Prophylactic intravenous calcium therapy for exchange blood transfusion in the newborn

Tinuade A Ogunlesi; Foluso EA Lesi; Olabisi Oduwole

doi:10.1002/14651858.CD011048.pub2

Tratamiento profiláctico con calcio intravenoso para la exanguinotransfusión en el recién nacido

Declaraciones de intereses de los autores

Versión publicada: 12 octubre 2017 Historial de versiones

https://doi.org/10.1002/14651858.CD011048.pub2

Contraer todo Desplegar todo

Resumen

disponible en

Antecedentes

La exanguinotransfusión es una forma de transfusión de sangre total mediante la cual todo el volumen sanguíneo se reemplaza en unas pocas horas. En la medicina perinatal y neonatal, la exanguinotransfusión se utiliza con mayor frecuencia en el tratamiento de la anemia grave o la hiperbilirrubinemia grave en la primera semana de vida. Se considera que la hipocalcemia, una de las morbilidades frecuentes asociadas con la exanguinotransfusión, se produce debido a los efectos quelantes del citrato utilizado habitualmente como un anticoagulante en la sangre del donante. Este trastorno se caracteriza por irritabilidad muscular y nerviosa, así como con arritmias cardíacas.

Objetivos

Determinar si la administración profiláctica de calcio reduce el riesgo de morbilidades relacionadas con la hipocalcemia y muerte entre los recién nacidos sometidos a una exanguinotransfusión.

Métodos de búsqueda

Se utilizó la estrategia de búsqueda estándar del Grupo Cochrane de Neonatología para buscar en el Registro Cochrane Central de Ensayos Controlados (CENTRAL 2016, número 5), MEDLINE vía PubMed (1966 hasta 29 junio 2016), Embase (1980 hasta 29 junio 2016) y en CINAHL (1982 hasta 29 junio 2016). También se buscaron ensayos controlados aleatorios y ensayos cuasialeatorios en bases de datos de ensayos clínicos, actas de congresos y listas de referencias de los artículos recuperados.

Criterios de selección

Todos los ensayos aleatorios y cuasialeatorios de tratamiento profiláctico con calcio intravenoso para la exanguinotransfusión en los recién nacidos.

Obtención y análisis de los datos

Dos autores de la revisión de forma independiente evaluaron y extrajeron los datos sobre los métodos, los participantes, las intervenciones y los resultados (media del calcio sérico total e ionizado antes y después de la exanguinotransfusión y la presencia de eventos adversos como hipoglucemia, apnea, paro cardíaco y muerte inmediatamente después de la exanguinotransfusión). Los resultados se informaron como diferencia de medias (DM) con intervalos de confianza (IC) del 95% para los resultados continuos y como cociente de riesgos (CR) y diferencia de riesgos (DR) con los IC del 95% para los resultados dicotómicos. Se evaluó la calidad utilizando la herramienta de evaluación Cochrane 'Riesgo de sesgo' y el sistema GRADE.

Resultados principales

Sólo se encontró un ensayo cuasialeatorio con 30 participantes que cumplió con los criterios de inclusión. En este ensayo pequeño los niveles de calcio sérico total e ionizado se midieron inmediatamente antes e después de la exanguinotransfusión. Todos los participantes se incluyeron en el análisis final y se informaron todos los resultados importantes.

Medidas de resultado principales

Hubo una muerte en cada grupo (CR 1,00; IC del 95%: 0,07 a 14,55; DR 0,00; IC del 95%: ‐0,18 a 0,18; participantes = 30; estudios = 1). El estudio no informó la presencia de arritmias cardíacas en la semana posterior a la exanguinotransfusión ni el número de lactantes con niveles de calcio sérico (total menor de 8 mg/dl [2 mmol/l] o ionizado menor de 4,4 mg/dl [1,1 mmol/l]).

La comparación pareada de exanguinotransfusión con gluconato de calcio al 10% intravenoso versus exanguinotransfusión sin calcio intravenoso (cambio a partir del valor inicial) mostró que el calcio sérico total medio se elevó en el grupo de intervención en comparación con el grupo control (DM ‐0,46; IC del 95%: ‐0,81 a ‐0,11; participantes = 30; estudios = 1). La evidencia de muy baja calidad también indicó un aumento de los niveles de calcio sérico ionizado medio en el grupo de intervención en comparación con el grupo control (DM ‐0,22; IC del 95%: ‐0,33 a ‐0,11; participantes = 30; estudios = 1).

