Preparación de la piel para la prevención de la infección después de una cesárea
Resumen
Antecedentes
El riesgo de mortalidad y morbilidad materna es mayor después de una cesárea que en el caso de un parto vaginal. Con el aumento de las tasas de cesárea, es importante minimizar los riesgos para la madre tanto como sea posible. Esta revisión se centró en diferentes preparaciones de la piel para prevenir la infección. Ésta es una actualización de una revisión publicada en 2018.
Objetivos
Comparar los efectos de diferentes agentes antisépticos, diferentes métodos de aplicación o diferentes formas de antisépticos utilizados para la preparación de la piel en el preoperatorio para prevenir la infección poscesárea.
Métodos de búsqueda
Para esta actualización, se hicieron búsquedas en el registro de ensayos del Grupo Cochrane de Embarazo y Parto (Cochrane Pregnancy and Childbirth's Trials Register), ClinicalTrials.gov, en la Plataforma de registros internacionales de ensayos clínicos de la Organización Mundial de la Salud (OMS) (ICTRP) (9 de julio de 2019), y en las listas de referencias de los estudios recuperados.
Criterios de selección
Ensayos aleatorizados y cuasialeatorizados, que evalúan cualquier tipo de preparación cutánea preoperatoria (agentes, métodos o formas). Se incluyeron estudios presentados sólo como resúmenes, si había suficiente información para evaluar el riesgo de sesgo.
Las comparaciones de interés en esta revisión fueron entre: diferentes agentes antisépticos (p.ej., alcohol, povidona yodada), diferentes métodos de aplicación de los antisépticos (p.ej., lavado, tinturas, paños), diferentes formas de antisépticos (p.ej., polvo, líquido), y también entre diferentes paquetes de preparación de la piel incluyendo una mezcla de agentes y métodos, como un paño plástico para la incisión, que puede o no estar impregnado con agentes antisépticos. La revisión se focalizó principalmente en la comparación entre diferentes agentes, con y sin el uso de paños.
Solo se incluyeron estudios que examinaron la preparación de la zona de incisión. Esta revisión no cubrió los estudios sobre el lavado de manos preoperatorio por el equipo de cirugía ni el baño preoperatorio.
Obtención y análisis de los datos
Tres autores de la revisión evaluaron de forma independiente todos los estudios potenciales para su inclusión, evaluaron el riesgo de sesgo, extrajeron los datos y verificaron su exactitud. La certeza de la evidencia se evaluó mediante los criterios GRADE.
Resultados principales
Se incluyeron 13 ensayos controlados aleatorizados (ECA) individuales, con un total de 6938 mujeres que se sometieron a una cesárea. Doce ensayos (6916 mujeres) aportaron datos a esta revisión. Las fechas de los ensayos variaron entre 1983 y 2016. Se realizaron seis ensayos en los Estados Unidos y el resto en la India, Egipto, Nigeria, Sudáfrica, Francia, Dinamarca e Indonesia.
En general, los estudios incluidos presentaban un bajo riesgo de sesgo en la mayoría de los dominios, aunque el alto riesgo de sesgo de detección suscitó algunas preocupaciones específicas en varios estudios. La duración de la estancia sólo se informó en una comparación.
Agentes antisépticos
Paraclorometaxilenol con yodo versus yodo solo
No se sabe con certeza si el paraclorometaxilenol con yodo marcó alguna diferencia en la incidencia de la infección del sitio quirúrgico (riesgos relativos (RR) 0,33; intervalo de confianza (IC) del 95%: 0,04 a 2,99; un ensayo, 50 mujeres) o la endometritis (RR 0,88; IC del 95%: 0,56 a 1,38; un ensayo, 50 mujeres) en comparación con el yodo solo, porque la certeza de la evidencia fue muy baja. No se informaron los eventos adversos (maternos o neonatales).
Gluconato de clorhexidina versus povidona yodada
Evidencia de certeza moderada indicó que el gluconato de clorhexidina, comparado con la povidona yodada, probablemente reduce ligeramente la incidencia de la infección del sitio quirúrgico (RR 0,72; IC del 95%: 0,58 a 0,91; ocho ensayos, 4323 mujeres). Este efecto todavía estaba presente en un análisis de sensibilidad después de eliminar cuatro ensayos de alto riesgo de sesgo para la evaluación de resultados (RR 0,87; IC del 95%: 0,62 a 1,23; cuatro ensayos, 2037 mujeres).
Evidencia de certeza baja indicó que el gluconato de clorhexidina, comparado con la povidona yodada, puede lograr poca o ninguna diferencia en la incidencia de la endometritis (RR 0,95; IC del 95%: 0,49 a 1,86; tres ensayos, 2484 mujeres). No se sabe con certeza si el gluconato de clorhexidina reduce la irritación de la piel materna o la reacción alérgica de la piel (RR 0,64; IC del 95%: 0,28 a 1,46; tres ensayos, 1926 mujeres; evidencia de muy baja certeza).
Un estudio pequeño (60 mujeres) informó de una reducción del crecimiento bacteriano a las 18 horas después de la cesárea para las mujeres que recibieron un preparado de gluconato de clorhexidina, en comparación con las mujeres que recibieron un preparado de povidona yodada (RR 0,23; IC del 95%: 0,07 a 0,70).
Métodos
Paños versus ningún paño
Esta comparación investigó el uso de un paño versus ningún paño, después de la preparación de la piel con antisépticos.
Evidencia de certeza baja indicó que el uso de un paño antes de la cirugía comparado con ningún paño, puede lograr poca o ninguna diferencia en la incidencia de la infección del sitio quirúrgico (RR 1,29; intervalo de confianza (IC) del 95%: 0,97 a 1,71; tres ensayos, 1373 mujeres), y probablemente logra poca o ninguna diferencia en la duración de la estancia en el hospital (diferencia de medias (DM) 0,10 días, IC del 95%: ‐0,27 a 0,46; un ensayo, 603 mujeres; evidencia de certeza moderada). Un ensayo comparó un lavado con alcohol y un paño con yodóforo con un lavado con yodóforo de cinco minutos solamente, y no informó infecciones del sitio quirúrgico en los grupos (79 mujeres, evidencia de certeza muy baja). No se conoce si la combinación de un minuto de lavado con alcohol y un paño reduce la incidencia de metritis, en comparación con cinco minutos de lavado, ya que la certeza de la evidencia fue muy baja (RR 1,62; IC del 95%: 0,29 a 9,16; un ensayo, 79 mujeres). Los estudios no informaron sobre eventos adversos (maternos o neonatales).
Conclusiones de los autores
Evidencia de certeza moderada indica que la preparación de la piel con gluconato de clorhexidina antes de la cesárea es probablemente un poco más efectiva para reducir la incidencia de la infección del sitio quirúrgico en comparación con la povidona yodada. En cuanto a otros resultados examinados, no se disponía de evidencia suficiente de los ECA incluidos. La mayoría de la evidencia de esta revisión se consideró de certeza muy baja. Lo anterior significa que para la mayoría de los hallazgos, la confianza en cualquier evidencia de un efecto de intervención es limitada, e indica la necesidad de estudios de investigación de mayor calidad. Por lo tanto, todavía no está claro qué tipo de preparación de la piel puede ser más efectivo para prevenir la infección del sitio quirúrgico después de la cesárea, o para reducir otros resultados indeseables para la madre y el recién nacido.
Se necesitan ECA bien diseñados, con tamaños de muestra más grandes. Entre las preguntas de alta prioridad, figuran la comparación de los tipos de antisépticos (especialmente yodo versus clorhexidina) y los métodos de aplicación (frotar, lavar o cubrir). Se encontraron dos estudios que están en curso; se incorporarán los resultados de estos estudios en futuras actualizaciones de esta revisión.
PICO
Resumen en términos sencillos
Preparación de la piel para la prevención de la infección después de una cesárea
Ésta es una actualización de una revisión publicada por primera vez en 2012 y actualizada en 2014 y 2018.
¿Cuál es el problema? El objetivo de esta revisión Cochrane fue determinar qué métodos de preparación de la piel antes de la cesárea fueron más efectivos para prevenir la infección después de la operación. Se recopilaron y analizaron todos los estudios que evaluaron la efectividad de los antisépticos utilizados para preparar la piel antes de hacer una incisión (o corte) para una cesárea. Solo se incluyó el análisis de los compuestos que se utilizaron para preparar el sitio quirúrgico en el abdomen antes de la cesárea; no se examinó el lavado de manos por parte del equipo de cirugía, ni la opción de bañar a la madre.
¿Por qué es importante?
