Corticosteroids for HELLP syndrome in pregnancy

Corticosteroids for HELLP syndrome in pregnancy (Review) Matchaba P, Moodley J This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2007, Issue http://www.thecochranelibrary.com Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd TABLE OF CONTENTS ABSTRACT PLAIN LANGUAGE SUMMARY BACKGROUND OBJECTIVES CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW SEARCH METHODS FOR IDENTIFICATION OF STUDIES METHODS OF THE REVIEW DESCRIPTION OF STUDIES METHODOLOGICAL QUALITY RESULTS DISCUSSION AUTHORS’ CONCLUSIONS POTENTIAL CONFLICT OF INTEREST ACKNOWLEDGEMENTS SOURCES OF SUPPORT REFERENCES TABLES Characteristics of included studies ANALYSES Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone Comparison 02 Dexamethasone versus betamethasone INDEX TERMS COVER SHEET GRAPHS AND OTHER TABLES Analysis 01.01 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 01 Maternal deaths Analysis 01.02 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 02 Postpartum sepsis Analysis 01.03 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 03 Mean platelet counts over 48 hours Analysis 01.04 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 04 Mean hospital stay post randomisation (days) Analysis 01.05 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 05 Neonatal deaths Analysis 01.06 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 06 Neonates with intraventricular hemorrhage Analysis 01.07 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 07 Neonates with respiratory distress syndrome Analysis 01.08 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 08 Neonates with minute Apgars less than Analysis 01.09 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 09 Weight at birth in grams Analysis 01.10 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 10 Neonates with retrolental fibroplasia Analysis 01.11 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 11 Number of cesarean section deliveries Analysis 01.12 Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 12 Time interval (hours) from randomisation to delivery Analysis 02.01 Comparison 02 Dexamethasone versus betamethasone, Outcome 01 Mean arterial pressure: adjusted time-averaged change from baseline Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 2 3 3 4 6 6 8 12 12 13 14 14 15 15 15 16 16 17 17 18 18 19 19 20 20 20 i Analysis 02.02 Comparison 02 Dexamethasone versus betamethasone, Outcome 02 Urinary output (mL/h): adjusted time-averaged change from baseline Analysis 02.03 Comparison 02 Dexamethasone versus betamethasone, Outcome 03 Platelet count (10-9 cells/L): adjusted time-averaged change from baseline Analysis 02.04 Comparison 02 Dexamethasone versus betamethasone, Outcome 04 LDH activity (U/L mean): adjusted time-averaged change from baseline Analysis 02.05 Comparison 02 Dexamethasone versus betamethasone, Outcome 05 AST activity (U/L): adjusted timeaveraged change from baseline Analysis 02.06 Comparison 02 Dexamethasone versus betamethasone, Outcome 06 Number of mothers with oliguria (less than 30 ml/hour for hours) Analysis 02.07 Comparison 02 Dexamethasone versus betamethasone, Outcome 07 Maternal pulmonary edema Analysis 02.08 Comparison 02 Dexamethasone versus betamethasone, Outcome 08 Number of participants needing acute antihypertensive therapy Analysis 02.09 Comparison 02 Dexamethasone versus betamethasone, Outcome 09 Neonates with a minute Apgar less than Analysis 02.10 Comparison 02 Dexamethasone versus betamethasone, Outcome 10 Neonates needing ventilatory support Analysis 02.11 Comparison 02 Dexamethasone versus betamethasone, Outcome 11 Neonates with respiratory distress syndrome Analysis 02.12 Comparison 02 Dexamethasone versus betamethasone, Outcome 12 Neonatal sepsis Analysis 02.13 Comparison 02 Dexamethasone versus betamethasone, Outcome 13 Neonatal hyperbilirubinemia Analysis 02.14 Comparison 02 Dexamethasone versus betamethasone, Outcome 14 Fetal or neonatal death Analysis 02.15 Comparison 02 Dexamethasone versus betamethasone, Outcome 15 Neonate time to discharge (days: mean) Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 21 21 21 22 22 23 23 24 24 25 25 26 26 27 ii Corticosteroids for HELLP syndrome in pregnancy (Review) Matchaba P, Moodley J This record should be cited as: Matchaba P, Moodley J Corticosteroids for HELLP syndrome in pregnancy Cochrane Database of Systematic Reviews 2004, Issue Art No.: CD002076 DOI: 10.1002/14651858.CD002076.pub2 This version first published online: 26 January 2004 in Issue 1, 2004 Date of most recent substantive amendment: 31 October 2003 ABSTRACT Background Hemolysis, elevated liver enzymes and low platelets (HELLP) syndrome is a severe form of pre-eclampsia Pre-eclampsia is a multi-system disease of pregnancy associated with an increase in blood pressure and increased perinatal and maternal morbidity and mortality Eighty per cent of women with HELLP syndrome present before term There are suggestions from observational studies that steroid treatment in HELLP syndrome may improve disordered maternal hematological and biochemical features and perhaps perinatal mortality and morbidity Objectives To summarise the evidence on the effects of corticosteroids on maternal and neonatal mortality and morbidity in women with HELLP syndrome Search strategy We searched the Cochrane Pregnancy and Childbirth Group trials register (October 2003) We scanned lists of references from review articles and primary studies Selection criteria Randomised and quasi-randomised trials evaluating the effects of adjunctive corticosteroids in patients diagnosed with HELLP syndrome were sought Data collection and analysis The