Medidas de resultado secundarias

Las reacciones adversas al tratamiento con calcio intravenoso incluyeron el paro cardíaco en un lactante del brazo de intervención (CR 3,00; IC del 95%: 0,13 a 68,26; DR 0,07; IC del 95%: ‐0,10 a 0,23; participantes = 30; estudios = 1). En los dos lactantes que murieron ocurrió apnea e hipoglucemia (CR 1,00; IC del 95%: 0,07 a 14,55; DR 0,00; IC del 95%: ‐0,18 a 0,18; participantes = 30; estudios = 1). No hubo datos disponibles sobre otros resultados secundarios importantes como el número de lactantes con reducción del magnesio sérico, reducción de la hormona paratiroidea, aumento de la calcitonina, presencia de crisis convulsivas, espasmo carpopedio, agitación e intervalo QTc prolongado en la electrocardiografía en el transcurso de una semana desde la exanguinotransfusión.

Conclusiones de los autores

Datos de muy baja calidad de un ensayo controlado cuasialeatorio indicaron que la media del calcio sérico total e ionizado aumentó en el grupo de estudio pero disminuyó en el grupo control inmediatamente después de la exanguinotransfusión. Sin embargo, los valores medios del calcio total e ionizado en ambos brazos de estudio permanecieron dentro de los rangos de referencia internacionales. Desafortunadamente no hubo datos disponibles para evaluar la tendencia del calcio sérico total e ionizado al final de la primera semana después de la exanguinotransfusión. Por lo tanto, debido a la calidad muy baja de la evidencia disponible, es difícil apoyar o rechazar el uso continuo del calcio intravenoso profiláctico en los recién nacidos sometidos a una exanguinotransfusión. Se alienta a los investigadores a realizar más ensayos con un diseño consistente y con un gran número de participantes y, en particular, que analicen el patrón de diferencias según la edad gestacional de los participantes, el tipo de anticoagulante administrado y el volumen de sangre utilizado.

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

Uso de calcio durante la exanguinotransfusión en el recién nacido

Pregunta de la revisión: ¿la administración de calcio durante la exanguinotransfusión reduce el riesgo de bajos niveles de calcio en sangre y el riesgo de muerte por bajos niveles de calcio en sangre entre los recién nacidos sometidos a una exanguinotransfusión?

Antecedentes: la exanguinotransfusión es un tipo de transfusión de sangre masiva (donde se toma sangre de una persona sana y se introduce en la sangre de un paciente enfermo) que predispone a la reducción de los niveles de calcio en sangre como resultado del efecto de unión (quelante) de los productos químicos anticoagulantes utilizados para preservar la sangre del donante. Los bajos niveles de calcio en sangre puede alterar la función de los músculos (incluido el corazón) y los nervios. En esta revisión, se investigaron los efectos beneficiosos del calcio administrado sobre los niveles de calcio en sangre y la seguridad de la administración de calcio a los recién nacidos sometidos a una exanguinotransfusión.

En las bases de datos médicas se efectuaron búsquedas de ensayos clínicos que administraran o no calcio a recién nacidos con edades de hasta 28 días de vida a partir del nacimiento a los que se les realizó una exanguinotransfusión y que monitorizaron una variedad de síntomas que incluyeron la muerte y los problemas del corazón.

Características de los estudios:se identificó solo un estudio mal diseñado para su inclusión.

Resultados: se encontró que administrar calcio a los recién nacidos elevó los niveles de calcio en sangre inmediatamente después de la exanguinotransfusión, mientras que los niveles de calcio en sangre se redujeron entre los recién nacidos que no recibieron calcio durante la exanguinotransfusión. Sin embargo, a pesar de las tendencias observadas en los valores de los niveles de calcio en sangre en ambos grupos del estudio, los valores promedios de calcio en sangre permanecieron dentro de límites normales. Además, la información disponible para esta revisión fue limitada, ya que los efectos de administrar o no el calcio durante la exanguinotransfusión no se estudiaron más allá del momento en que se realizó.