Las infecciones de las incisiones quirúrgicas son las terceras infecciones adquiridas en el hospital que más frecuentemente se comunican. Las mujeres que dan a luz por cesárea están expuestas a la infección de gérmenes, ya presentes en la propia piel de la madre, o de fuentes externas. El riesgo de infección después de una cesárea puede ser diez veces mayor que el del parto vaginal. Por lo tanto, la prevención de la infección mediante la preparación adecuada de la piel antes de hacer la incisión es una parte importante de los cuidados generales que les brindan a las mujeres antes del parto por cesárea. Un antiséptico es una sustancia que se aplica para eliminar las bacterias que pueden causar daño a la madre o al recién nacido cuando se multiplican. Los antisépticos incluyen el yodo o la povidona yodada, el alcohol, la clorhexidina y el paraclorometaxileno. Se pueden aplicar en forma de líquidos o polvos, exfoliantes, tinturas, hisopos, o en "paños" impregnados que se adhieren a la piel, que el cirujano luego corta. También se pueden aplicar paños no impregnados, una vez que la piel ha sido frotada, con el objetivo de reducir la propagación de cualquier bacteria restante durante la cirugía. Es importante conocer si alguno de estos antisépticos o métodos funciona mejor que otros.
¿Qué evidencia se encontró? Esta revisión actualizada incluyó 13 ensayos con 6938 mujeres. Se realizaron seis ensayos en los Estados Unidos; los restantes ensayos se realizaron en Nigeria, Sudáfrica, Francia, Dinamarca, Indonesia, India y Egipto. La revisión examinó qué fue mejor para las mujeres y los recién nacidos cuando se trataba de resultados importantes, entre ellos: la infección del sitio donde el cirujano hizo el corte para realizar la cesárea; la inflamación del revestimiento del útero (metritis y endometritis); el tiempo que la mujer permaneció en el hospital; y cualquier otro efecto adverso como la irritación de la piel de la mujer, o cualquier repercusión informada sobre el recién nacido. No todos los 13 ensayos exploraron todos estos resultados, y la evidencia para cada resultado se basó generalmente en los resultados de mucho menos de 6938 mujeres.
Gran parte de la evidencia que se encontró fue de calidad relativamente deficiente, debido a las limitaciones en la forma de realizar los estudios. Esto significa que no existe mucha seguridad acerca de la mayoría de los hallazgos. La evidencia indicó que en las mujeres a las que se les preparó la piel con el agente gluconato de clorhexidina, es probable que haya una ligera reducción de la incidencia de la infección del sitio quirúrgico en comparación con las mujeres a las que se les preparó la piel con povidona yodada. En cuanto a los demás resultados, hubo poca o ninguna diferencia entre los diversos agentes antisépticos y los métodos de aplicación en relación a la endometritis, la irritación de la piel o la reacción alérgica de la madre. En un estudio, se observó una reducción del crecimiento bacteriano en la piel a las 18 horas de la cesárea en las mujeres que recibieron un preparado cutáneo con gluconato de clorhexidina, en comparación con las mujeres que recibieron el preparado cutáneo con povidona yodada, pero se necesitan más datos para ver si esta intervención reduce realmente las infecciones en las mujeres.
¿Qué significa esto?
La evidencia disponible de los ensayos que se han realizado no fue suficiente para determinar cuál es el mejor tipo de preparación de la piel para prevenir la infección del sitio quirúrgico después de una cesárea. Se necesitan más estudios de investigación de calidad alta. Se encontraron dos estudios que aún están en curso. Los resultados de estos estudios se incorporarán a esta revisión en futuras actualizaciones.
Authors' conclusions
Summary of findings
| Parachlorometaxylenol with iodine versus iodine alone | ||||||
|---|---|---|---|---|---|---|
| Population: women undergoing caesarean section Comparison: iodine alone | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of participants | Certainty of the evidence | Comments | |
| Risk with iodine alone | Risk with parachlorometaxylenol with iodine | |||||
| Surgical site infection | Study population | RR 0.33 | 50 | ⊕⊝⊝⊝ | ||
| 120 per 1000 | 40 per 1000 | |||||
| Endometritis | Study population | RR 0.88 | 50 | ⊕⊝⊝⊝ | ||
| 640 per 1000 | 563 per 1000 | |||||
| Length of stay | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in the included study. |
| Adverse events (maternal or neonatal) | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in the included study. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a Wide confidence interval crossing the line of no effect, single study with small sample size (imprecision ‐2). b Blinding of outcome assessor was at high risk of bias (risk of bias ‐1). | ||||||
| Chlorhexidine gluconate compared to povidone iodine | ||||||
|---|---|---|---|---|---|---|
| Population: women undergoing caesarean section Comparison: povidone iodine | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | No of participants | Certainty of the evidence | Comments | |
| Risk with povidone iodine | Risk with chlorhexidine gluconate | |||||
| Surgical site infection | Study population | RR 0.72 | 4323 | ⊕⊕⊕⊝ | ||
| 75 per 1000 | 54 per 1000 | |||||
| Endometritis | Study population | RR 0.95 | 2484 | ⊕⊕⊝⊝ | ||
| 14 per 1000 | 13 per 1000 | |||||
| Length of stay | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in any of the included studies. |
| Adverse events (maternal) ‐ skin irritation or allergic skin reaction | Study population | RR 0.64 | 1926 | ⊕⊝⊝⊝ | No neonatal adverse events were reported in any of the included studies. | |
| 15 per 1000 | 9 per 1000 | |||||
| *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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a Selection bias was unclear and blinding of outcome assessor was high risk of bias (risk of bias ‐1). b Wide confidence interval crossing the line of no effect (imprecision ‐1). c Wide confidence interval crossing the line of no effect and few events (imprecision ‐2). | ||||||
| Drape compared to no drape | ||||||
|---|---|---|---|---|---|---|
| Population: women undergoing caesarean section Comparison: no drape | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | No of participants | Certainty of the evidence | Comments | |
| Risk with no drape | Risk with drape | |||||
| Surgical site infection | Study population | RR 1.29 | 1373 | ⊕⊕⊝⊝ | ||
| 112 per 1000 | 144 per 1000 | |||||
| Metritis | 49 per 1000 | 79 per 1000 (14 to 447) | RR 1.62 (0.29 to 9.16) | 79 | ⊕⊝⊝⊝ VERY LOW a,c | |
| Length of stay (days) | The mean length of stay with no drape was 5.7 days | The mean number of days with a drape was 0.10 higher | ‐ | 603 | ⊕⊕⊕⊝ | |
| Adverse events (maternal or neonatal) | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in any of the included studies. |
| *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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a Selection bias was unclear and blinding of outcome assessor was at high risk of bias (risk of bias ‐1). b Wide 95% CI (imprecision ‐1). c Single study with small sample size and wide 95% CI (imprecision ‐2). | ||||||
Background
Description of the condition
Caesarean section is an increasingly common major surgical procedure performed on women (WHO 2015). For example, in 2015, every third birth (32%) in the USA was a caesarean delivery (Martin 2015). The increasing rate of caesarean birth worldwide in both high‐ and low‐income countries, is well established, and a concern to many (Thomas 2001). Between 2000 and 2015, the global rate of caesarean section almost doubled from 12.1% to 21.1% (Boerma 2018). The risk of maternal morbidity and mortality is higher in caesarean section birth than in vaginal birth; postoperative infection is a common component of morbidity. With the increase in caesarean sections, it is important that the risks to the mother are minimised as far as possible (Thomas 2001). This review focused on different agents, methods and forms of application for preoperative skin preparation to prevent infection; it did not include studies of preoperative handwashing of the surgical team and preoperative bathing.
Women who give birth by caesarean section are exposed to both endogenous (internal) and exogenous (external) sources of infection during birth. Exposure to a hospital environment places these women at risk of developing hospital‐acquired infections. The rate of postcaesarean infection has been estimated to be 10 times greater than that after vaginal birth (Henderson 1995).
The Centers for Disease Control and Prevention (CDC) estimates that 27 million surgical procedures are performed in the USA each year. The CDC's National Nosocomial Infections Surveillance system reports that surgical site infections are the third most frequently reported nosocomial infection, accounting for 14% to 16% of all such infections (CDC 2005). Preventing infection by properly preparing the skin before incision is thus a vital part of the overall care given to women during caesarean birth.
The incidence of abdominal incisional infections following caesarean section ranges from 3% to 15%. A postcaesarean surgical site infection is a bacterial infection in the surgical incision following an abdominal birth. Women who develop a postcaesarean surgical site infection typically experience a temperature of 38.0°C (100.4°F) or higher, and lower abdominal pain (Cunningham 2018). Abdominal incisional abscesses that develop following caesarean birth usually cause fever on about the fourth postoperative day. In many cases, these are preceded by uterine infection, and fever persists from the first or second postoperative day. Wound redness (erythema) and drainage may also be present. Organisms causing these infections are usually the same as those isolated from amniotic fluid at caesarean birth, but hospital‐acquired pathogens may also be the cause (Lewis 2013).