two authors independently applied inclusion criteria, assessed trial quality and extracted relevant data Main results Of the five studies reviewed (n = 170), three were conducted antepartum and two postpartum Four of the studies randomised participants to standard therapy or dexamethasone One study compared dexamethasone with betamethasone Dexamethasone versus control There were no significant differences in the primary outcomes of maternal mortality and morbidity due to placental abruption, pulmonary oedema and liver hematoma or rupture Of the secondary maternal outcomes, there was a tendency to a greater platelet count increase over 48 hours, statistically significantly less mean number of hospital stay days (weighted mean difference (WMD) -4.50, 95% confidence interval (CI) -7.13 to -1.87), mean interval (hours) to delivery (41 ± 15) versus (15 ± 4.5) (p = 0.0068) in favour of women allocated to dexamethasone There were no significant differences in perinatal mortality or morbidity due to respiratory distress syndrome, need for ventilatory support, intracerebral hemorrhage, necrotizing enterocolitis and a five minute Apgar less than seven The mean birthweight was significantly greater in the group allocated to dexamethasone (WMD 247.00, 95% CI 65.41 to 428.59) Dexamethasone versus betamethasone There were no significant differences in all the maternal and perinatal mortality and in primary morbidity outcomes Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd Women randomised to dexamethasone fared significantly better for: oliguria, mean arterial pressure, mean increase in platelet count, mean increase in urinary output and liver enzyme elevations Authors’ conclusions There is insufficient evidence to determine whether adjunctive steroid use in HELLP syndrome decreases maternal and perinatal mortality, major maternal and perinatal morbidity PLAIN LANGUAGE SUMMARY More research is needed to determine if corticosteroids can improve outcomes for women and babies affected by HELLP syndrome in pregnancy Pre-eclampsia, also known as toxaemia, is a potentially life-threatening condition of pregnancy It involves high blood pressure (hypertension) and protein in the urine HELLP syndrome (hemolysis, elevated liver enzymes and low platelets) is a severe form of preeclampsia, which can cause a tendency to bleed and other complications Corticosteroids may be able to normalise some of the abnormal biochemical changes caused by HELLP, as well as reduce hypertension The review of trials of corticosteroids for women with HELLP (both during pregnancy and after the birth) found too little evidence to be sure of the effects, but more research is worthwhile BACKGROUND Hemolysis, elevated liver enzymes and low platelets (HELLP) syndrome is a severe form of pre-eclampsia (Pritchard 1954) Preeclampsia is a condition associated with an increase in blood pressure, protein loss via the urine, and oedema in pregnancy and occurs in 5% to 8% of all pregnancies Its cause is not known, but it remains a major cause of perinatal and maternal morbidity and mortality (Duley 1992) HELLP syndrome occurs in about 20% of severely pre-eclamptic women and does so before term in 80% of cases; 10% of these occur before 27 weeks’ gestation (Sibai 1993) Premature delivery contributes to the very high rate of perinatal morbidity and mortality About a third of the cases are diagnosed for the first time postpartum (Sibai 1993) Although it is agreed that hemolysis, liver dysfunction and thrombocytopenia must be present to make a diagnosis of HELLP syndrome, there is disagreement over the specific biochemical and hematological criteria to be used For the purposes of this review, the criteria specified by Sibai (Sibai 1986) and Martin (Martin 1991) will be used Sibai’s criteria include hemolysis as evidenced by an abnormal peripheral smear, lactate dehydrogenase (LDH) greater than 600 IU/L or Total Bilirubin (TB) greater than 1.2 ULN; elevated liver enzymes as evidenced by an aspartate transaminase (AST) greater than 70 IU/L and platelets less than 100 000/mm³ Martin’s criteria include hemolysis as evidenced by a falling hematocrit, LDH greater than 164 IU/L or a bleeding diathesis; elevated liver enzymes as evidenced an AST greater than 48 IU/L and alanine transaminase (ALT) greater than 24 IU/L and platelets less than 100 000 mm³ Maternal mortality (1% to 3.5%) and morbidity are increased by the hypertensive complications of pre-eclampsia, the bleeding tendency that results from the low platelets and liver dysfunction (Sibai 1993; Weinstein 1985; Martin 1991) Since HELLP syndrome is common before term, any intervention that prolongs the time from diagnosis to delivery by stabilising the clinical condition, may decrease perinatal morbidity and mortality Steroids have been shown in observational studies (Magann 1993; Clark 1986; Yeast 1987) to improve the deranged maternal biochemical and haematological indices associated with the syndrome This benefit has also been shown in HELLP syndrome diagnosed postpartum (Martin 1997; Yalcin 1998) However, it may be argued that delaying delivery may worsen the maternal morbidity and increase both maternal and fetal mortality Glucocorticoids have been shown to improve fetal lung maturation, decreasing the occurrence and severity of respiratory distress syndrome and the overall fetal morbidity and mortality after preterm birth (Crowley 1999) Dexamethasone at doses up to a total of 48 mg over a 48 hour period may be given intramuscularly in order to improve fetal lung function However, the steroid dosing regimens that have been suggested for HELLP syndrome are different from the one used for fetal lung