Conclusión: sobre la base de los resultados de esta revisión, no existe evidencia de buena calidad para apoyar ni rechazar el uso continuo del calcio durante la exanguinotransfusión. No es posible establecer conclusiones ya que la evidencia encontrada es limitada y de calidad muy baja y podría cambiar si se dispone de más resultados de estudios más grandes y mejor diseñados.

Authors' conclusions

Implications for practice

Prophylactic administration of intravenous calcium prevents the reduction in the total and ionised serum calcium in neonates receiving exchange blood transfusion (EBT) immediately after the procedure. This has the potential of preventing clinically significant EBT‐related hypocalcaemia but it is uncertain if this likely benefit is of any clinical significance or if the perceived benefit would be retained and justifiable hours and days after the procedure. Therefore, with the high risk of bias in the available trial and the accrued very low quality of evidence in the index review, there is, at present, not enough evidence to support or reject prophylactic administration of intravenous calcium to neonates receiving EBT.

Implications for research

This review shows the very low‐quality evidence on potential benefits/harms of prophylactic administration of intravenous calcium to neonates receiving EBT. This observation shows a significant gap in evidence‐based knowledge and lack of reliable data to inform clinical practice for a common clinical intervention in perinatal and neonatal medicine, particularly in the parts of the world where the procedure is still frequently carried out. Well‐designed and appropriately powered randomised controlled trials are required to investigate the clinical utility or otherwise of prophylactic administration of intravenous calcium to neonates receiving EBT to prevent hypocalcaemia‐related morbidities and mortality. These studies should address the trend of total and ionised serum calcium during and after EBT, the hormonal changes which accompany the procedure, as well as the pattern of cardiac response when calcium is administered to neonates during the procedure. In addition, it is important to build into the studies the likely effects of other morbidities associated with EBT and how these effects may vary with or without calcium administration. Specific attention should be paid to the maturity of the neonates as determined by the estimated gestational age, the type of anticoagulant used to preserve donor blood, and the volume of blood transfused. While double‐volume EBT is no longer frequently performed in the technologically advanced parts of the world due to more efficient management of neonatal hyperbilirubinaemia, single‐volume EBT is still widely performed in the management of severe neonatal anaemia the world over. Therefore, these studies should aim at findings from multicentre research across low‐ to high‐income countries as the application of the findings would, most likely, be global.

Summary of findings

Open in table viewer

Summary of findings for the main comparison. Pair‐wise comparison of exchange blood transfusion (EBT) plus 10% calcium gluconate intravenously versus EBT alone (change from baseline) for exchange blood transfusion in the newborn

Pair‐wise comparison of EBT plus 10% calcium gluconate intravenously versus EBT alone (change from baseline) for EBT in the newborn
Patient or population: neonates receiving EBT Settings: Intervention: EBT + 10% calcium gluconate IV versus EBT alone (change from baseline)
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	EBT	EBT + 10% calcium gluconate IV
All‐cause death Number of events	Study population		See comment	30 (1 study)	⊕⊕⊝⊝ Low^a	Risks were calculated from pooled risk differences.
	67 per 1000	67 per 1000 (‐113 to 247)
	Moderate
	67 per 1000	67 per 1000 (‐114 to 248)
Mean total serum calcium levels Mean difference	‐	The mean total serum calcium levels in the intervention group was 0.46 lower (0.81 to 0.11 lower)	‐	30 (1 study)	⊕⊝⊝⊝ Very low^b,c,d	‐
Mean ionised serum calcium levels Mean ionised serum calcium	‐	The mean ionised serum calcium levels in the intervention group was 0.22 lower (0.33 to 0.11 lower)	‐	30 (1 study)	⊕⊝⊝⊝ Very low^b,c,d	‐
Adverse reaction: cardiac arrest Number of events	Study population		See comment	30 (1 study)	⊕⊝⊝⊝ Very low^b,c,d	Risks were calculated from pooled risk differences.
	0 per 1000	0 per 1000 (0 to 0)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Adverse reaction: apnoea and hypoglycaemia Number of events	Study population		See comment	30 (1 study)	⊕⊝⊝⊝ Very low^b,c,d	Risks were calculated from pooled risk differences.
	67 per 1000	67 per 1000 (‐113 to 247)
	Moderate
	67 per 1000	67 per 1000 (‐114 to 248)
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). CI: confidence interval; EBT: exchange blood transfusion; IV: intravenous; RR: risk ratio.
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
^aQuasi‐randomised controlled study with no blinding; we are very uncertain if death was due to the effect of intervention or EBT. ^bQuasi‐randomised controlled study with no blinding; we are very uncertain about the estimate of effect. ^cOne small study with very few participants. ^dQuasi‐randomised control trial, baseline characteristics not reported and no information on the length of follow‐up.