Some women are more likely than others to develop a postcaesarean surgical site infection. Women at increased risk include those who are obese; have diabetes or an immunosuppressive disorder (HIV infection); have chorioamnionitis (infection of the amniotic fluid and fetal membrane) during labour; anaemia; or are taking corticosteroids (by mouth or intravenously (Cunningham 2018)).
In addition to surgical site infections, another common source of morbidity is postcaesarean metritis, including endometritis, an infection that develops within the lining of the uterus after birth. Despite the use of routine antibiotics before or during surgery (perioperative prophylaxis), estimates of metritis following caesarean range form 10% to 20% (Normand 2001).
Description of the intervention
Proper preparation of an incision site involves removing surface dirt and oil with a soap or detergent scrub plus applying a topical antimicrobial agent that will reduce the bacterial population to a minimal level. In surgical patients, the choice of surgical scrub and the duration of scrubbing have not been shown to make any significant difference in the rate of surgical site infection in either clean or clean‐contaminated wounds (such as caesarean skin incision (Dumville 2015; Mangram 1999)).The use of plastic adhesive drape is one of the common method of preventing surgical site infection.
Antiseptics to prevent infection have been in use for over 150 years. Antiseptic handwash solution was first introduced by Semmelweis, in 1847, at the Vienna Maternity Hospital, to reduce maternal mortality due to puerperal sepsis (Loudon 2002). Later, in 1864, Lister introduced carbolic acid spray preparation for the operative site. Since then, many solutions (including alcoholic iodine, mercuric compounds, and ether) have been used to prepare the operative site. However, as another Cochrane Review has shown, there is uncertainty about which antiseptic skin preparation is the most effective for preventing postoperative surgical site infections (Dumville 2015). Iodophore (on its own or as an alcohol‐containing agent) and chlorhexidine gluconate are the primary skin disinfectants used; the CDC's Prevention Guideline for the Prevention of Surgical Site Infection, published in 2017, recommends the use of alcohol‐containing preparations unless there is a contraindication to alcohol (CDC 2017). Iodophore is effective against bacteria, fungi, viruses and spore‐forming bacteria, and its disinfecting effect lasts for a long time. However, it cannot be used on mucous membranes and does not have an immediate antiseptic effect.Chlorhexidine gluconate is characterised by its immediate antiseptic effect, although it cannot kill the spores (Johansson 2007).
There are six types of antiseptics that are designed for topical application: iodine or iodophors, alcohol, chlorhexidine gluconate, hexachlorophene, parachlorometaxylenol, and triclosan (Dumville 2015; Larson 1988). For the purpose of this review, antiseptic agents can be applied in the form of liquids, solutions, or powders, or delivered on impregnated drapes.
How the intervention might work
The removal of transient bacteria and reduction of the number of existing organisms by antiseptic is recommended prior to surgery by several organisations, for example, the Royal College of Surgeons of England (Leaper 2001), the Centers for Disease Control and Prevention (Mangram 1999), and the Association of Operating Room Nurses (AORN 2002). Several antiseptic agents are available for preoperative preparation of skin at the incision site. Leclair 1990 described an antiseptic as 'a chemical agent that reduces the microbial population on the skin'. It is suggested that the ideal agent would kill all bacteria, fungi, viruses, protozoa, tubercle bacilli, and spores; be nontoxic, hypoallergenic, and safe to use in all body regions; not be absorbed; have residual activity, and be safe for repetitive use (Dumville 2015; Hardin 1997).
Antiseptics for preoperative skin preparation should be broad‐spectrum and fast‐acting, and contain an antimicrobial ingredient that results in significant reduction in the number of micro‐organisms on intact skin (Larson 1988). The primary action of antiseptics includes both the mechanical removal and chemical killing, and the inhibition of both contaminating and colonising flora.
Why it is important to do this review
There is a Cochrane Review on the use of preoperative skin antiseptics for preventing infections (Dumville 2015). However, although the scope of the Dumville 2015 review included clean and clean‐contaminated surgical operations, including caesarean section, the focus was solely on preventing surgical site infections. In our review, the focus extended to preventing all types of infection, such as endometritis or metritis. In this review, we did not look at different methods of surgical incision for caesarean, because that was the topic of another Cochrane Review (Mathai 2013).
Objectives
To compare the effects of different antiseptic agents, different methods of application, or different forms of antiseptic used for preoperative skin preparation for preventing postcaesarean infection.
Methods
Criteria for considering studies for this review
Types of studies
All published and unpublished randomised controlled trials (RCTs) and quasi‐RCTs, including cluster‐RCTs, evaluating any described type of preoperative skin preparation agents, methods and forms of application for caesarean section were eligible for inclusion. No quasi‐ or cluster‐RCTs were identified for inclusion. Cross‐over studies were not eligible for inclusion. We included studies presented only as abstracts, if they provided enough information.
Types of participants
Pregnant women undergoing elective or emergency caesarean section.
Types of interventions
Comparisons between different antiseptic agents used for caesarean section skin preparation (e.g. alcohol, povidone iodine), different methods of antiseptic application (e.g. scrub, paint, drape), or different forms of antiseptic (e.g. powder, liquid).
We only included studies involving the preparation of the incision area. We excluded studies of preoperative handwashing of the surgical team and preoperative bathing. Other Cochrane Reviews cover other methods for preventing infection at caesarean section (e.g. antimicrobial application, skin shaving).
A related Cochrane Review covers vaginal preparation with antiseptics before caesarean section (Haas 2020).
Types of outcome measures
Primary outcomes
-
Surgical site infection (as defined by trialists)
-
Metritis or endometritis, or both
Secondary outcomes
-
Length of stay
-
Maternal mortality
-
Repeat surgery
-
Re‐admission resulting from infection
-
Reduction of skin bacteria colony count*
-
Adverse events (maternal or neonatal)*
*Outcome not prespecified in our published protocol (Hadiati 2008); see Differences between protocol and review.
Search methods for identification of studies
The following search methods section of this review is based on a standard template used by Cochrane Pregnancy and Childbirth.
Electronic searches
For this update, we searched Cochrane Pregnancy and Childbirth’s Trials Register by contacting their Information Specialist (9 July 2019)
The Register is a database containing over 25,000 reports of controlled trials in the field of pregnancy and childbirth. It represents over 30 years of searching. For full current search methods used to populate Pregnancy and Childbirth’s Trials Register including the detailed search strategies for CENTRAL, MEDLINE, Embase and CINAHL; the list of handsearched journals and conference proceedings, and the list of journals reviewed via the current awareness service, please follow this link.
Briefly, Cochrane Pregnancy and Childbirth’s Trials Register is maintained by their Information Specialist and contains trials identified from:
-
monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL);
-
weekly searches of MEDLINE (Ovid);
-
weekly searches of Embase (Ovid);
-
monthly searches of CINAHL (EBSCO);
-
handsearches of 30 journals and the proceedings of major conferences;
-
weekly current awareness alerts for a further 44 journals plus monthly BioMed Central email alerts.
Search results are screened by two people and the full text of all relevant trial reports identified through the searching activities described above is reviewed. Based on the intervention described, each trial report is assigned a number that corresponds to a specific Pregnancy and Childbirth review topic (or topics), and is then added to the Register. The Information Specialist searches the Register for each review using this topic number rather than keywords. This results in a more specific search set that has been fully accounted for in the relevant review sections (Included studies; Excluded studies; Ongoing studies).
In addition, we searched ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP) for unpublished, planned, and ongoing trial reports (9 July 2019) using the methods detailed in Appendix 1.
Searching other resources
We searched the reference lists of all included studies and review articles.
We did not apply any language restrictions.
Data collection and analysis
For methods used in the previous version, please see Hadiati 2012.
For this update, we used the following methods, which are based on a standard template used by the Cochrane Pregnancy and Childbirth.
Selection of studies
Three review authors independently assessed all the potential studies we identified as a result of the search strategy, for inclusion. We resolved any disagreement through discussion, or consulted a fourth person if required.
Data extraction and management
We designed a form to extract data. We also extracted information on the dates of the study, sources of trial funding, and the trial authors' declarations of interest. For eligible studies, at least two review authors extracted the data, using the agreed form. We resolved discrepancies through discussion, or consulted a third review author. if required. We entered data into Review Manager 5 software, and checked for accuracy (RevMan 2014).
Assessment of risk of bias in included studies
Two review authors independently assessed the risk of bias for each study, using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We resolved any disagreement by discussion with the third review author.
(1) Random sequence generation (checking for possible selection bias)
For each included study, we described the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups.
We assessed the method as:
-
low risk of bias (any truly random process, e.g. random number table; computer random number generator);
-
high risk of bias (any non‐random process, e.g. odd or even date of birth; hospital or clinic record number);
-
unclear risk of bias.
(2) Allocation concealment (checking for possible selection bias)
For each included study, we described the method used to conceal the allocation sequence and determine whether intervention allocation could have been foreseen in advance of, or during recruitment, or changed after assignment.