maturation The maternal and fetal effects of such a regimen have not been quantified Also, the number of women with HELLP syndrome who have been treated with steroids is probably small This review therefore aims to summarise existing evidence of the effects of steroids in HELLP syndrome Any effects may be different according to the gestational age, severity of disease, parity and whether the mother is antepartum or postpartum The effect estimates will be explored for these factors Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd OBJECTIVES (1) To evaluate the effects of corticosteroids in women with HELLP syndrome during pregnancy or shortly after delivery The following primary hypothesis will be tested: In comparing women given steroids with control/placebo groups, there will be no difference in: (a) maternal morbidity and mortality; (b) perinatal morbidity and mortality (2) In addition, we will explore whether the possible effects of steroids on the mother differ according to: (a) gestational age at treatment; and (b) whether the mother is antepartum or postpartum CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW Types of studies All randomised controlled trials and trials which used pseudorandomised methods, such as alternate allocation Types of participants All antepartum and postpartum women diagnosed clinically and by biochemical parameters as having the HELLP syndrome Types of intervention Any corticosteroid versus placebo or no treatment Types of outcome measures Primary outcomes (1) maternal mortality; (2) perinatal mortality; (3) maternal morbidity, namely: • presence of liver hematoma and rupture; • pulmonary oedema; • renal failure; • abruptio placentae; • average time taken to return to a normal urine output; • average time taken to return to a normal blood pressure; • average time to delivery; • mode of delivery (2) Neonatal • the mean gestational age and weight at birth; • the mean Apgar scores at birth SEARCH METHODS FOR IDENTIFICATION OF STUDIES See: Cochrane Pregnancy and Childbirth Group methods used in reviews We searched the Cochrane Pregnancy and Childbirth Group trials register (October 2003) The Cochrane Pregnancy and Childbirth Group’s trials register is maintained by the Trials Search Co-ordinator and contains trials identified from: quarterly searches of the Cochrane Central Register of Controlled Trials (CENTRAL); monthly searches of MEDLINE; handsearches of 30 journals and the proceedings of major conferences; weekly current awareness search of a further 37 journals Details of the search strategies for CENTRAL and MEDLINE, the list of handsearched journals and conference proceedings, and the list of journals reviewed via the current awareness service can be found in the ’Search strategies for identification of studies’ section within the editorial information about the Cochrane Pregnancy and Childbirth Group Trials identified through the searching activities described above are given a code (or codes) depending on the topic The codes are linked to review topics The Trials Search Co-ordinator searches the register for each review using these codes rather than keywords (4) perinatal morbidity, namely: • respiratory distress syndrome and ventilatory support required; METHODS OF THE REVIEW • intracerebral haemorrhage; • necrotizing enterocolitis Secondary outcomes (1) Maternal After randomisation, the • average time taken to return to normal biochemical and haematological parameters; The inclusion criteria were applied to all identified trials by the two authors independently Each included trial was assessed in terms of adequacy of concealment of allocation, generation of allocation sequence, blinding and follow up of subjects (Clarke 2000) Data extraction was undertaken by the two reviewers independently, who collected the following information for each trial: Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd • number of women; • women’s characteristics including ethnic origin, parity, whether the women were antepartum or postpartum, gestational age at diagnosis, number of babies in current pregnancy, the mode of delivery and time from diagnosis to delivery; • the haematological and biochemical results; In none of the five included studies was blinding described in the methods section There was significant loss to follow up in one study (Magann 1994a) Only 25 of the original 40 participants randomised were accounted for in the results section Intention to treat analysis was not performed in this study The other studies had no loss to follow up • morbidity and mortality outcomes; • the type, dose and duration of steroid used; • other medication used RESULTS Five studies involving 170 women are included Babies’ characteristics including ethnic origin, gestational age at birth, weight, Apgar scores and morbidity and mortality outcomes Primary outcomes It was intended to explore whether any heterogeneity of effect was explained by whether the woman was antepartum or postpartum, the gestational age, severity of disease and parity Trials that compared dexamethasone plus standard therapy versus standard therapy alone There were four trials that used this study design (Magann 1994a; Magann 1994b; Yalcin 1998; Vigil-De Gracia 1997) DESCRIPTION OF STUDIES Seven published studies were identified that met the inclusion criteria One of the studies (Kadanali 1997) has not been included in the data analysis because it is being translated into English and another (Isler 2003) has not been included because we have asked for the original data Of the five studies that have been analysed, four have compared steroid use (dexamethasone) with placebo (Isler 2001; Magann 