Background

Exchange blood transfusions (EBT) have been reported as accounting for close to two‐thirds of blood transfusions among newborns in low‐ and middle‐income countries (Pam 2004; Ogunlesi 2011). On the contrary, centres in high‐income countries rarely employ EBT in the care of newborn infants (MacLennan 2001; Flaherman 2012). This difference is due to the varying pattern of indications for EBT (such as haemolytic diseases and other causes of hyperbilirubinaemia (increased levels of bilirubin in the blood) in the newborn infants) in low‐ and middle‐income countries compared to high‐income countries.

Description of the condition

EBT is a form of whole blood transfusion in which the total blood volume is replaced within a period of 24 hours. In perinatal and neonatal medicine, EBT is most often used for rapid removal of toxins (such as bilirubin and sensitised red cells), or to replace red blood cells. Two types of EBT are commonly used in newborns; single‐volume or double‐volume exchange transfusion. Single‐volume EBT implies transfusion with 80 mL/kg to 85 mL/kg of blood, which is the estimated total blood volume of the infant, whereas double‐volume EBT implies transfusion with double the estimated total blood volume of an infant (160 mL/kg to 170 mL/kg) (Maisels 2001). Traditionally, double‐volume EBT is used in the management of severe hyperbilirubinaemia, while single‐volume EBT is commonly used in the management of severe neonatal anaemia, especially within the first week of life. Double‐volume EBT is used in infants with severe hyperbilirubinaemia arising from haemolytic (blood) diseases such as blood group incompatibilities (particularly ABO and Rhesus), enzymopathies (such as glucose‐6‐phosphate dehydrogenase deficiency), and membranopathies (causing red blood cell anomalies such as erythrocytosis and elliptocytosis). Other indications include septicaemia, prematurity, cephalohematoma or subgaleal haematoma (collections of blood in the scalp), disseminated intravascular coagulation (clotting anomaly), and metabolic derangements such as intractable hypoglycaemia (low blood sugar levels).

Description of the intervention

Fresh whole blood, usually preserved with acid citrate dextrose or citrate phosphate dextrose, is used for EBT. The calculated volume of blood is administered in small aliquots of 5 mL to 10 mL, depending on the weight of the infant, using a three‐way valve. The valve is connected to an umbilical venous catheter through which the transfusion is performed over about two hours. When done effectively, the double‐volume procedure replaces between 85% and 90% of the red cells in circulation, and reduces the serum bilirubin level to about 60% of the pre‐EBT level, while the single‐volume procedure raises the haematocrit (level of red blood cells) of the infant close to that of the blood used for the procedure (Maisels 2001). Complications during EBT may arise from the process of umbilical catheterisation, the biochemical properties of the anticoagulant used, and the presence of microbes in the blood.

EBT‐related morbidities include hypocalcaemia (low blood calcium), hypothermia (low body temperature), hypoglycaemia, metabolic acidosis (overproduction of acid), hyperkalaemia (high blood potassium), thrombocytopenia (low platelet concentration), cardiac arrhythmias (irregularities in heart beats), air embolism, apnoea (cessation of breathing), septicaemia (infection of the blood), omphalitis (infections of the umbilical stump), necrotising enterocolitis (tissue death in the bowel), and intestinal perforation (Patra 2004; Sanpavat 2005; Gharehbaghi 2010; Hosseinpour Sakha 2010).