We assessed the methods as:
-
low risk of bias (e.g. telephone or central randomisation; consecutively numbered sealed opaque envelopes);
-
high risk of bias (open random allocation; unsealed or non‐opaque envelopes, alternation; date of birth);
-
unclear.
(3.1) Blinding of participants and personnel (checking for possible performance bias)
For each included study, we described the method used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. We considered that studies were at low risk of bias if they were blinded, or if we judged that the lack of blinding could not have affected the results. We assessed blinding separately for different outcomes or classes of outcomes.
We assessed the methods as:
-
low, high, or unclear risk of bias for participants;
-
low, high, or unclear risk of bias for personnel.
(3.2) Blinding of outcome assessment (checking for possible detection bias)
For each included study, we described the method used, if any, to blind outcome assessors from knowledge of which intervention a participant received. We assessed blinding separately for different outcomes or classes of outcomes.
We assessed methods used to blind outcome assessment as:
-
low, high, or unclear risk of bias.
(4) Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations)
For each included study, and for each outcome or class of outcomes, we described the completeness of data, including attrition and exclusions from the analysis. We stated whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomised participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups, or were related to outcomes. Where sufficient information was reported, we re‐included missing data in the analyses that we undertook. We assessed methods as:
-
low risk of bias (e.g. no missing outcome data; missing outcome data balanced across groups);
-
high risk of bias (e.g. numbers or reasons for missing data imbalanced across groups; 'as treated' analysis done with substantial departure of intervention received from that assigned at randomisation);
-
unclear risk of bias.
(5) Selective reporting bias
For each included study, we described how we investigated the possibility of selective outcome reporting bias, and what we found.
We assessed the methods as:
-
low risk of bias (where it was clear that all of the study’s prespecified outcomes and all expected outcomes of interest to the review were reported);
-
high risk of bias (where not all the study’s prespecified outcomes were reported; one or more reported primary outcomes were not prespecified; outcomes of interest were reported incompletely and so could not be used; study failed to include results of a key outcome that would have been expected to have been reported);
-
unclear risk of bias.
(6) Other sources of bias
For each included study, we reported any important concerns we had about other possible sources of bias.
We assessed whether each study was free of other problems that could put it at risk of bias:
-
low risk of bias;
-
high risk of bias;
-
unclear whether there is risk of other bias.
(7) Overall risk of bias
We made explicit judgements about whether studies were at high risk of bias, according to the criteria given in the CochraneHandbook for Systematic Reviews of Interventions (Higgins 2011). With reference to (1) to (6) above, we assessed the likely magnitude and direction of the bias, and whether we considered it was likely to impact on the findings. We explored the impact of the level of bias by undertaking sensitivity analyses (see Sensitivity analysis).
Measures of treatment effect
Dichotomous data
For dichotomous data, we presented results as summary risk ratio with 95% confidence intervals.
Continuous data
For continuous outcomes, we used the mean difference with 95% confidence intervals, as outcomes were measured in the same way between trials. In future updates of this review, if there are trials that measure the same outcome, but use different methods, we will use the standardised mean difference.
Unit of analysis issues
Cluster‐randomised trials
We did not identify any cluster‐randomised trials for inclusion in our review. In future updates, if we identify any cluster‐randomised trials, we will include them in the analyses along with individually‐randomised trials. We will consider it reasonable to combine the results from both if there is little heterogeneity between the study designs and the interaction between the effect of intervention and the choice of randomisation unit is considered to be unlikely. We will adjust their sample sizes using the methods described in the Cochrane Handbook for Systematic Reviews of Interventions, using an estimate of the intracluster correlation co‐efficient (ICC) derived from the trial (if possible), from a similar trial, or from a study of a similar population (Higgins 2011). If we use ICCs from other sources, we will report this, and conduct sensitivity analyses to investigate the effect of variation in the ICC.
We will also acknowledge heterogeneity in the randomisation unit, and perform a sensitivity analysis to investigate the effects of the randomisation unit.
Cross‐over trials
Cross‐over trials were not eligible for inclusion in this review.
Studies with more than two intervention arms
If studies included multiple intervention groups (Cordtz 1989; Ngai 2015), we included only the arms relevant to our research question and each arm was included in the analysis only once (Higgins 2011).
Dealing with missing data
For included studies, we noted levels of attrition. We explored the impact of including studies with high levels of missing data in the overall assessment of treatment effect with a sensitivity analysis.
For all outcomes, we carried out analyses, as far as possible, on an intention‐to‐treat basis, i.e. we included all participants randomised to each group in the analyses, and all participants were analysed in the group to which they were allocated, regardless of whether or not they received the allocated intervention. The denominator for each outcome in each trial was the number randomised minus any participants whose outcomes were known to be missing.
Assessment of heterogeneity
We assessed statistical heterogeneity in each meta‐analysis using the Tau², I², and Chi² statistics. We regarded heterogeneity as substantial if the I² was greater than 30%, and either the Tau² was greater than zero, or there was a low P value (less than 0.10) in the Chi² test for heterogeneity.
Assessment of reporting biases
In future updates, if there are 10 or more studies in the meta‐analysis, we will investigate reporting biases (such as publication bias) using funnel plots. We will assess funnel plot asymmetry visually. If asymmetry is suggested by a visual assessment, we will perform exploratory analyses to investigate it (Sterne 2017).
Data synthesis
We carried out statistical analysis using Review Manager 5 software (RevMan 2014). We used fixed‐effect meta‐analysis for combining data where it was reasonable to assume that studies were estimating the same underlying treatment effect: i.e. where trials were examining the same intervention, and the trials’ populations and methods were judged to be sufficiently similar. Where there was sufficient clinical heterogeneity to expect that the underlying treatment effects differ between trials, or if we detect substantial statistical heterogeneity, we used a random‐effects meta‐analysis to produce an overall summary, where we considered an average treatment effect across trials to be clinically meaningful. We treated the random‐effects summary as the average of the range of possible treatment effects, and discuss the clinical implications of treatment effects differing between trials. If the average treatment effect was not considered to be clinically meaningful, we would not have combined the trials.
In future updates, if we use random‐effects analyses, we will present the results as the average treatment effect with its 95% confidence interval, and the estimates of Tau² and I².
Subgroup analysis and investigation of heterogeneity
We planned to investigate substantial heterogeneity by using subgroup analyses and sensitivity analyses, and if an overall summary is meaningful, use random‐effects analysis to produce it. The outcome with substantial heterogeneity in this update, did not include enough trials to make performing these analyses meaningful. Instead, we used a random‐effects analysis for this outcome, and in future updates, we will carry out the following subgroup analyses.
-
Risk of infection (high versus low risk)
-
Duration of skin preparation
-
Dose of preparation
We will restrict subgroup analyses to the primary outcomes.
We will assess differences between subgroups by using interaction tests available in RevMan 2014. We will report the results of subgroup analyses quoting the Chi² statistic and P value, and the interaction test I² value.
To assess the effect of the addition of alcohol to povidone iodine, we performed a post hoc subgroup analysis comparing 'chlorhexidine plus alcohol versus povidone iodine plus alcohol' versus 'chlorhexidine plus alcohol versus povidone iodine alone', see Analysis 2.1.
Sensitivity analysis
We carried out a sensitivity analysis to explore the effects of blinded outcome assessment on the results of the review. Studies in which blinded outcome assessment (for surgical site infection) was inadequate (i.e. high risk of bias) were excluded from the analysis to assess for any substantive difference to the overall result.
In future updates, we will also carry out sensitivity analysis to explore the effect of allocation concealment on the results of the review, and exclude studies with are high risk for this domain. We will also carry out a sensitivity analysis to explore the effects of fixed‐effect or random‐effects analyses for outcomes with statistical heterogeneity. If, in future updates, we include cluster‐randomised trials along with the individually‐randomised trials, we will carry out sensitivity analysis to investigate the effect of the randomisation unit.
Summary of findings and assessment of the certainty of the evidence
We used the GRADE approach to evaluate the certainty of the evidence, as outlined in the GRADE Handbook (GRADE Handbook; GRADE Working Group 2004). The GRADE approach uses five considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of the body of evidence for specific outcomes. The evidence can be downgraded from 'high certainty' by one level for serious (or by two levels for very serious) limitations, depending on assessments for risk of bias, indirectness of evidence, serious inconsistency, imprecision of effect estimates, or potential publication bias (GRADE Working Group 2004; Schünemann 2009). In this review, we used the GRADE approach to assess the following outcomes for all comparisons in the review.
-
Surgical site infection
-
Metritis or endometritis, or both
-
Length of stay*
-
Adverse events (maternal or neonatal)*
*Outcome for GRADE assessment was not prespecified in our previous review (Hadiati 2014); see Differences between protocol and review.
