Home > Health & Research > Health Education Campaigns & Programs > Cochrane Neonatal Review > Phenobarbital prior to preterm birth for preventing neonatal periventricular haemorrhage

Phenobarbital prior to preterm birth for preventing neonatal periventricular haemorrhage

Skip sharing on social media links
Share this:

Authors

Caroline A Crowther1, Danielle D Crosby2

Background - Methods - Results - Characteristics of Included Studies - References - Data Tables & Graphs


1Liggins Institute, The University of Auckland, Auckland, New Zealand [top]
2ARCH: Australian Research Centre for Health of Women and Babies, The Robinson Research Institute, Discipline of Obstetrics and Gynaecology, The University of Adelaide, Adelaide, Australia [top]

Citation example: Crowther CA, Crosby DD. Phenobarbital prior to preterm birth for preventing neonatal periventricular haemorrhage. Cochrane Database of Systematic Reviews 2010, Issue 1. Art. No.: CD000164. DOI: 10.1002/14651858.CD000164.pub2.

This is a Cochrane Pregnancy and Childbirth External Web Site Policy systematic review.

Contact person

Caroline A Crowther

Liggins Institute
The University of Auckland
Private Bag 92019
85 Park Road
Auckland
New Zealand

E-mail: c.crowther@auckland.ac.nz
E-mail 2: caroline.crowther@adelaide.edu.au

Dates

Assessed as Up-to-date: 10 January 2011
Date of Search: 20 December 2010
Next Stage Expected: 31 January 2012
Protocol First Published: Issue 1, 1997
Review First Published: Issue 1, 1997
Last Citation Issue: Issue 1, 2010

What's new

Date / Event Description
20 December 2010
Updated

Format updated; search updated, no new trials added.

History

Date / Event Description
11 November 2009
New citation: conclusions not changed

New author helped prepare the last update.

14 April 2008
Amended

Converted to new review format.

30 March 2008
Updated

Search updated. Thorp 2003 study found which is a follow-up report on the Thorp 1994 trial - this reports developmental outcomes in children at age seven.

19 March 2003
Updated

Search updated. Three new references found relating to existing included trials. No new trials. One publication awaiting assessment which appears to be an additional report of Thorp 1994.

30 September 2000
Updated

Search updated. Three new references found, including one new trial.

[top]

Abstract

Background

Preterm infants are at risk of periventricular haemorrhage (PVH). Phenobarbital might prevent ischaemic injury or reduce fluctuations in blood pressure and blood flow in the brain.

Objectives

To assess the benefits and harms of giving phenobarbital to women at risk of imminent very preterm birth with the primary aim of preventing PVH in the infant.

Search methods

We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (20 December 2010).

Selection criteria

Randomised trials with reported data that compared neonatal and maternal outcomes following prenatal exposure to phenobarbital, with outcomes in controls with or without placebo.

Data collection and analysis

We independently assessed trial eligibility and quality and extracted data. We included eligible trials in the initial analysis and prespecified sensitivity analyses to evaluate the effect of trial quality.

Main results

Nine trials (1752 women) were included. Analyses of all included trials showed a significant reduction in the rates of all grades of PVH (risk ratio (RR) 0.65, 95% confidence interval (CI) 0.50 to 0.83; nine trials; 1591 women) and severe grades PVH (3 and 4) (RR 0.41, 95% CI 0.20 to 0.85; eight trials; 1527 women) in infants whose mothers had been given prenatal phenobarbital. These results were influenced by trials of poor quality which contributed excessive weight in the analysis due to their higher rates of severe PVH. When only the two higher quality trials were included, these beneficial effects disappeared for all grades of PVH (RR 0.90, 95% CI 0.75 to 1.08; two trials; 945 women), and severe grades of PVH (RR 1.05, 95% CI 0.60 to 1.83; two trials; 945 women).

No difference was found in the incidence of neurodevelopmental abnormalities at paediatric follow up at 18 to 24 months or seven years of age between children born to mothers given prenatal phenobarbital and children not so exposed.

Maternal sedation was more likely in women receiving phenobarbital (RR 2.06, 95% CI 1.79 to 2.37; one trial; 576 women).

Authors' conclusions

The evidence in this review does not support the use of prophylactic maternal phenobarbital administration to prevent PVH in preterm infants or to protect them from neurological disability in childhood. Phenobarbital administration may lead to maternal sedation. If any future trials are carried out, they should measure neurodevelopmental status at follow up.

[top]

Plain language summary

Phenobarbital prior to preterm birth for preventing neonatal periventricular haemorrhage

 

Evidence does not support phenobarbital treatment to women giving birth before 34 weeks to decrease the risk of bleeding into the babies' brains.