Hypocalcaemia (defined as serum total calcium less than 8 mg/dL (2 mmol/L) or serum ionised calcium less than 4.4 mg/dL (1.1 mmol/L)) is routinely prevented with calcium administration during or immediately after EBT (Nelson 1989; Maisels 2001). However, studies have shown that not all newborn infants with EBT‐related hypocalcaemia get calcium treatment (Steiner 2007). Hypocalcaemia may be asymptomatic, but some of the common manifestations include jitteriness, abnormal cry, lethargy, irritability, apnoea, hypotonia (low muscle tone), carpopedal spasm (affecting hands and feet), or seizures (Maisels 2001). These clinical features do not reflect the severity of calcium deficiency, and may occur in any combination. Evidence of a prolonged QTc interval on electrocardiography is suggestive of severe hypocalcaemia requiring treatment. The use of clinical assessment for hypocalcaemia may be justifiable in low‐ and middle‐income countries where laboratory facilities required for the detection of hypocalcaemia may not be available.

This intervention may be administered in pulses during the procedure, or as a slow bolus (dose) immediately after the procedure.

As a result of the risk of significant morbidity and mortality associated with hypocalcaemia, some practitioners ‐ and even standard textbooks of neonatal medicine ‐ suggest routine prophylactic administration of calcium to infants undergoing EBT as described above. Although, some practitioners still advocate routine prophylactic calcium administration for EBT, most current guidelines on transfusion practices are silent on the need for routine calcium administration to infants requiring EBT or stipulate that routine administration of calcium is not recommended (NICE 2010). It has yet to be confirmed that this practice is necessary, safe, and effective.

How the intervention might work

Hypocalcaemia that occurs during EBT is thought to arise from the chelating (metal binding) effects of citrate commonly used as an anticoagulant in transfused blood. Serum calcium levels have been shown to fall significantly during EBT without accompanying clinical manifestations, and to rise transiently to hypercalcaemic (high calcium level) ranges following calcium administration (Nelson 1989; Lo 1990). The two extremes of this fluctuation pose great risk of cardiac arrhythmias to the infant undergoing EBT. However, there have been suggestions that changes in ionised serum calcium level during EBT are not remarkable (Kreuger 1975; Jasso‐Gutierrez 1982). Furthermore, serum calcium levels have been reported to normalise within 24 hours of EBT, even without calcium therapy (Nelson 1989).

The interaction between the chelating agent and serum calcium is complex. When blood containing citrate is administered, the citrate binds calcium and forms unionised calcium complexes. Citrate causes a fall in ionised serum calcium levels resulting in a compensatory stimulation of parathormone secretion and mobilisation of calcium and phosphorus in order to normalise the serum calcium level. However, use of heparin as an anticoagulant for blood preserves the serum ionised calcium levels, in addition to those of albumin and parathyroid hormone, during EBT (Milner 1975). As citrate is more commonly used as an anticoagulant than heparin, newborn infants receiving EBT are expected to develop hypocalcaemia. Therefore, newborn infants receiving EBT with citrated blood should benefit from routinely given prophylactic intravenous calcium to prevent EBT‐related hypocalcaemia.

Potential adverse effects

The adverse effects of intravenous calcium include vomiting, constipation, rapid vasodilation with hypotension (widening of blood vessels causing low blood pressure), bradycardia (slow heart rate), cardiac dysrhythmias (irregular heartbeat), and possibly cardiac arrest in severe cases. It also causes intense irritation at the site of administration and ultimately, tissue necrosis if extravasation (leakage from the vein) occurs.

Why it is important to do this review

Infants undergoing EBT may develop extremely low serum levels of total and ionised calcium. Some infants undergoing EBT may develop hypocalcaemia and require treatment during EBT, while others may not require treatment with calcium (Steiner 2007). Fluctuations in serum calcium levels may cause dangerously high serum calcium levels with the risk of cardiac toxicity.

It is important to know the extent to which hypocalcaemia, which requires treatment, occurs during EBT, and if it is necessary and safe to administer prophylactic intravenous calcium routinely to infants requiring EBT.

Objectives

To determine whether the use of prophylactic calcium reduces the risk of hypocalcaemia‐related morbidities and death among newborn infants receiving EBT.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials and quasi‐randomised controlled trials (individually or cluster‐randomised) that compared the outcome of newborn infants undergoing EBT who received or did not receive prophylactic calcium.