We used GRADEpro GDT to import data from Review Manager 5 to create a 'Summary of findings' table, which presents a summary of the intervention effect and a measure of certainty according to the GRADE approach for each of the outcomes listed above (GRADEpro GDT; RevMan 2014).
Results
Description of studies
Results of the search
See: Figure 1.
Study flow diagram
For this update, we assessed six new trial reports and reassessed the four studies that were still ongoing in the previous version of the review. We included two new studies (three reports) and added an additional report to an already included study. We also excluded four studies and two are ongoing.
Included studies
We included 13 trials with a total of 6938 women. SeeCharacteristics of included studies.
Method, trial dates, and sample sizes
All of the included studies were randomised controlled trials (RCTs) (Aworinde 2016; Cordtz 1989; Fahmi 2017; Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Pello 1990; Saha 2019; Salama 2016; Springel 2017; Tuuli 2016; Ward 2001). We did not include any quasi‐ or cluster‐RCTs. The trials were conducted between 1983 and 2016. The trial dates were not provided in three studies (Fahmi 2017; Magann 1993; Pello 1990). Sample sizes ranged from 22 women (Pello 1990), to 1404 women (Ngai 2015).
Settings
Six studies were conducted in the USA (Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Springel 2017; Tuuli 2016), one study in Denmark (Cordtz 1989), one in France (Pello 1990), one in Nigeria (Aworinde 2016), one in South Africa (Ward 2001), one in Indonesia (Fahmi 2017), one in India (Saha 2019), and one study was conducted in Egypt (Salama 2016). Of the included trials, 10 were single‐centre trials (Aworinde 2016; Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Saha 2019; Salama 2016; Springel 2017; Tuuli 2016; Ward 2001), two were conducted in multiple centres (Cordtz 1989; Fahmi 2017), and one trial did not provide any details about the facility (Pello 1990).
Participants
Six of the trials recruited women with either scheduled or emergency caesarean section (Fahmi 2017; Magann 1993; Saha 2019; Salama 2016; Springel 2017; Tuuli 2016). Magann 1993 did not include women undergoing caesarean section for fetal distress with inadequate time for skin preparation. Three trials only recruited women with scheduled caesarean section (Aworinde 2016; Kunkle 2015; Ngai 2015). The remaining four trials did not specify type of caesarean section (Cordtz 1989; Lorenz 1988; Pello 1990; Ward 2001). Pello 1990 included women with a male fetus, who were undergoing caesarean section. Kunkle 2015 recruited women aged 18 to 45 years undergoing scheduled caesarean delivery at 36 gestational weeks or greater. Participants in Ngai 2015 were all women in their 37th week of gestation, who were undergoing scheduled or non‐emergency caesarean delivery. Most trials stated that they excluded women with a known sensitivity or allergy to one of the antiseptics used.
Interventions and comparisons
Different antiseptic agents/preparations
Two different antiseptic preparation comparisons were made in the 12 trials. The Magann 1993 trial compared a five‐minute scrub with parachlorometaxylenol followed by a 10% povidone‐iodine scrub and normal saline irrigation of the pelvis and subcutaneous tissue at uterine closure and fascial closure (special preparation) in the experimental group, versus a 7.5% povidone‐iodine surgical scrub followed by 10% povidone iodine and normal saline irrigation of the pelvis and subcutaneous tissue at uterine closure and fascial closure (standard preparation) in the control group. Eight trials compared preoperative application of chlorhexidine gluconate versus povidone iodine, without the use of a drape in either the intervention or control arms (Aworinde 2016; Fahmi 2017; Kunkle 2015; Ngai 2015; Saha 2019; Salama 2016;Springel 2017; Tuuli 2016).Of these eight trials, three trials (Fahmi 2017; Saha 2019; Springel 2017) compared 2% chlorhexidine versus 10% povidone iodine. One trial used antiseptics including 70% isopropyl alcohol in both groups (Fahmi 2017), and the other two trials used antiseptics including 70% isopropyl alcohol in the chlorhexidine group only (Saha 2019; Springel 2017).
Different methods
Four trials compared the use of drape versus no drape, and the drapes used were impregnated with an antiseptic agent (Lorenz 1988; Pello 1990) or without an antiseptic agent (Cordtz 1989; Ward 2001). Cordtz 1989 assessed the effect of incisional plastic drapes versus no drape, combined with standard iodine disinfection with 2.5% iodine in 70% ethanol for all women. For the purpose of this review, we did not include data from this trial relating to additional arms in which women were re‐disinfected, because this secondary disinfection of the skin around the incision took place shortly before skin closure, not before skin incision. Ward 2001 compared plastic incisional drape with no drape. Before surgery in all women, the abdomen was washed with chlorhexidine soap, and then swabbed with 0.5% chlorhexidine in 80% alcohol solution for 30 seconds. In the Lorenz 1988 trial, the comparison of skin preparation for caesarean section was between a one‐minute scrub with 70% isopropyl alcohol followed by application of iodophor‐impregnated adhesive film in the experimental group, versus a five‐minute iodophor scrub followed by application of iodophor solution in the control group. The Pello 1990 trial compared two skin preparations using different agents and either drape or no drape (chlorhexidine 0.5% versus 70% alcohol plus an IOBAN 2 drape).
Outcomes
Surgical site infection (as defined by trialists)
Twelve studies reported on our primary outcome of surgical site infection (Aworinde 2016; Cordtz 1989; Fahmi 2017; Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Saha 2019; Salama 2016; Springel 2017; Tuuli 2016; Ward 2001). Surgical site infection was assessed from delivery to 30 days postoperative in Aworinde 2016, Springel 2017, Tuuli 2016 and Saha 2019; at three and seven days postoperative in Fahmi 2017; at three days and two weeks in Kunkle 2015; on two separate measurements at least 24 hours postoperative in Lorenz 1988; on two separate occasions six hours apart after the first 24 hours in Magann 1993; at two and six weeks postcaesarean in Ngai 2015; and at seven days and 30 days in Salama 2016. Cordtz 1989 did not describe the time of assessment of surgical site infection. Trials used various definitions for surgical site infections. Cordtz 1989 defined a possible wound infection as localised erythema, serous secretion, or both without the presence of pus, and infected as the presence of pus regardless of the results of bacteriological examination. Kunkle 2015 defined surgical site infection as the presence of purulent drainage, cellulitis, or the need for incision and drainage, or treatment with antibiotics for a clinical diagnosis of infection. In Lorenz 1988, surgical site infection was defined as infectious morbidity with (1) erythema and tenderness of the wound or separation of skin edges, and (2) no uterine tenderness or malodorous, discoloured lochia. Magann 1993 defined this outcome as hyperemic skin incision and a fluctuant mass, which when opened, contained purulent material. In Ward 2001, surgical site infection was diagnosed if two of three features were present: (1) erythematous cellulitis, (2) seropurulent discharge from the wound, and (3) positive swab culture (organisms and leucocytes). In Salama 2016, surgical site infection was diagnosis by pain, tenderness, swelling, redness, heat, purulent discharge from the incision or deliberate reopening of the surgical wound. Fahmi 2017, Ngai 2015, Springel 2017, Tuuli 2016 and Saha 2019 based the diagnosis of surgical site infection on Centers for Disease Control and Prevention (CDC) criteria. One trial did not define surgical site infection in the report (Aworinde 2016).
Metritis or endometritis, or both
Five trials reported on endometritis (Lorenz 1988; Magann 1993; Salama 2016; Springel 2017; Tuuli 2016). In Lorenz 1988, metritis was defined as infectious morbidity with either (1) uterine tenderness on bimanual examination, or (2) no other site of infection identified on physical examination, urinalysis, or urine culture, and if indicated, a chest roentgenogram. In Magann 1993, endometritis was identified through physical examination, urine, and blood cultures. Springel 2017 followed the definition of endometritis according to the CDC, and Tuuli 2016 and Salama 2016 did not provide any definition for this outcome.
Length of stay
Tuuli 2016 and Ward 2001 reported our secondary outcome, length of stay. However, Tuuli 2016 reported only median length of hospital stay, so we only described data in a narrative way.
Re‐admission resulting from infection
Three trials reported on re‐admission resulting from infection (Salama 2016; Springel 2017; Tuuli 2016).
Reduction of skin bacteria colony count
Lorenz 1988 reported on reduction of skin bacteria colony counts, and Kunkle 2015 reported on bacterial growth at 18 hours. Salama 2016 reported on positive post‐sterilisation skin cultures.
Adverse events (maternal or neonatal)
Three trials reported on maternal adverse events involving skin reactions, with two (Aworinde 2016; Tuuli 2016) reporting on erythema and skin irritation, and one (Salama 2016) reporting allergic reaction. Only one trial reported on neonatal adverse events, which was the concentration of iodine in the cord blood (Pello 1990). No other maternal or neonatal adverse events were reported.