1994a; Magann 1994b; Yalcin 1998: Vigil-De Gracia 1997) One study compared two different steroid regimens (Isler 2001) The total number of trial participants in the five studies is 170 Three of the reviewed studies were conducted antepartum (Isler 2001; Magann 1994b; Vigil-De Gracia 1997) The other two were conducted postpartum (Magann 1994a; Yalcin 1998) METHODOLOGICAL QUALITY Three studies employed adequate randomisation and allocation concealment methods (Isler 2001; Magann 1994a; Magann 1994b) However, in Magann 1994b, women allocated to standard treatment (control) had higher mean platelet counts than those allocated to steroid treatment In the two other studies (Yalcin 1998; Vigil-De Gracia 1997), there was no mention of the randomisation method used, nor whether allocation concealment was implemented Furthermore in VigilDe Gracia 1997, the groups randomly allocated to standard therapy differed in the maternal platelet count when compared with the group allocated to steroid therapy, the platelet count being lower in the group randomly allocated to receive steroids None of the trials were placebo-controlled Maternal mortality There was one maternal death in the group randomised to standard therapy (Vigil-De Gracia 1997, n = 34) and this was not statistically significant (relative risk (RR) 0.33, 95% confidence interval (CI) 0.01 to 7.65) Perinatal mortality There was no statistically significant difference in neonatal deaths (RR 0.36, 95% CI 0.04 to 3.02) in the one study (n = 25) Maternal morbidity There were no cases of liver hematoma or rupture, pulmonary oedema, renal failure or placental abruption in either group Perinatal morbidity i) Intraventricular hemorrhage events occurred in one study (Magann 1994b, n = 25) and the difference was not statistically significant (RR 7.54, 95% CI 0.43 to 132.35) ii) Respiratory distress syndrome events occurred in one study (Magann 1994b) and there was no statistically significant difference (RR 1.00, 95% CI 0.25 to 4.00) iii) Retrolental fibroplasia occurred in one neonate allocated to placebo (Magann 1994b) but the difference was not statistically significant (RR 0.36, 95% CI 0.02 to 8.05) No intracerebral hemorrhagic events nor necrotizing enterocolitis were recorded Trial that compared dexamethasone and betamethasone There was only one study with this design (Isler 2001, n = 40) Maternal mortality There were no maternal deaths Perinatal mortality There was no significant difference in neonatal deaths when dexamethasone was compared with betamethasone (RR 0.95, 95% CI 0.15 to 6.08) Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd Maternal morbidity There were no cases of liver hematoma or rupture, pulmonary oedema, or abruptio placentae in either group There was a statistically significant difference in maternal oliguria (RR 0.06, 95% CI 0.00 to 0.93) in favour of participants randomised to dexamethasone iii) The mean increase in platelet count (WMD 8.10, 95% CI 6.23 to 9.97) iv) The mean decrease in lactate dehydrogenase activity (U/L) (WMD -54.20, 95% CI -88.22 to -20.18) v) The mean decrease in aspartate transaminase activity (U/L) (WMD -30.30, 95% CI -36.06 to -24.54) Perinatal morbidity There was a tendency for fewer neonates allocated to dexamethasone to have ventilatory support or respiratory distress syndrome when compared with those allocated to betamethasone However, this was not statistically significant (RR 0.54, 95% CI 0.19 to 1.56) In this study, no cases of intracerebral hemorrhage and necrotizing enterocolitis were recorded The number of participants needing acute antihypertensive therapy in the dexamethasone group differed significantly statistically compared with those allocated to betamethasone (RR 0.29, 95% CI 0.12 to 0.73) Secondary outcomes Trials that compared dexamethasone plus standard therapy versus standard therapy alone Maternal morbidity i) There was no statistically significant difference in postpartum sepsis (RR 2.00, 95% CI 0.20 to 19.78) in one study (n = 30) and cesarean sections (RR 0.93, 95% CI 0.66 to 1.31) between the two groups (one study, n = 34) ii) There was a tendency to a greater platelet count increase over 48 hours in participants allocated to dexamethasone (weighted mean difference (WMD) 40.60, 95% CI -26.12 to 107.32) but this result must be interpreted with caution because the data are skewed and are derived from only one small study (Vigil-De Gracia 1997, n=34) iii) There was a statistically significant difference in the mean number of hospital stay days postrandomisation (WMD -4.50, 95% CI -7.13 to -1.87) in favour of participants allocated to dexamethasone iv) There was a significant difference in the mean interval (hours) from randomisation to delivery (41 ± 15) versus (15 ± 4.5) (p = 0.0068) in favour of women allocated to dexamethasone in the single study that looked at this outcome (Magann 1994b, n = 25) Neonatal morbidity The number of neonates with a five minute Apgar less than seven did not differ significantly (RR 1.00 95% CI 0.25 to 4.00): one study, n = 24 The mean weight at birth was significantly greater in the group allocated to steroids (WMD 247.00, 95% CI 65.41 to 428.59) Trial that compared dexamethasone and betamethasone Maternal morbidity There was a statistically significant difference in favour of participants allocated to dexamethasone in the adjusted time-average change from baseline in the following secondary outcomes: i) The mean arterial pressure decrease (WMD -7.50, 95% CI -8.37 to -6.63) ii) The mean increase in urinary output (WMD 24.80, 95% CI 19.58 to 30.02) Neonatal morbidity There were no statistically significant differences between the two groups with regards