Types of participants

Newborn infants aged from birth to 28 days who required EBT (single‐ or double‐volume) for any reason and irrespective of postmenstrual age (whether or not the infant was premature).

Types of interventions

We considered for inclusion infants who received prophylactic intravenous calcium irrespective of the dose and type of calcium salt administered and the time of calcium administration (during or within one hour of EBT). The control arm included infants who did not receive intravenous calcium or received placebo.

Types of outcome measures

Primary outcomes

All‐cause death until discharge.
Presence of cardiac dysrhythmias within one week of EBT.
Number of infants with total serum calcium levels less than 8 mg/dL (2 mmol/L) or ionised calcium levels less than 4.4 mg/dL (1.1 mmol/L) within one week of EBT.

Secondary outcomes

Number of infants with serum magnesium under 1.5 mg/dL within one week of EBT.
Number of infants with serum parathormone under 10 pg/mL within one week of EBT.
Number of infants with serum calcitonin over 10 pg/mL within one week of EBT.
Duration of hospitalisation (in days).
Incidence of partial or generalised seizures within one week of EBT (clinically diagnosed as tonic extension of limbs or clonic, jerky movements of the limbs, or diagnosed by electroencephalography.
Incidence of carpopedal spasms within one week of EBT (clinically diagnosed as sustained contraction of the muscles of the extremities).
Incidence of jitteriness within one week of EBT (clinically diagnosed as brief myoclonic twitching of the muscles of the extremities occurring in clusters).
Incidence of prolonged QTc interval on electrocardiography within one week of EBT.
Adverse reactions to calcium therapy within one week of EBT (skin necrosis, vomiting, constipation, hypotension, apnoea, cardiac arrest) occurring singly or in combination.
1. Skin necrosis defined as an obvious discontinuity in the skin layer at the site of calcium administration.
2. Vomiting defined as a reflex act of expelling the gastric contents.
3. Constipation defined as a frequency of bowel opening that was less than three times per week.
4. Hypotension defined as mean arterial pressure less than 30 mmHg.
5. Apnoea defined as cessation of breathing for more than 20 seconds associated with cyanosis (blue tinge to skin) and bradycardia.
6. Cardiac arrest defined as a sudden cessation of cardiac output and effective circulation.

Outcomes included post hoc:

Mean total serum calcium levels
Mean ionised serum calcium levels

Search methods for identification of studies

Electronic searches

We used the criteria and standard methods of Cochrane and the Cochrane Neonatal Review Group (see the Cochrane Neonatal Group search strategy for specialized register).

We conducted a comprehensive search including: Cochrane Central Register of Controlled Trials (CENTRAL 2016, Issue 5) in the Cochrane Library; MEDLINE via PubMed (1966 to 29 June 2016); Embase(1980 to 29 June 2016); and CINAHL (1982 to 29 June 2016) using the following search terms: (transfusion OR exchange transfusion OR exchange blood transfusion OR blood transfusion) AND (calcium)), plus database‐specific limiters for RCTs and neonates (see Appendix 1 for the full search strategies for each database). We applied no language restrictions.

We searched clinical trials registries for ongoing or recently completed trials (clinicaltrials.gov; the World Health Organization International Trials Registry and Platform www.whoint/ictrp/search/en/, and the ISRCTN Registry).

Searching other resources

We handsearched journals; published abstracts; and proceedings of major conferences such as the International Neonatology Conference, conferences of the Royal College of Paediatrics and Child Health, and the American Academy of Pediatrics. We searched the reference lists of articles obtained. We searched the Web of Science for relevant citations in journal databases and in the conference proceedings citation index.

Data collection and analysis

Selection of studies

Two review authors (TAO and FEAL) independently assessed the eligibility of trials identified through searching for inclusion in the review. We retrieved full reports of the potentially relevant trials, and independently determined if they met the inclusion criteria using a pretested eligibility form. We listed all excluded studies, along with the reasons for excluding them (see Characteristics of excluded studies table). We would have ensured that trials with multiple publications were included only once, and where the multiple publications included different but relevant outcomes, we would have included all the publications in the review. We would have grouped multiple publications so that each study, rather than each publication, was the unit of analysis. We resolved any disagreements through discussion.