Babies born very early (before 34 weeks) are at risk of bleeding in the brain (periventricular haemorrhage). This can be a cause of brain damage that might lead to disability including cerebral palsy. Phenobarbital may prevent injury to the brain by stabilising blood pressure and blood flow in the brain. Possible adverse effects of phenobarbital for the women include drowsiness, gastrointestinal upset and development of a rash.

Nine trials involving 1752 women were included in the review. The trials with low risk of bias found that phenobarbital given to women immediately prior to a very preterm birth did not decrease the risk of bleeding in the brains of the babies. No differences in child development were found on follow up at 18 to 24 months or at seven years.

Maternal sedation was more likely in women receiving phenobarbital. The use of prenatal corticosteroids, known to reduce rates of periventricular haemorrhage, varied between trials and may have influenced findings.

[top]

Background

Description of the condition

Preterm infants born before 34 weeks' gestation have a significant risk of developing brain injury in association with periventricular haemorrhage (PVH). Severe forms of such lesions are strong predictors of later neurodevelopmental abnormalities, including cerebral palsy.

The exact cause of these lesions is uncertain although it has been hypothesised that brain ischaemia followed by haemorrhage, usually arising initially in the subependymal germinal matrix, is the most likely explanation. Fluctuations in blood pressure and cerebral perfusion around birth and during early postnatal adaptation might also be involved. Most haemorrhages are manifest on head ultrasound within 72 hours of birth although the severity may change after this time (Vohr 2000).

Haemorrhages are usually graded for severity:
grade 1 - confined to the germinal matrix;
grade 2 - in the lateral ventricle;
grade 3 - distending the ventricle;
grade 4 - intracerebral.

The latter two grades have been associated with long-term neurological sequelae.

Description of the intervention

Intravenous or oral phenobarbital given to women at risk of preterm birth.

How the intervention might work

Phenobarbital is a potential neuroprotective agent that might act by preventing ischaemic injury or by reducing the fluctuations in blood pressure and cerebral perfusion (Goddard 1987; Wimberley 1982). Phenobarbital has been suggested as a postnatal treatment and has been the subject of another Cochrane review (Whitelaw 2007). As many haemorrhages are thought to originate close to the time of birth, prophylactic prenatal rather than postnatal treatment may be preferable. Possible adverse effects for the women of phenobarbital include drowsiness, gastrointestinal upset, and a rash.

For further background and discussion regarding prevention and treatment of PVH, see Horbar 1992.

Why it is important to do this review

PVH has severe effects on the infant's brain and subsequent development. There is a need to identify the most safe and efficacious method to prevent PVH in preterm infants to improve mortality and neurodevelopmental outcomes.

[top]

Objectives

To assess the benefits and harms of phenobarbital administered to women at risk of imminent very preterm birth with the primary aim of preventing periventricular haemorrhage in the infant.

[top]

Methods

Criteria for considering studies for this review

Types of studies

All published, unpublished and ongoing randomised and quasi-randomised trials with reported data which compared outcomes for women who were given phenobarbital with outcomes for women who were given a different treatment, placebo or no treatment.

Types of participants

Women at risk of imminent very preterm birth (before 34 weeks' gestation).

Types of interventions

Phenobarbital administered to women intravenously or orally before birth with outcomes in controls with or without placebo.

Types of outcome measures

Primary outcomes
  • Neonatal mortality;
  • early infant death (within three days of life);
  • neonatal neurological and other morbidity;
  • long-term neurodevelopment;
  • maternal morbidity.

Search methods for identification of studies

Electronic searches

We searched the Cochrane Pregnancy and Childbirth Group’s Trials Register by contacting the Trials Search Co-ordinator (20 December 2010). 

The Cochrane Pregnancy and Childbirth Group’s Trials Register is maintained by the Trials Search Co-ordinator and contains trials identified from: 

  1. quarterly searches of the Cochrane Central Register of Controlled Trials (CENTRAL);
  2. weekly searches of MEDLINE;
  3. handsearches of 30 journals and the proceedings of major conferences;
  4. weekly current awareness alerts for a further 44 journals plus monthly BioMed Central email alerts. 

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 ‘Specialized Register’ section within the editorial information about the Cochrane Pregnancy and Childbirth Group External Web Site Policy

Trials identified through the searching activities described above are each assigned to a review topic (or topics). The Trials Search Co-ordinator searches the register for each review using the topic list rather than keywords. 

Searching other resources

We also searched the citation lists of relevant publications, review articles and included studies.

We did not apply any language restrictions.

Data collection and analysis

Selection of studies

Two review authors independently assessed for inclusion all the potential studies we identified as a result of the search strategy. We resolved any disagreement through discussion or, if required, we consulted a third person. We assessed included studies for quality and methodological details without consideration of the results.