to the number of neonates with a five minute Apgar less than seven (RR 0.95, 95% CI 0.22 to 4.14), neonatal sepsis (RR 4.76, 95% CI 0.24 to 93.19), neonatal hyperbilirubinemia (RR 2.85, 95% CI 0.32 to 25.07) and mean time to discharge in days (WMD -5.40, 95% CI -19.53 to 8.73) There was no trial that compared betamethasone plus standard therapy versus standard therapy alone DISCUSSION Trials that compared participants randomised with steroid (dexamethasone) use and placebo (Magann 1994a; Magann 1994b; Vigil-De Gracia 1997; Yalcin 1998) HELLP syndrome is a severe form of pre-eclampsia (Pritchard 1954), whose cause is unknown and remains a major cause of perinatal morbidity and mortality (Duley 1992) In this review there was no statistically significant difference in the primary outcome of maternal mortality (relative risk (RR) 0.33, 95% confidence interval (CI) 0.01 to 7.65) There was one maternal death in the group allocated to placebo in a trial with a total of 34 participants (Vigil-De Gracia 1997) Besides the small sample size of this trial, the randomisation method used was not mentioned and the methodology used for allocation concealment is unclear Furthermore the postrandomisation patient characteristics differed significantly with regards to platelet count, before treatment was instituted There were four neonatal deaths in one small trial (Magann 1994b, n = 25), which were not statistically significant (RR 0.36, 95% CI 0.04 to 3.02) Although allocation concealment and randomisation were adequate in this trial, maternal platelet levels differed significantly postrandomisation before treatment commenced (p = 0.034) and the ratio of black to female participants was 2:1 Maternal morbidity primary outcomes of pulmonary oedema, presence of a liver hematoma or rupture, renal and abruptio placentae did not differ significantly between the two groups of participants Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd None of the primary perinatal morbidity outcomes, respiratory distress syndrome, need for ventilatory support, intracerebral hemorrhage and necrotizing enterocolitis, differed significantly The only secondary maternal morbidity outcome that was statistically significant in favour of steroid use was the mean number of days of hospital stay (weighted mean difference (WMD) -4.50, 95% CI -7.13 to -1.87) The number of neonates with Apgar scores less than seven after five minutes were not significantly different The mean weight at birth differed significantly (WMD 247.00, 95% CI 65.41 to 428.59) in one study (Magann 1994b) In summary, because of the small sample sizes in the trials reviewed, there is no evidence that either supports or refutes the use of steroids in HELLP syndrome antenatally and in the postpartum period in order to decrease or increase maternal and perinatal mortality and the primary morbidity outcomes of interest in this review One trial compared dexamethasone use with betamethasone (Isler 2001): There were no maternal or neonatal primary outcome differences with regards to death and maternal morbidity, with the exception of oliguria (less than 30 ml urine/hour for two hours) (RR 0.06, 95% CI 0.00 to 0.93), in favour of dexamethasone The main findings were in the maternal hematological and biochemical parameters In all instances of adjusted time-average change from baseline for mean arterial pressure decrease, mean increase in the platelet count, mean decrease in lactate dehydrogenase and aspartate transaminase activity and the need for acute antihypertensive therapy, women allocated to dexamethasone fared better significantly than those allocated to betamethasone This is supportive of the observational data referenced above that had similar findings There were no significant differences in the secondary neonatal outcomes of Apgar score of less than seven after five minutes, neonatal sepsis, hyperbilirubinaemia and time to discharge from hospital In summary, because of the small sample size there is no evidence from this trial that indicates that when dexamethasone is given intravenously (10 mg 12 hourly before delivery), that there is a statistically significant difference in the maternal biochemical and hematological parameters when compared with betamethasone (12 mg intramuscularly every 24 hours) cient power to detect a significant difference in the primary maternal and neonatal outcomes of death and severe morbidity, there is no evidence to support the addition of dexamethasone or betamethasone to standard therapy in HELLP syndrome However, it is important to note that glucocorticoids have been shown to improve fetal lung maturity and overall fetal mortality and morbidity when the appropriate dosing regimens are used (Crowley 1999) Implications for research The high maternal and perinatal mortality and morbidity associated with HELLP syndrome makes it imperative that a large study with the appropriate design be done to determine the effect of antenatal and postpartum steroids on the primary and secondary outcomes chosen in this review Steroids are relatively cheap and if found efficacious in decreasing the primary outcomes sought in the review, they would provide a cost-effective intervention for HELLP syndrome Because HELLP syndrome is not very common (Sibai 1993), a global multi-centre study design may be the most efficient study design to use in order to recruit sufficient patients to adequately answer the review question POTENTIAL CONFLICT OF INTEREST We certify that we have no affiliations or involvement in any organisation or entity with a direct financial interest in the subject matter of the review (e.g employment, consultancy, stock ownership, honoraria