Data extraction and management

Two review authors (TAO and OO) independently extracted data from the eligible study using the pretested data extraction form. One review author (OO) entered the data into Review Manager 5 (RevMan 2014), while the second review author (TAO) cross‐checked the data for completeness and accuracy. We extracted the number of participants randomised and the number analysed in each group for each reported outcome.

For continuous outcomes, we extracted the number of participants for each treatment arm, arithmetic means and standard deviations (SDs). We extracted data on reported adverse events as dichotomous outcomes.

We attempted to contact the trial authors where the relevant details were not recorded or were unclear.

Assessment of risk of bias in included studies

Two review authors (TAO and OO) independently assessed the risk of bias for the included study using the guidelines provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We assessed the risk of bias in relation to random sequence generation, allocation concealment, blinding, handling of incomplete outcome data, selective outcome reporting, and other bias, giving ratings of 'low risk of bias', 'high risk of bias,' and 'unclear risk of bias' for each domain.

Randomisation sequence generation

If the included study was an RCT, we would have described the method used to generate the allocation sequence to allow an assessment of whether it should produce comparable groups.

The assessment would have been graded as follows:

low risk of bias (truly random processes such as the use of a table of random numbers or computer‐generated random numbers);
high risk of bias (non‐random processes such as use of hospital record numbers or dates of birth);
unclear risk of bias.

Allocation concealment

If the included study was an RCT, we would have described the method used to conceal allocation to interventions prior to assignment and assessed whether intervention allocation could have been predicted before or changed after recruitment. The methods would have been graded as follows:

low risk of bias (use of central allocation which may be web‐based or pharmacy‐controlled; sequentially numbered, opaque, sealed envelopes);
high risk of bias (unsealed, non‐opaque assignment envelopes or not sequentially numbered, allocation by alternation or rotation);
unclear risk of bias (method of concealment not described in significant detail to allow a definite judgement).

Blinding of participants and researchers

For an RCT, we would have described the methods used to blind study participants and researchers from knowledge of which intervention a participant received. Blinded studies, or studies in which non‐blinding was not likely to have affected the results significantly, would have been classified as having a low risk of bias. Non‐blinded studies would have been classified as having a high risk of bias.

Incomplete outcome data

If the included study was an RCT, we would have described the methods used to account for incomplete outcome data, with regards to the amount, nature, and handling of incomplete outcome data. In instances where outcome data were not completely reported, we would have attempted to obtain missing data from the study authors. We would have extracted and reported on data on attrition and exclusions as well as the numbers involved (compared with total randomised), reasons for attrition/exclusion where reported or obtained from investigators, and any re‐inclusions in analyses performed by review authors if such data were available. Unbiased follow‐up would have been considered to have happened when at least 80% of the participants were followed up. Based upon this, we would have judged whether the researchers dealt with incomplete data.

Selective outcome reporting

We attempted to assess the possibility of selective outcome reporting by the investigators in the included trial. We could not examine the study protocols as the authors of the study could not be reached; based upon this, we would have judged whether reports of the study were free from suggestion of selective outcome reporting.

Other bias

We explored other sources of bias, particularly the sources of funding of the included study and other study peculiarities.

Measures of treatment effect

The type of treatment effect used in describing each of the listed outcomes depended on the type of data extracted for the specific outcome.

Continuous data

We reported the mean difference (MD) for continuous outcomes. We presented all measures of effect with their corresponding 95% confidence intervals (CI). We extracted postintervention values and utilised the mean and SD for the analysis.

Binary data

We analysed binary outcomes by calculating the risk ratio (RR), with 95% CIs.

Unit of analysis issues

Not applicable.

Dealing with missing data

We attempted to contact the original authors of included study to request unreported data (e.g. baseline characteristics and number of loss to follow‐up) but received no response. Therefore, we performed the analysis using the available data.