The other outcomes investigated by this review, maternal mortality and repeated surgery, were not been reported in any of the included trials.
Sources of trial funding
Seven trials did not mention source of funding (Cordtz 1989; Fahmi 2017; Lorenz 1988; Magann 1993; Ngai 2015; Pello 1990; Ward 2001). The Obafemi Awowolowo University Teaching Hospital Complex, Nigeria supported Aworinde 2016. The University of Southern California supported Kunkle 2015 and Edward Henry Springel, MD supported Springel 2017. Tuuli 2016 received funding from the Women’s Reproductive Health Research Career Development grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health, and the Department of Obstetrics and Gynecology, Washington University School of Medicine in St. Louis. The primary investigator (Dr Luwang) was sponsor in the trial by Saha 2019. The Ain Shams Maternity Hospital in Cairo, Egypt, was sponsor in the trial by Salama 2016.
Trialists' declarations of interest
Nine studies did not report declarations of interest (Aworinde 2016; Cordtz 1989; Fahmi 2017; Lorenz 1988; Magann 1993; Ngai 2015; Pello 1990; Saha 2019; Ward 2001). Three studies stated no conflict of interest (Kunkle 2015; Salama 2016; Springel 2017). In Tuuli 2016, the first author reported having received a grant from the National Institutes of Health during the conduct of the study, but all other authors declared no conflict of interest.
Excluded studies
We excluded 10 trials (Bianco 2018; Brown 1984; Jindal 2019; Kosus 2010; Lukabwe 2018; NCT00528008; NCT01700803; NCT02027324; Nili 2015; Robins 2005). Bianco 2018 was excluded as it focused on preoperative washing on the surgical site by the surgical team. The Brown 1984 study was excluded as the trial did not present the results for caesarean section cases separately from other surgical cases. Jindal 2019 was excluded as they compared different antiseptic agents and different surgical closing techniques. The Kosus 2010 trial was excluded as they used an antibiotic compared with or without Rifamycin SV (250 mg) before closure of subcutaneous tissue, in addition to povidone iodine 10% for preoperative antisepsis and after closure of the skin. Lukabwe 2018 was excluded as they focused on preoperative bathing. One trial (NCT00528008) was stopped following interim analysis and results of interim analysis are unclear. Two trials (NCT01700803; NCT02027324) were excluded as they focused on preoperative handwashing of the surgical team. We excluded one study because it was a cohort study and not a randomised controlled trial (Nili 2015). Robins 2005 compared the effectiveness of chlorhexidine spray and single‐use sachets for skin preparation before regional nerve blockade, not for the preparation of the incision site before caesarean section.
For further details of the excluded studies, seeCharacteristics of excluded studies.
Risk of bias in included studies
See 'Risk of bias' tables and figures for the 13 included studies in Characteristics of included studies, Figure 2, and Figure 3.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
'Risk of bias' graph: review authors' judgements about each risk of bias item for each included study
Allocation
Sequence generation
Sequence generation (computer‐generated sequence) was adequate in nine trials (Aworinde 2016; Fahmi 2017; Lorenz 1988; Magann 1993; Ngai 2015; Saha 2019; Salama 2016; Tuuli 2016; Ward 2001). The remaining four trials provided insufficient or no details about sequence generation, and therefore, we judged them to be of unclear risk of bias (Cordtz 1989; Kunkle 2015; Pello 1990; Springel 2017).
Allocation concealment
The majority of included trials provided insufficient or no details regarding the measures taken to ensure that the treatment allocation could not be foreseen, and so we assessed them as unclear risk of bias for this item (Cordtz 1989; Fahmi 2017; Kunkle 2015; Lorenz 1988; Magann 1993; Pello 1990; Springel 2017; Tuuli 2016). In four trials, allocation concealment was performed using sequentially numbered sealed envelopes, so we assessed the trials at low risk of bias (Aworinde 2016; Ngai 2015; Salama 2016; Ward 2001). Saha 2019 used an open list of random numbers and was therefore judged to be of high risk of bias for allocation concealment.
Blinding
Blinding of participants and personnel (performance bias)
Although blinding was not reported in most of the trials, it was unlikely that this lack of blinding could have caused different treatments or behaviour between groups, such that outcomes would be affected, and therefore, we judged performance bias to be low in 11 trials (Aworinde 2016; Cordtz 1989; Fahmi 2017; Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Salama 2016; Springel 2017; Tuuli 2016; Ward 2001). In Saha 2019 and Pello 1990, the information provided was too limited to exclude any performance bias.
Blinding of outcome assessment (detection bias)
Seven trials were at high risk of detection bias (Aworinde 2016; Cordtz 1989; Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Springel 2017). In these trials, outcome assessors were not blinded, and some outcomes involved an important element of subjective assessment. It was possible that lack of blinding could have caused differences in the reported outcomes, especially with respect to our primary outcome of surgical site infection, and therefore, this brings into question the certainty of the estimated treatment effect. Outcome assessment was blinded in four trials, so we judged them at low risk of bias (Fahmi 2017; Salama 2016; Tuuli 2016; Ward 2001). In Pello 1990, blinding of outcome assessment was not described, but lack of blinding was unlikely to affect the outcome because it was objectively measured, and therefore was at low risk of bias. Saha 2019 did not provide sufficient information to exclude detection bias and was judged to be unclear.
Incomplete outcome data
In 11 trials, we judged attrition bias to be low, as all women were followed up until the end of the study, or the number of missing participants was balanced between intervention groups, and it was unlikely that the missing data substantially influenced the outcomes (Aworinde 2016; Cordtz 1989; Fahmi 2017; Kunkle 2015; Lorenz 1988; Magann 1993; Ngai 2015; Salama 2016; Springel 2017; Tuuli 2016; Ward 2001). Two studies were judged as unclear risk of bias, because the number of women in each group was not reported (Pello 1990; Saha 2019).
Selective reporting
We found no evidence of selective reporting in four trials, and assessed them at low risk of bias (Aworinde 2016; Kunkle 2015; Ngai 2015; Springel 2017). Protocols were not available for us to assess originally planned outcomes in six trials, and so these we assessed as unclear risk of reporting bias (Cordtz 1989; Fahmi 2017; Lorenz 1988; Magann 1993; Pello 1990; Ward 2001). Saha 2019 was also assessed as unclear because only an abstract of this study was available which may not report all the outcomes assessed. We judged Salama 2016 and Tuuli 2016 as being at a high risk of reporting bias, because some prespecified outcomes were not reported in the trial report.
Other potential sources of bias
Overall, we judged 11 trials to be at low risk of bias for other potential sources of bias. In the Cordtz 1989 trial, antibiotics for prophylaxis therapy were given to about 10% of the women in the study. Even though more women in the drape group (10.7%) received antibiotics than in the group without drapes (8.2%), infections were more common in the drape group. Therefore, the 'real' effect of skin preparation may have been different, but this was likely to be small, because of the overall small difference between groups. In the Lorenz 1988, Magann 1993, and Kunkle 2015 trials, the use of prophylaxis antibiotics was not specified in the study report. Saha 2019 and Pello 1990 did not provide enough information to exclude other bias.
Effects of interventions
See: Summary of findings 1 Parachlorometaxylenol with iodine versus iodine alone for preventing infection following caesarean section; Summary of findings 2 Chlorhexidine gluconate compared to povidone iodine for preventing infection following caesarean section; Summary of findings 3 Drape compared to no drape for preventing infection following caesarean section
We included 13 trials, involving 6938 women, in this review. One small study, involving 22 women, only reported neonatal outcomes, and did not contribute any data towards any of the prespecified outcomes of this review (Pello 1990). We were only able to conduct meta‐analyses for Comparisons 2 (8 trials) and 3 (3 trials).
Comparison 1: Parachlorometaxylenol with iodine versus iodine alone
One trial, involving 50 women, contributed data to this comparison (Magann 1993). See summary of findings Table 1.
Primary outcomes
Surgical site infection
We were uncertain if parachlorometaxylenol with iodine made any clear difference to the reduction of surgical site infection when compared with iodine alone, because the certainty of the evidence was very low (risk ratio (RR) 0.33, 95% confidence interval (CI) 0.04 to 2.99; 1 trial, 50 women; very low‐certainty evidence; Analysis 1.1).
Metritis or endometritis, or both
We were uncertain if parachlorometaxylenol with iodine made any difference to the reduction of metritis, endometritis, or both, when compared with iodine alone, because the certainty of the evidence was very low (RR 0.88, 95% CI 0.56 to 1.38; 1 trial, 50 women; very low‐certainty evidence; Analysis 1.2).
Secondary outcomes
Length of stay
No data were reported for length of stay.
Maternal mortality
No data were reported for maternal mortality.
Repeat surgery
No data were reported for repeated surgery.
Re‐admission resulting from infection
No data were reported for re‐admission resulting from infection.