Data extraction and management

We designed a form to extract data. For eligible studies, two review authors independently extracted the data using the agreed form. We resolved discrepancies through discussion or, if required, we consulted a third person. There was no blinding of authorship. We entered data into Review Manager software (RevMan 2008) and checked for accuracy.

When information regarding any of the above was unclear, we attempted to contact authors of the original reports to provide further details.

Assessment of risk of bias in included studies

Two review authors independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2009). Any disagreement was resolved by discussion or by involving a third assessor.

(1) Sequence generation (checking for possible selection bias)

We described for each included study the methods used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should have produced comparable groups.

We assessed the methods as:

  • adequate (any truly random process, e.g. random number table; computer random number generator);
  • inadequate (any non-random process, e.g. odd or even date of birth; hospital or clinic record number); or
  • unclear.
(2) Allocation concealment (checking for possible selection bias)

We described for each included study the method used to conceal the allocation sequence in sufficient detail and determined whether intervention allocation could have been foreseen in advance of, or during recruitment or changed after assignment.

We assessed the methods as:

  • adequate (e.g. telephone or central randomisation; consecutively numbered sealed opaque envelopes);
  • inadequate (open random allocation; unsealed or non-opaque envelopes, alternation; date of birth);
  • unclear.
(3) Blinding (checking for possible performance bias)

We described for each included study all the methods 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:

  • adequate, inadequate or unclear for participants;
  • adequate, inadequate or unclear for personnel;
  • adequate inadequate or unclear for outcome assessors.
(4) Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations)

We described for each included study, and for each outcome or class of outcomes, 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, or could be supplied by the trial authors, we re-included missing data in the analyses which we undertook.

We assessed methods as:

  • adequate (e.g. where there was no missing data or where reasons for missing data are balanced across groups);
  • inadequate (e.g. where missing data are likely to be related to outcomes or are not balanced across groups);
  • unclear (e.g. where there is insufficient reporting of attrition or exclusions to permit a judgement to me made).
(5) Selective reporting bias

We described for each included study how we investigated the possibility of selective outcome reporting bias and what we found.

We assessed the methods as:

  • adequate (where it is clear that all of the study’s pre-specified outcomes and all expected outcomes of interest to the review have been reported);
  • inadequate (where not all the study’s pre-specified outcomes have been reported; one or more reported primary outcomes were not pre-specified; outcomes of interest are reported incompletely and so cannot be used; study fails to include results of a key outcome that would have been expected to have been reported);
  • unclear.
(6) Other sources of bias

We described for each included study any important concerns we had about other possible sources of bias. For example, was there a potential source of bias related to the specific study design? Was the trial stopped early due to some data-dependent process? Was there extreme baseline imbalance? Has the study been claimed to be fraudulent?

We assessed whether each study was free of other problems that could put it at risk of bias:

  • yes;
  • no;
  • unclear.
(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 Handbook (Higgins 2009). With reference to (1) and (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 through undertaking sensitivity analyses - see Sensitivity analysis.

Measures of treatment effect

We carried out statistical analysis using the Review Manager software (RevMan 2008). We used fixed-effect meta-analysis for combining data in the absence of significant heterogeneity if trials were sufficiently similar. If significant heterogeneity was found, we explored this by sensitivity analysis followed by random-effects if required.

Dichotomous data

For dichotomous data, we presented results as summary risk ratio with 95% confidence intervals.

Continuous data

For continuous data, we used the mean difference if outcomes were measured in the same way between trials. We used the standardised mean difference to combine trials that measured the same outcome, but used different methods. If there was evidence of skewness, this was reported.

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 by using sensitivity analysis.

For all outcomes, we carried out analyses, as far as possible, on an intention-to-treat basis, i.e. we attempted to include 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.

Where data were not reported for some outcomes or groups, we attempted to contact the study authors for further information.

Assessment of heterogeneity

We applied tests of heterogeneity between trials, if appropriate, using the I² statistic. With no significant heterogeneity, we pooled data using a fixed-effect model. If we identified high levels of heterogeneity among the trials (I² exceeding 50%), we explored it by prespecified subgroup analysis and performed sensitivity analysis. We used a random-effects meta-analysis as an overall summary if this was considered appropriate.

Assessment of reporting biases

Where we suspected reporting bias (see 'Selective reporting bias' above), we attempted to contact study authors asking them to provide missing outcome data. Where this was not possible, and the missing data were thought to introduce serious bias, the impact of including such studies in the overall assessment of results was explored by a sensitivity analysis. 

Where we suspected publication bias (e.g. where only statistically significant results are reported), this was explored using funnel plots (Higgins 2009). We involved the project statistician in the interpretation of such analysis.

