.16]

8.15 Mobility (12 months, using TUG; lower values indicate better mobility) Show forest plot	2	575	Mean Difference (IV, Random, 95% CI)	‐2.74 [‐6.82, 1.35]

8.16 Mobility (12 months, using 9‐point mobility scale; lower scores indicate better mobility) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

8.17 Mobility (12 months; able to ambulate independently) Show forest plot	2	175	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.71, 1.31]

8.17.1 Modern design	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.53, 1.17]
8.17.2 First generation uncemented stem	1	135	Risk Ratio (M‐H, Random, 95% CI)	1.09 [0.87, 1.36]
8.18 Late mobility (> 24 months; able to ambulate independently) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

8.19 Early mortality (≤ 4 months) Show forest plot	6	725	Risk Ratio (M‐H, Random, 95% CI)	0.77 [0.42, 1.42]

8.19.1 Modern design	4	465	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.32, 2.41]
8.19.2 First generation uncemented stem design	1	180	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.30, 1.44]
8.19.3 Age of design is unknown	1	80	Risk Ratio (M‐H, Random, 95% CI)	3.00 [0.13, 71.51]
8.20 Mortality (12 months) Show forest plot	11	2667	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.83, 1.22]

8.20.1 Modern design	10	2487	Risk Ratio (M‐H, Random, 95% CI)	1.03 [0.82, 1.28]
8.20.2 First generation uncemented stem design	1	180	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.52, 1.42]
8.21 Late mortality (> 24 months) Show forest plot	8	931	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.81, 1.23]

8.21.1 Modern design	7	751	Risk Ratio (M‐H, Random, 95% CI)	0.97 [0.72, 1.32]
8.21.2 First generation uncemented stem design	1	180	Risk Ratio (M‐H, Random, 95% CI)	0.97 [0.86, 1.10]
8.22 Unplanned return to theatre (end of follow‐up) Show forest plot	10	2594	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.37, 1.07]

8.22.1 Modern design	9	2414	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.59, 1.25]
8.22.2 First generation uncemented stem design	1	180	Risk Ratio (M‐H, Random, 95% CI)	0.28 [0.12, 0.66]
8.23 Length of hospital stay (days) Show forest plot	3	306	Mean Difference (IV, Random, 95% CI)	0.80 [‐1.12, 2.73]

8.24 Pain (12 months: data not combined; lower scores indicate less pain) Show forest plot	9	1435	Std. Mean Difference (IV, Random, 95% CI)	‐0.13 [‐0.38, 0.12]

8.25 Late pain (> 24 months) Show forest plot	2		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

8.26 Pain (> 24 months: categorical data: no pain) Show forest plot	1	135	Risk Ratio (M‐H, Fixed, 95% CI)	1.47 [1.07, 2.00]

8.27 Early pain (≤ 4 months: higher scores indicate less pain) Show forest plot	5	572	Std. Mean Difference (IV, Random, 95% CI)	0.10 [‐0.10, 0.30]

8.28 Discharge destination (own home) Show forest plot	2	1612	Risk Ratio (M‐H, Random, 95% CI)	0.97 [0.87, 1.08]

8.29 Discharge destination (geriatric ward) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

8.30 Adverse events related to implant, fracture, or both Show forest plot	14		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

8.30.1 Postoperative perioprosthetic fracture	3	1557	Risk Ratio (M‐H, Random, 95% CI)	1.08 [0.70, 1.66]
8.30.2 Prosthetic loosening	4	1889	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.17, 2.41]
8.30.3 Deep infection	8	2343	Risk Ratio (M‐H, Random, 95% CI)	0.87 [0.50, 1.54]
8.30.4 Superficial infection	10	2495	Risk Ratio (M‐H, Random, 95% CI)	1.25 [0.67, 2.30]
8.30.5 Dislocation	12	2719	Risk Ratio (M‐H, Random, 95% CI)	1.96 [1.17, 3.27]
8.31 Adverse events unrelated to implant, fracture, or both Show forest plot	8		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