Reduction of skin bacteria colony count
No data were reported for reduction of skin bacteria colony count.
Adverse events (maternal or neonatal)
No adverse events were reported for mother or baby in the included study.
Comparison 2: Chlorhexidine gluconate verus povidone iodine
Eight trials, involving 4807 women, contributed data to this comparison (Aworinde 2016; Fahmi 2017; Kunkle 2015; Ngai 2015; Saha 2019; Salama 2016; Springel 2017; Tuuli 2016). See summary of findings Table 2.
Primary outcomes
Surgical site infection
Moderate certainty of evidence suggested that chlorhexidine gluconate before caesarean section probably slightly reduces the incidence of surgical site infection compared with povidone iodine (RR 0.72, 95% CI 0.58 to 0.91; 8 trials, 4323 women; moderate‐certainty evidence; Analysis 2.1). The effect of chlorhexidine gluconate compared with povidone iodine was still present in sensitivity analysis (RR 0.87, 95% CI 0.62 to 1.23; P = 0.44; 4 trials, 2037 women) after removing four trials at high risk of bias for outcome assessment (Aworinde 2016; Kunkle 2015; Ngai 2015; Springel 2017). To assess the effect of the addition of alcohol to povidone iodine we performed a post hoc subgroup analysis comparing 'chlorhexidine plus alcohol versus povidone iodine plus alcohol' versus 'chlorhexidine plus alcohol versus povidone iodine alone'. We found no evidence of a subgroup difference according to the test for subgroup differences: Chi² = 0.25, df = 1 (P = 0.61), I² = 0%. However, it should be noted that there were too few trials included to carry out any meaningful subgroup analysis.
Metritis or endometritis, or both
Low certainty of evidence indicated that using chlorhexidine gluconate before caesarean section, when compared with povidone iodine, may make little or no difference to the reduction of endometritis (RR 0.95, 95% CI 0.49 to 1.86; 3 trials, 2484 women; low certainty of evidence; Analysis 2.2).
Secondary outcomes
Length of stay
In Tuuli 2016, a median length of hospital stay of four days (interquartile range three to four days) was reported for both the chlorhexidine gluconate group and the povidone iodine group; there was no difference in these reported data between groups.
Maternal mortality
No data were reported for maternal mortality in any of the included studies.
Repeat surgery
No data were reported for repeated surgery in any of the included studies.
Re‐admission resulting from infection
The results from three trials reporting on this outcome suggested that there may be little or no difference in re‐admissions due to infection between the chlorhexidine gluconate and the povidone‐iodine groups (average RR 0.51, 95% CI 0.25 to 1.02; 3 trials, 2484 women; Analysis 2.3).
Reduction of skin bacteria colony count
Results from one small trial suggested that women who received chlorhexidine gluconate skin preparation may have slightly reduced bacterial growth at 18 hours after caesarean section compared with women who received skin preparation with povidone iodine (RR 0.23, 95% CI 0.07 to 0.70; 1 trial, 60 women; Analysis 2.4).
Adverse events (maternal or neonatal)
We are unclear about the effects for skin irritation or allergic skin reaction (RR 0.64, 95% CI 0.28 to 1.46; 3 trials, 1926 women; very low certainty of evidence; Analysis 2.5), because the evidence was very‐low certainty. Wide confidence intervals crossing the line of no effect were also present for skin reactions (RR 0.79, 95% CI 0.32 to 1.96; 1 trial, 374 women; Analysis 2.5), and erythema (RR 1.13, 95% CI 0.57 to 2.26; 2 trials, 1521 women; Analysis 2.5). No neonatal adverse events were reported in the included studies.
Comparison 3: Drape versus no drape
Four trials involving 2046 women contributed data to this comparison (Cordtz 1989; Ward 2001; Lorenz 1988; Pello 1990). See summary of findings Table 3
Primary outcomes
Surgical site infection
For women undergoing caesarean section, low certainty of the evidence suggested that using a drape before surgery compared with no drape may make little or no difference to surgical site infection (risk ratio (RR) 1.29, 95% confidence interval (CI) 0.97 to 1.71; 3 trials, 1373 women; low‐certainty evidence; Analysis 3.1).
Metritis or endometritis, or both
One trial, involving 79 women, was measured (Lorenz 1988). We were uncertain whether the combination of a one‐minute alcohol scrub with an iodophor drape clearly reduced the occurrence of metritis when compared with a five‐minute iodophor scrub, because the certainty of the evidence was very low (RR 1.62, 95% CI 0.29 to 9.16; 1 trial, 79 women; very low‐certainty evidence; Analysis 3.2).
Secondary outcomes
Length of stay
The length of stay was measured in one trial involving 603 women, which compared the use of a drape with no drape (Ward 2001). Moderate‐certainty evidence suggested that the use of a drape probably makes little or no difference to the length of time spent in hospital (mean difference (MD) 0.10 day, 95% CI ‐0.27 to 0.46; Analysis 3.3).
Maternal mortality
No data were reported for maternal mortality in either of the included studies.
Repeat surgery
No data were reported for repeated surgery in either of the included studies.
Re‐admission resulting from infection
No data were reported for re‐admission resulting from infection in either of the included studies.
Reduction of skin bacteria colony count
The single trial (Lorenz 1988) providing data on this comparison suggested that there may be little or no difference between groups for reduced skin bacteria colony counts (MD 0.07 colony forming unit per plate, 95% CI ‐0.34 to 0.48; 1 trial, 79 women; Analysis 3.4).
Adverse events (maternal or neonatal)
Pello 1990 compared skin preparation with chlorhexidine 0.5% and 70% alcohol with a drape (IOBAN 2), and reported the newborn's exposure to iodine. Cord blood iodine concentration (CBI) was higher in the IOBAN 2 drape group (18.38 ± 20.34 versus 6.44 ± 0.66 μg/100 mL; P < 0.05). There was no clear difference between the two groups in 48‐hour urine iodine excretion or thyroid‐stimulating hormone (TSH) blood concentration on the fifth day. No maternal adverse events were reported in the included study.
Discussion
Summary of main results
This updated review included 13 trials and 6938 women, with 12 trials and 6916 women contributing data to the meta‐analyses. In relation to the primary outcomes of metritis or endometritis, our analyses suggested that there may be little or no difference between different skin preparations for caesarean section. In one comparison, we found that there is probably a slight reduction in surgical site infection with chlorhexidine gluconate compared to povidone iodine, based on seven trials including 4323 women. We assessed the evidence for this result to be of moderate certainty due to limitations in study design. The effect of chlorhexidine gluconate compared with povidone iodine was still present in sensitivity analysis after removing four trials at high risk of bias for outcome assessment (Aworinde 2016; Kunkle 2015; Ngai 2015; Springel 2017). The evidence for the other comparisons between skin preparations in this review all came from only one or two trials, involved small numbers of women, and also yielded effect estimates with wide confidence intervals in most cases. There were also wide‐ranging concerns regarding risk of bias. In some studies, it was not possible to know whether randomisation processes were adequate, and we were particularly concerned about the high number of studies where there was no blinding of subjective assessments relating to infection. Moreover, in assessing surgical site infection, pooled results may have also been only indirectly comparable, due to variability between the included studies in the criteria used to assess surgical site infection.
Regarding the secondary outcomes, we were unclear about the effect between chlorhexidine gluconate and povidone iodine for maternal skin irritation or allergic skin reaction because the certainty of the evidence was found to be very low. However, the included trials provided either very little or no data for almost all of our specified secondary outcomes, most notably, adverse neonatal effects, maternal mortality, or repeat surgery. Therefore, this review was limited in its power to detect meaningful differences between antiseptic agents and methods of skin preparation, with respect to most of the outcomes under consideration.
Overall completeness and applicability of evidence
Of the 13 included trials, seven were reasonably large, while the other six trials each recruited only small numbers of women. Twelve trials contributed to the evaluation of four main comparisons, limiting the ability to pool results (we were only able to carry out meta‐analysis for two comparisons). Twelve trials reported on surgical site infection, while five trials reported on endometritis; only one to two trials reported on the remaining outcomes, or not at all. Four trials (Lorenz 1988; Magann 1993; Salama 2016; Springel 2017) used a mix of the co‐intervention with scrubbing; we were unable to assess interventions separately. Therefore, it was not possible to fully address the objectives of this review.
We are unable to draw robust conclusions regarding the different agents and methods of skin preparation for preventing infection following caesarean section. The body of evidence was too small; although we included 13 studies involving 6938 women, the evidence available for each comparison and outcome under investigation reported on too few women and was too limited in scope.