Assessment of heterogeneity

It was not feasible to assess for heterogeneity because we included only one quasi‐randomised study. However, we would have identified statistical heterogeneity in each of the included study using the Chi² test and I² statistic. We would also have used visual assessment of forest plots to identify obvious overlaps and outliers. We would have agreed on how to handle heterogeneity depending on the degree indicated by the I² statistic. As an approximate guide, I² values of less than 25% would have been treated as indicating no heterogeneity, values of 25% to 49% as indicative of low heterogeneity, values of 50% to 74% as indicative of moderate heterogeneity and values of 75% or above as indicative of high heterogeneity (Higgins 2003). When heterogeneity was significant, outliers would have been investigated and we would have attempted to explain the heterogeneity. When heterogeneity was significant and could not be explained, we would have used subgroup analyses and sensitivity analysis to resolve the discrepancy.

Assessment of reporting biases

We examined the included study for within‐study selective outcome reporting. We did not have access to the study protocol, therefore, we could not detect or investigate selective reporting.

Data synthesis

We did not perform a meta‐analysis because only one study was included in this review.

Quality of evidence

We used the GRADE approach, as outlined in the GRADE handbook for grading quality of evidence and strength of recommendations (Schünemann 2013), to assess the quality of evidence for the following (clinically relevant) outcomes: all‐cause mortality, mean total serum calcium levels (total less than 8 mg/dL (2 mmol/L); ionised calcium less than 4.4 mg/dL (1.1 mmol/L)) within one week of EBT and adverse reactions to intervention (apnoea, hypoglycaemia, and cardiac arrest).

Two review authors (TAO and OO) independently assessed the quality of the evidence for each of the outcomes above. We considered evidence from randomised controlled trials as high quality but downgraded the evidence one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates, and presence of publication bias. We used the GRADEpro Guideline Development Tool to create a 'Summary of findings' table to report the quality of the evidence.

The GRADE approach results in an assessment of the quality of a body of evidence in one of four grades.

High: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate: 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: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Subgroup analysis and investigation of heterogeneity

We planned the following subgroup analysis.

Preterm (less than 37 weeks' gestation) versus term (37 weeks' gestation or greater).
Type of anticoagulant in the blood transfused (citrate versus heparin).
Volume of blood used (single‐volume EBT versus double‐volume EBT).
Timing of administration of calcium (administration during EBT versus administration immediately after EBT).

Sensitivity analysis

We planned to agree on eligibility criteria (age range of participants, comparator group, range of doses of intervention, and the range of time points to be evaluated in the analysis), type of data to be analysed (dichotomous or continuous), and the method of data analysis to be adopted. We planned to conduct sensitivity analyses if we encountered disparities with respect to any of the eligibility criteria earlier agreed upon. We planned to input missing data with replacement values and assume all the participants with missing data had a poor outcome. We would have contacted the Cochrane Neonatal Group's statistician when this decision was taken. We planned to perform sensitivity analysis to assess how sensitive the findings were if we assumed that all those participants with missing data had a poor outcome. We planned to use a 'Summary of Findings' table to report the findings of sensitivity analyses.

Results

Description of studies

Results of the search

Our search yielded 383 studies after removing duplicates (Figure 1), and 382 of these were clearly irrelevant. The one potentially relevant study involved 30 neonates and met our inclusion criteria (Locham 2002).

Figure 1

Study flow diagram.

Included studies

One study with high risk of bias conducted on 30 healthy appropriate for gestational age babies with neonatal jaundice, requiring EBT. Participants were allocated alternatively to two groups; study group and control group. There was no specific randomisation or blinding. Fifteen participants in study group received 1 mL of 10% calcium gluconate intravenously for every 100 mL of citrate phosphate dextrose blood exchanged. The control group received only EBT.

Excluded studies

We excluded no studies.

Risk of bias in included studies

We included one small study with an overall high risk of bias as there was no specific randomisation or blinding done (Locham 2002). Participants were allocated alternatively to two groups; study group and control group (Figure 2; Figure 3).

Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

The study was at high risk of selection bias as allocation was not concealed.

Blinding

The study was at high risk of performance bias and detection bias as there was no blinding.

Incomplete outcome data

The study was at low risk of attrition bias as all participants were included in the final analysis.

Selective reporting

The study was at unclear risk of reporting bias as, since we could not access the study protocol, it was unclear if there was selective reporting. However, all important outcomes were reported.

Other potential sources of bias

The study was at unclear risk of other bias as the study did not report baseline characteristics of participants and length of follow‐up.