8.31.1 Acute kidney injury	2	1561	Risk Ratio (M‐H, Random, 95% CI)	1.09 [0.62, 1.92]
8.31.2 Blood transfusion	2	285	Risk Ratio (M‐H, Random, 95% CI)	2.14 [1.27, 3.61]
8.31.3 Cerebrovascular accident	4	657	Risk Ratio (M‐H, Random, 95% CI)	1.63 [0.63, 4.21]
8.31.4 Pneumonia/chest infection (reported at > 4 months)	5	613	Risk Ratio (M‐H, Random, 95% CI)	0.87 [0.38, 2.00]
8.31.5 Myocardial infarction	4	460	Risk Ratio (M‐H, Random, 95% CI)	1.48 [0.48, 4.58]
8.31.6 Urinary tract infection	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.01, 3.46]
8.31.7 Venous thromboembolic phenomena (DVT)	4	486	Risk Ratio (M‐H, Random, 95% CI)	4.25 [0.86, 21.06]
8.31.8 Venous thromboembolic phenomena (pulmonary embolism)	5	673	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.14, 1.63]

Comparison 8. THA vs HA

Ir a la tabla de la revisión

Comparison 9. THA: single articulation vs dual‐mobility

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Early functional status (≤ 4 months, using different scales; higher scores indicate better function) Show forest plot	2	78	Std. Mean Difference (IV, Random, 95% CI)	‐0.33 [‐0.78, 0.12]

9.2 Functional status (12 months, using OHS and HHS; higher scores indicate better function) Show forest plot	2	79	Std. Mean Difference (IV, Random, 95% CI)	‐0.60 [‐1.05, ‐0.15]

9.3 HRQoL (using EQ‐5D, range of scores from 0 to 1; higher scores indicate better quality of life) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

9.3.1 Early (≤ 4 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
9.3.2 At 12 months	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
9.4 Mortality (12 months) Show forest plot	2	82	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.08, 4.77]

9.5 Adverse events related to the implant, fracture, or both Show forest plot	2		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

9.5.1 Deep infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
9.5.2 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
9.5.3 Dislocation	2		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
9.6 Adverse events unrelated to the implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

9.6.1 Venous thromboembolic phenomena	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Comparison 9. THA: single articulation vs dual‐mobility

Ir a la tabla de la revisión

Comparison 10. THA: short stem vs standard stem

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 Functional status (at 24 months; using HHS, range of scores from 0 to 100; higher scores indicate better function) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

10.2 Mobility Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

10.3 Mortality (12 months) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

10.4 Pain Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

10.5 Adverse events related to the implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

10.5.1 Intraoperative periprosthetic fracture	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
10.5.2 Superficial infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
10.5.3 Dislocation	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
10.6 Adverse events unrelated to the implant, fracture, or both Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

10.6.1 Acute kidney injury	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
10.6.2 Chest infection/pneumonia	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
10.6.3 Urinary tract infection	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

Comparison 10. THA: short stem vs standard stem

Ir a la tabla de la revisión

	Idioma de la Revisión Cochrane Escoja su idioma de preferencia para las revisiones Cochrane y otros contenidos. Las secciones sin una traducción aparecerán en inglés.

	Idioma de la web Escoja su idioma de preferencia para la web de la Biblioteca Cochrane.

Idioma de la Revisión Cochrane

Idioma de la web

Resumen

Antecedentes

Objetivos

Métodos de búsqueda

Criterios de selección

Obtención y análisis de los datos

Resultados principales

Conclusiones de los autores

PICO

PICO

Population

Intervention

Comparison

Outcome

Resumen en términos sencillos

Cirugía de prótesis de cadera en adultos

Resumen visual

Authors' conclusions

Implications for practice

Implications for research

Summary of findings

Background

Description of the condition

Epidemiology

Burden of disease

Types of hip fracture

Description of the intervention

Arthroplasty

Hemiarthroplasty

Total hip arthroplasty

Component fixation

Cemented systems

Uncemented systems

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Critical outcomes

Other important outcomes

Related

Unrelated

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Summary of findings and assessment of the certainty of the evidence

Results

Description of studies

Results of the search

Included studies

Types of studies and setting

Types of participants

Types of interventions

Types of outcome measures

Sources of funding and declarations of interest

Excluded studies

Studies awaiting classification

Ongoing studies

Risk of bias in included studies

Allocation

Blinding

Incomplete outcome data

Selective reporting