Quality of the evidence
The included studies were subject to some important methodological limitations. Regarding the primary outcome of surgical site infection, most of the studies did not blind outcome assessors, and we could not always be confident in the adequacy of randomisation processes. When we assessed the certainty of the evidence using GRADE criteria, we found it ranged from very low to moderate. We found very‐low certainty of evidence when comparing parachlorometaxylenol with iodine versus iodine alone, due to wide confidence intervals crossing the line of no effect and small sample sizes, and lack of blinding of outcome assessment for surgical site infection and endometritis (summary of findings Table 1). We found moderate certainty of evidence when comparing chlorhexidine gluconate with povidone iodine, for surgical site infection, low certainty of evidence for endometritis, and very low certainty of evidence for adverse events (maternal ‐ skin irritation or allergic skin reaction), due to wide 95% confidence intervals and lack of blinding for outcome assessment (summary of findings Table 2). When comparing the use of a drape versus no drape, we found low‐certainty evidence for surgical site infection, very‐low certainty evidence for metritis and moderate‐certainty evidence for length of stay, with downgrading due to wide confidence intervals crossing the line of no effect, small sample sizes, and lack of blinding of outcome assessment (summary of findings Table 3).
Potential biases in the review process
We believe it is unlikely that we missed any eligible published studies, because of the comprehensive search strategy employed by Cochrane Pregnancy and Childbirth. However, there may be relevant unpublished trials that we were unable to locate. Screening of the studies, data extraction, 'Risk of bias' and GRADE assessments were carried out independently by at least two review authors, therefore ensuring reliable data were available for the review. Professor Hadiati is a named author on Fahmi 2017, but was not involved in the screening process and'Rrisk of bias' assessment.
Agreements and disagreements with other studies or reviews
The Cochrane Review of skin preparation for clean surgery found preoperative skin preparation with 0.5% chlorhexidine in methylated spirits was associated with lower rates of surgical site infections following clean surgery than alcohol‐based povidone iodine, which is consistent with our findings (Dumville 2015). Rather than iodine alone, alternate chlorhexidine might be effective, although the evidence for both comparisons was low certainty of evidence. We need to interpret the results with caution. However, caesarean section is regarded as 'clean‐contaminated' surgery, and so the effect of antiseptic skin preparation may be more evident for this type of surgery because of exposure to both internal and external sources of infection during birth. Because of the limited certainty of the evidence presently available, we were unable to explore this hypothesis further in this review.
Study flow diagram
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
'Risk of bias' graph: review authors' judgements about each risk of bias item for each included study
Comparison 1: Parachlorometaxylenol with iodine versus iodine alone, Outcome 1: Surgical site infection
Comparison 1: Parachlorometaxylenol with iodine versus iodine alone, Outcome 2: Endometritis
Comparison 2: Chlorhexidine gluconate versus povidone iodine, Outcome 1: Surgical site infection
Comparison 2: Chlorhexidine gluconate versus povidone iodine, Outcome 2: Endometritis
Comparison 2: Chlorhexidine gluconate versus povidone iodine, Outcome 3: Re‐admission resulting from infection
Comparison 2: Chlorhexidine gluconate versus povidone iodine, Outcome 4: Bacterial growth 18 hours
Comparison 2: Chlorhexidine gluconate versus povidone iodine, Outcome 5: Adverse events (maternal)
Comparison 3: Drape versus no drape, Outcome 1: Surgical site infection
Comparison 3: Drape versus no drape, Outcome 2: Metritis
Comparison 3: Drape versus no drape, Outcome 3: Length of stay
Comparison 3: Drape versus no drape, Outcome 4: Reduction of skin bacteria colony counts
| Parachlorometaxylenol with iodine versus iodine alone | ||||||
|---|---|---|---|---|---|---|
| Population: women undergoing caesarean section Comparison: iodine alone | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of participants | Certainty of the evidence | Comments | |
| Risk with iodine alone | Risk with parachlorometaxylenol with iodine | |||||
| Surgical site infection | Study population | RR 0.33 | 50 | ⊕⊝⊝⊝ | ||
| 120 per 1000 | 40 per 1000 | |||||
| Endometritis | Study population | RR 0.88 | 50 | ⊕⊝⊝⊝ | ||
| 640 per 1000 | 563 per 1000 | |||||
| Length of stay | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in the included study. |
| Adverse events (maternal or neonatal) | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in the included study. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a Wide confidence interval crossing the line of no effect, single study with small sample size (imprecision ‐2). b Blinding of outcome assessor was at high risk of bias (risk of bias ‐1). | ||||||
| Chlorhexidine gluconate compared to povidone iodine | ||||||
|---|---|---|---|---|---|---|
| Population: women undergoing caesarean section Comparison: povidone iodine | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | No of participants | Certainty of the evidence | Comments | |
| Risk with povidone iodine | Risk with chlorhexidine gluconate | |||||
| Surgical site infection | Study population | RR 0.72 | 4323 | ⊕⊕⊕⊝ | ||
| 75 per 1000 | 54 per 1000 | |||||
| Endometritis | Study population | RR 0.95 | 2484 | ⊕⊕⊝⊝ | ||
| 14 per 1000 | 13 per 1000 | |||||
| Length of stay | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in any of the included studies. |
| Adverse events (maternal) ‐ skin irritation or allergic skin reaction | Study population | RR 0.64 | 1926 | ⊕⊝⊝⊝ | No neonatal adverse events were reported in any of the included studies. | |
| 15 per 1000 | 9 per 1000 | |||||
| *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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a Selection bias was unclear and blinding of outcome assessor was high risk of bias (risk of bias ‐1). b Wide confidence interval crossing the line of no effect (imprecision ‐1). c Wide confidence interval crossing the line of no effect and few events (imprecision ‐2). | ||||||
| Drape compared to no drape | ||||||
|---|---|---|---|---|---|---|
| Population: women undergoing caesarean section Comparison: no drape | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | No of participants | Certainty of the evidence | Comments | |
| Risk with no drape | Risk with drape | |||||
| Surgical site infection | Study population | RR 1.29 | 1373 | ⊕⊕⊝⊝ | ||
| 112 per 1000 | 144 per 1000 | |||||
| Metritis | 49 per 1000 | 79 per 1000 (14 to 447) | RR 1.62 (0.29 to 9.16) | 79 | ⊕⊝⊝⊝ VERY LOW a,c | |
| Length of stay (days) | The mean length of stay with no drape was 5.7 days | The mean number of days with a drape was 0.10 higher | ‐ | 603 | ⊕⊕⊕⊝ | |
| Adverse events (maternal or neonatal) | ‐ | ‐ | ‐ | ‐ | ‐ | This outcome was not reported in any of the included studies. |
| *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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a Selection bias was unclear and blinding of outcome assessor was at high risk of bias (risk of bias ‐1). b Wide 95% CI (imprecision ‐1). c Single study with small sample size and wide 95% CI (imprecision ‐2). | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Surgical site infection Show forest plot | 1 | 50 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.33 [0.04, 2.99] |
| 1.2 Endometritis Show forest plot | 1 | 50 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.88 [0.56, 1.38] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Surgical site infection Show forest plot | 8 | 4323 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.72 [0.58, 0.91] |
| 2.1.1 Chlorhexidine plus alcohol versus povidone iodine plus alcohol | 4 | 2663 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.62 [0.45, 0.87] |
| 2.1.2 Chlorhexidine plus alcohol versus povidone iodine | 4 | 1660 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.84 [0.61, 1.15] |
| 2.2 Endometritis Show forest plot | 3 | 2484 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.95 [0.49, 1.86] |
| 2.3 Re‐admission resulting from infection Show forest plot | 3 | 2484 | Risk Ratio (M‐H, Random, 95% CI) | 0.51 [0.25, 1.02] |
| 2.4 Bacterial growth 18 hours Show forest plot | 1 | 60 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.23 [0.07, 0.70] |
| 2.5 Adverse events (maternal) Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2.5.1 Any skin reaction | 1 | 374 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.79 [0.32, 1.96] |
| 2.5.2 Erythema | 2 | 1521 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.13 [0.57, 2.26] |
| 2.5.3 Skin irritation or allergic skin reaction | 3 | 1926 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.64 [0.28, 1.46] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 3.1 Surgical site infection Show forest plot | 3 | 1373 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.29 [0.97, 1.71] |
| 3.1.1 Iodine | 1 | 691 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.42 [0.98, 2.04] |
| 3.1.2 Chlorhexidine | 1 | 603 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.11 [0.70, 1.76] |
| 3.1.3 Isopropyl alcohol scrub versus iodophor scrub | 1 | 79 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
| 3.2 Metritis Show forest plot | 1 | 79 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.62 [0.29, 9.16] |
| 3.3 Length of stay Show forest plot | 1 | 603 | Mean Difference (IV, Fixed, 95% CI) | 0.10 [‐0.27, 0.46] |
| 3.4 Reduction of skin bacteria colony counts Show forest plot | 1 | 79 | Mean Difference (IV, Fixed, 95% CI) | 0.07 [‐0.34, 0.48] |