and expert testimony) The researchers wrote this review independently and Novartis Pharmaceuticals has no commercial or intellectual interest in the topic or products reviewed ACKNOWLEDGEMENTS Contributions: Jim Neilson, Michel Boulvain, the Cochrane Pregnancy and Childbirth Group’s panel of consumers SOURCES OF SUPPORT External sources of support AUTHORS’ CONCLUSIONS • No sources of support supplied Implications for practice Internal sources of support Until a large enough study that is adequately designed with suffi- • Medical Research Council SOUTH AFRICA Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd REFERENCES References to studies included in this review Isler 2001 {published data only} Isler C, Barrilleaux P, Magann E, Bass J, Martin J A prospective, randomized trial comparing the efficacy of dexamethasone and betamethasone for the treatment of antepartum HELLP (hemolysis, elevated liver enzymes, and low platelet count) syndrome [Un estudio prospectivo, randomizado que compara la eficacia de la dexametasona y betametasona para el tratamiento anteparto del sindron de hellp (hemolisis, elevacion de enzimas hepaticas y plaquetopenia)] Revista Chilena de Obstetricia y Ginecologia 2001;66(3):248–50 Isler CM, Barrilleaux PS, Magann EF, Bass JD, Martin JN A prospective, randomized trial comparing the efficacy of dexamethasone and betamethasone for the treatment of antepartum HELLP (hemolysis, elevated liver enzymes, and low platelet count) syndrome American Journal of Obstetrics and Gynecology 2001;184(7):1332–7 Magann 1994a {published data only} Magann EF, Perry KG, Meydrech EF, Harris RL, Suneet PC, Martin JN Postpartum corticosteroids: accelerated recovery from the syndrome of hemolysis, elevated liver enzymes, and low platelets (HELLP) American Journal of Obstetrics Gynecology 1994;171(4): 1154–8 Kadanali 1997 Kadanali S, Kucukozkan T, Bukam B Helpful effect of high-dose corticosteroid use on Hellp syndrome Jinekoloji Ve Obstetrik Dergisi 1997;11:55–8 Additional references Clark 1986 Clark SL, Phelan JR, Allen SH, Golde SR Antepartum reversal of hematologic abnormalities associated with the HELLP syndrome A report of three cases Journal of Reproductive Medicine 1986;31:70–2 Clarke 2000 Clarke M, Oxman AD, editors Cochrane Reviewers’ Handbook 4.1 [updated June 2000] In: Review Manager (RevMan) [Computer program] Version 4.1 Oxford, England: The Cochrane Collaboration, 2000 Crowley 1999 Crowley P Prophylactic corticosteroids for preterm birth (Cochrane Review) In: The Cochrane Library, 4, 1999 Oxford: Update software Duley 1992 Duley L Maternal mortality associated with hypertensive disorders of pregnancy in Africa, Asia, Latin America and the Caribbean British Journal of Obstetrics and Gynaecology 1992;99:547–53 Magann 1994b {published data only} Magann EF, Bass D, Chauhan SP, Sullivan D, Martin RW, Martin JN Antepartum corticosteroids: disease stabilization in patients with the syndrome of hemolysis, elevated liver enzymes, and low platelets (HELLP) American Journal of Obstetrics and Gynecology 1994;171: 1148–53 Magann 1993 Magann EF, Martin RW, Issacs JD, Blake PG, Morrison JC, Martin JN Jr Corticosteroids for the enhancement of fetal lung maturity: impact on the gravida with preeclampsia and the HELLP syndrome Australian and New Zealand Journal of Obstetrics and Gynaecology 1993;33:127–31 Magann EF, Bass D, Chauhan SP, Sullivan DL, Martin RW, Martin JN Antepartum corticosteroids: disease stabilization in patients with HELLP syndrome American Journal of Obstetrics and Gynecology 1994;170:410 Martin 1991 Martin JN Jr, Blake PG, Perry KG, McCaul JF, Hess LW, Martin RW The natural history of HELLP syndrome: patterns of disease progression and regression American Journal of Obstetrics and Gynecology 1991;164:1500–9 Vigil-De Gracia 1997 {published data only} Vigil-De Gracia P, Garcia-Caceres E Dexamethasone in the postpartum treatment of HELLP syndrome International Journal of Gynecology & Obstetrics 1997;59:217–21 Yalcin 1998 {published data only} Yalcin OT, Sener T, Hassa H, Ozalp S, Okur A Effects of postpartum corticosteroids in patients with HELLP syndrome International Journal of Gynecology & Obstetrics 1998;61:141–8 References to studies awaiting assessment Isler 2003 Isler C, Magann E, Rinehart B, Terrone D, Bass J, Martin J Jr Dexamethasone versus betmethasone for postpartum hellp syndrome: a randomized prospective clinical trial of comparative efficacy American Journal of Obstetrics and Gynecology 2000;182(1 Pt 2):S87 Isler CM, Magann EF, Rinehart BK, Terrone DA, Bass JD, Martin JN Jr Dexamethasone compared with betamethasone for glucocorticoid treatment of postpartum hellp syndrome International Journal of Gynecology & Obstetrics 2003;80(3):291–7 Martin 1997 Martin JN, Perry KG Jr, Blake PG, May WA, Moore A, Robinette L Better maternal outcomes are achieved with dexamethasone therapy for postpartum HELLP (hemolysis, elevated liver enzymes and thrombocytopenia) syndrome American Journal of Obstetrics and Gynecology 1997;177:1011–7 Pritchard 1954 Pritchard J, Weisman R, Ratnoff O, Vosburgh G Intravascular haemolysis, thrombocytopenia and other haematologic abnormalities associated with severe toxemia of pregnancy New England Journal of Medicine 1954;250:89–98 Sibai 1986 Sibai BM, Taslimi MM, el-Nazer A, Aman E, Mabie BC, Ryan GM Maternal-perinatal outcome associated with the syndrome of hemolysis, elevated liver enzymes, and low platelets in severe pre-eclampsiaeclampsia American Journal of Obstetrics and Gynecology 1986;155: 501–9 Sibai 1993 Sibai BM, Ramadan MK, Usta I, Salama M, Mercer BM, Friedman SA Maternal morbidity and mortality in 442 pregnancies with Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 06 Neonates with intraventricular hemorrhage 07 Neonates with respiratory distress syndrome 08 Neonates with minute Apgars less than 09 Weight at birth in grams 10 Neonates with retrolental fibroplasia 11 Number of cesarean section deliveries 12 Time interval (hours) from randomisation to delivery 25 Relative Risk (Fixed) 95% CI 7.54 [0.43, 132.35] 24 Relative Risk (Fixed) 95% CI 1.00 [0.25, 4.00] 24 Relative Risk (Fixed) 95% CI 1.00 [0.25, 4.00] 25 Weighted Mean Difference (Fixed) 95% CI 25 Relative Risk (Fixed) 95% CI 247.00 [65.41, 428.59] 0.36 [0.02, 8.05] 34 Relative Risk (Fixed) 95% CI 0.93 [0.66, 1.31] Weighted Mean Difference (Fixed) 95% CI Totals not selected Comparison 02 Dexamethasone versus betamethasone Outcome title 01 Mean arterial pressure: adjusted time-averaged change from baseline 02 Urinary output (mL/h): adjusted time-averaged change from baseline 03 Platelet count (10-9 cells/L): adjusted time-averaged change from baseline 04 LDH activity (U/L mean): adjusted time-averaged change from baseline 05 AST activity (U/L): adjusted time-averaged change from baseline 06 Number of mothers with oliguria (less than 30 ml/hour for hours) 07 Maternal pulmonary edema 08 Number of participants needing acute antihypertensive therapy 09 Neonates with a minute Apgar less than 10 Neonates needing ventilatory support 11 Neonates with respiratory distress syndrome 12 Neonatal sepsis 13 Neonatal hyperbilirubinemia 14 Fetal or neonatal death No of studies No of participants 40 Weighted Mean Difference (Fixed) 95% CI -7.50 [-8.37, -6.63] 40 Weighted Mean Difference (Fixed) 95% CI 24.80 [19.58, 30.02] 40 Weighted Mean Difference (Fixed) 95% CI 8.10 [6.23, 9.97] 40 Weighted Mean Difference (Fixed) 95% CI -54.20 [-88.22, -20.18] 40 Weighted Mean Difference (Fixed) 95% CI -30.30 [-36.06, -24.54] 40 Relative Risk (Fixed) 95% CI 0.06 [0.00, 0.93] 1 40 40 Relative Risk (Fixed) 95% CI Relative Risk (Fixed) 95% CI Not estimable 0.29 [0.12, 0.73] 39 Relative Risk (Fixed) 95% CI 0.95 [0.22, 4.14] 39 Relative Risk (Fixed) 95% CI 0.54 [0.19, 1.56] 39 Relative Risk (Fixed) 95% CI 0.54 [0.19, 1.56] 1 39 39 39 Relative Risk (Fixed) 95% CI Relative Risk (Fixed) 95% CI Relative Risk (Fixed) 95% CI 4.76 [0.24, 93.19] 2.85 [0.32, 25.07] 0.95 [0.15, 6.08] Statistical method Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd Effect size 13 15 Neonate time to discharge (days: mean) 39 Weighted Mean Difference (Fixed) 95% CI -5.40 [-19.53, 8.73] INDEX TERMS Medical Subject Headings (MeSH) Adrenal Cortex Hormones [∗ therapeutic use]; Betamethasone [therapeutic use]; Dexamethasone [therapeutic use]; HELLP Syndrome [∗ drug therapy]; Randomized Controlled Trials MeSH check words Female; Humans; Pregnancy COVER SHEET Title Corticosteroids for HELLP syndrome in pregnancy Authors Matchaba P, Moodley J Contribution of author(s) PT Matchaba conceived the idea, and wrote the protocol and review J Moodley assisted with the writing of the protocol and the review Issue protocol first published 2000/2 Review first published 2004/1 Date of most recent amendment 24 August 2005 Date of most recent SUBSTANTIVE amendment 31 October 2003 What’s New Information not supplied by author Date new studies sought but none found Information not supplied by author Date new studies found but not yet included/excluded 31 October 2003 Date new studies found and included/excluded 18 July 2002 Date authors’ conclusions section amended Information not supplied by author Contact address Dr Patrice Tinaye Matchaba Medical Director Novartis Pharmaceutical Corporation CD and MA, Building 122 One Health Plaza East Hanover New Jersey NJ 07936 - 108 USA E-mail: patrice.matchaba@pharma.novartis.com Tel: +1 862 7788443 Fax: +1 973 7813579 DOI 10.1002/14651858.CD002076.pub2 Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 14 Cochrane Library number CD002076 Editorial group Cochrane Pregnancy and Childbirth Group Editorial group code HM-PREG GRAPHS AND OTHER TABLES Analysis 01.01 Review: Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 01 Maternal deaths Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 01 Maternal deaths Study Vigil-De Gracia 1997 Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 0/17 1/17 100.0 0.33 [ 0.01, 7.65 ] 17 17 100.0 0.33 [ 0.01, 7.65 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=0.69 p=0.5 0.1 0.2 0.5 Favour Dexamethasone Analysis 01.02 Review: 10 Favour control Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 02 Postpartum sepsis Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 02 Postpartum sepsis Study Yalcin 1998 Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 2/15 1/15 100.0 2.00 [ 0.20, 19.78 ] 15 15 100.0 2.00 [ 0.20, 19.78 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=0.59 p=0.6 0.1 0.2 0.5 Favour Dexamethasone Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 10 Favour control 15 Analysis 01.03 Review: Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 03 Mean platelet counts over 48 hours Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 03 Mean platelet counts over 48 hours Study Dexamethasone Vigil-De Gracia 1997 Total (95% CI) Control N Mean(SD) N Mean(SD) 17 110.26 (118.00) 17 69.66 (76.00) 17 Weighted Mean Difference (Fixed) Weight 95% CI (%) 17 Weighted Mean Difference (Fixed) 95% CI 100.0 40.60 [ -26.12, 107.32 ] 100.0 40.60 [ -26.12, 107.32 ] Test for heterogeneity: not applicable Test for overall effect z=1.19 p=0.2 -100.0 -50.0 50.0 Favour control Analysis 01.04 Review: 100.0 Favour Dexamethasone Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 04 Mean hospital stay post randomisation (days) Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 04 Mean hospital stay post randomisation (days) Study Yalcin Dexamethasone 1998 Total (95% CI) Control N Mean(SD) N Mean(SD) 15 6.00 (4.10) 15 10.50 (3.20) 15 Weighted Mean Difference (Fixed) Weight Weighted Mean Difference (Fixed) 95% CI (%) 95% CI 15 100.0 -4.50 [ -7.13, -1.87 ] 100.0 -4.50 [ -7.13, -1.87 ] Test for heterogeneity: not applicable Test for overall effect z=3.35 p=0.0008 -10.0 -5.0 Favour Dexamethasone 5.0 10.0 Favour control Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 16 Analysis 01.05 Review: Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 05 Neonatal deaths Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 05 Neonatal deaths Study Magann 1994b Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 1/12 3/13 100.0 0.36 [ 0.04, 3.02 ] 12 13 100.0 0.36 [ 0.04, 3.02 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=0.94 p=0.3 0.1 0.2 0.5 Favour Dexamethasone Analysis 01.06 Review: 10 Favour control Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 06 Neonates with intraventricular hemorrhage Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 06 Neonates with intraventricular hemorrhage Study Magann 1994b Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 3/12 0/13 100.0 7.54 [ 0.43, 132.35 ] 12 13 100.0 7.54 [ 0.43, 132.35 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=1.38 p=0.2 0.001 0.01 0.1 Favour Dexamethasone Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 10 100 1000 Favour control 17 Analysis 01.07 Review: Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 07 Neonates with respiratory distress syndrome Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 07 Neonates with respiratory distress syndrome Study Magann 1994b Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight n/N n/N 95% CI (%) Relative Risk (Fixed) 95% CI 3/12 3/12 100.0 1.00 [ 0.25, 4.00 ] 12 12 100.0 1.00 [ 0.25, 4.00 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=0.00 p=1 0.1 0.2 0.5 Favour Dexamethasone Analysis 01.08 Review: 10 Favour control Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 08 Neonates with minute Apgars less than Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 08 Neonates with minute Apgars less than Study Magann 1994b Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 3/12 3/12 100.0 1.00 [ 0.25, 4.00 ] 12 12 100.0 1.00 [ 0.25, 4.00 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=0.00 p=1 0.1 0.2 0.5 Favour Dexamethasone Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 10 Favour control 18 Analysis 01.09 Review: Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 09 Weight at birth in grams Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 09 Weight at birth in grams Study Magann 1994b Total (95% CI) Dexamethasone Control N Mean(SD) N Mean(SD) 12 1758.00 (230.00) 13 1511.00 (233.00) 12 Weighted Mean Difference (Fixed) Weight 95% CI (%) 13 Weighted Mean Difference (Fixed) 95% CI 100.0 247.00 [ 65.41, 428.59 ] 100.0 247.00 [ 65.41, 428.59 ] Test for heterogeneity: not applicable Test for overall effect z=2.67 p=0.008 -1000.0 -500.0 500.0 1000.0 Favour control Analysis 01.10 Review: Favour Dexamethasone Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 10 Neonates with retrolental fibroplasia Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 10 Neonates with retrolental fibroplasia Study Magann 1994b Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 0/12 1/13 100.0 0.36 [ 0.02, 8.05 ] 12 13 100.0 0.36 [ 0.02, 8.05 ] Total events: (Dexamethasone), (Control) Test for heterogeneity: not applicable Test for overall effect z=0.65 p=0.5 0.1 0.2 0.5 Favour Dexamethasone Corticosteroids for HELLP syndrome in pregnancy (Review) Copyright © 2007 The Cochrane Collaboration Published by John Wiley & Sons, Ltd 10 Favour control 19 Analysis 01.11 Review: Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 11 Number of cesarean section deliveries Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 11 Number of cesarean section deliveries Study Vigil-De Gracia 1997 Total (95% CI) Dexamethasone Control Relative Risk (Fixed) Weight Relative Risk (Fixed) n/N n/N 95% CI (%) 95% CI 13/17 14/17 100.0 0.93 [ 0.66, 1.31 ] 17 17 100.0 0.93 [ 0.66, 1.31 ] Total events: 13 (Dexamethasone), 14 (Control) Test for heterogeneity: not applicable Test for overall effect z=0.42 p=0.7 0.1 0.2 0.5 Favour Dexamethasone Analysis 01.12 Review: 10 Favour control Comparison 01 Dexamethasone plus standard treatment versus standard treatment alone, Outcome 12 Time interval (hours) from randomisation to delivery Corticosteroids for HELLP syndrome in pregnancy Comparison: 01 Dexamethasone plus standard treatment versus standard treatment alone Outcome: 12 Time interval (hours) from randomisation to delivery Study Dexamathasone Magann 1994b Control N Mean(SD) N Mean(SD) 12 41.00 (15.00) 13 15.00 (4.50) Weighted Mean Difference (Fixed) Weighted Mean Difference (Fixed) 95% CI 95% CI 26.00 [ 17.17, 34.83 ] -100.0 -50.0 Favour control Analysis 02.01 Review: 50.0 100.0 Favour Dexamethasone Comparison 02 Dexamethasone versus betamethasone, Outcome 01 Mean arterial pressure: adjusted time-averaged change from baseline Corticosteroids for HELLP syndrome in pregnancy Comparison: 02 Dexamethasone versus betamethasone Outcome: 01 Mean arterial pressure: adjusted time-averaged change from baseline Study Isler 2001 Total (95% CI) Dexamethasone Betamethasone N Mean(SD) N Mean(SD) 19 -15.60 (1.40) 21 -8.10 (1.40) 19 Weighted Mean Difference (Fixed) Weight Weighted Mean Difference (Fixed) 95% CI (%) 95% CI 21 100.0 -7.50 [ -8.37, -6.63 ] 100.0 -7.50 [ -8.37, -6.63 ] Test for heterogeneity: not applicable Test for overall effect z=16.92 p