Data synthesis

We carried out statistical analysis using the Review Manager software (RevMan 2008). We used fixed-effect meta-analysis for combining data where it was reasonable to assume that studies are estimating the same underlying treatment effect: i.e. where trials were examining the same intervention, and the trials' populations and methods were judged sufficiently similar. If there was clinical heterogeneity sufficient to expect that the underlying treatment effects differed between trials, or if substantial statistical heterogeneity was detected, we used random-effects meta-analysis to produce an overall summary if an average treatment effect across trials was considered clinically meaningful. The random-effects summary was treated as the average range of possible treatment effects and we discussed the clinical implications of treatment effects differing between trials. If the average treatment effect was not clinically meaningful, we did not combine trials.

If we used random-effects analyses, the results were presented as the average treatment effect with its 95% confidence interval, and the estimates of I².

If appropriate, we synthesised data from studies where results are expressed as dichotomous and continuous data.  We involved the project statistician before attempting to synthesise different measures of treatment effect.

Subgroup analysis and investigation of heterogeneity

We planned subgroup analyses based on gestational age or birthweight strata, or both, but were not performed because of insufficient data.

We conducted the following planned subgroup analyses classifying whole trials by interaction tests as described by Deeks 2001:

  1. gestational age;
  2. birthweight;
  3. gestational age and birthweight.

We were unable to perform the planned subgroup analyses due to insufficient data.

Sensitivity analysis

If we identified high levels of heterogeneity among the trials (I² exceeding 50%), we explored it by prespecified subgroup analysis and performed sensitivity analysis. A random-effects meta-analysis was used as an overall summary if this was considered appropriate. This involved analysis based on the rating of selection bias and attrition bias. This was done by excluding trials given an inadequate rating in the quality assessment for allocation of concealment and then inadequate (20% or more or unclear) for exclusions. We did post hoc sensitivity analysis by removing the trial with dual intervention (Thorp 1994) in which both phenobarbital and vitamin K were compared with placebo control.

[top]

Results

Description of studies

Results of the search

We identified nine trials of prenatal phenobarbital for the prevention of periventricular haemorrhage (PVH) of which all nine met our inclusion criteria. One thousand seven hundred and fifty-two women were recruited into the trials.

Included studies

The nine included randomised controlled trials (RCTs) involved 1752 women (Detroit 1 1986; Detroit 2 1990; Detroit 3 1996; Italy 1988; Kaempf 1990; Mexico 1998; Morales 1986; Rayburn 1986; Thorp 1994). Some of the RCTs involved more than one intervention. Four trials (Detroit 1 1986; Detroit 3 1996; Morales 1986; Thorp 1994) used intravenous (IV) followed by oral phenobarbital treatment. Four trials used IV phenobarbital only (Detroit 2 1990, Italy 1988; Kaempf 1990; Mexico 1998). One study only used oral phenobarbital (Rayburn 1986). One study (Thorp 1994) used both vitamin K and phenobarbital as the treatment.

Details of each study are given in the Characteristics of included studies table.

Excluded studies

No studies identified from the search were excluded from the analysis.

Risk of bias in included studies

Allocation (selection bias)

All except four of the trials (Detroit 1 1986; Detroit 2 1990; Italy 1988; Morales 1986) used formal randomisation, although the degree of concealment was uncertain in many of the studies (Kaempf 1990; Mexico 1998; Rayburn 1986). For Detroit 3 1996 and Thorp 1994, randomisation was in the pharmacy. Allocation concealment was not met in the Morales 1986 trial as they used quasi-randomisation utilising the last digit of the medical record number. In this trial, imbalance in prognostic factors, such as corticosteroid use favouring the treatment group, also suggested non-randomisation.

Blinding (performance bias and detection bias)

Only the Detroit 3 1996, Kaempf 1990, Mexico 1998 and Thorp 1994 trials were adequately blinded. In the Rayburn 1986 trial, although a placebo was used, this was unblinded at birth to determine the dose of phenobarbital which was given to all neonates. A placebo was not used in the Detroit 1 1986, Detroit 2 1990, Italy 1988, and Morales 1986 trials and were therefore not adequately blinded.

All trials report that the assessment of the primary outcome of PVH was blinded, apart from Rayburn 1986 where this is unclear.

Incomplete outcome data (attrition bias)

The rate of exclusion of women after randomisation was high in many of the studies (Detroit 1 1986, 27%; Detroit 2 1990, 31%; Detroit 3 1996, 12%; Kaempf 1990, 29%; Mexico 1998, 12%; Rayburn 1986, 49%; Thorp 1994, 10%). Exclusions were not stated for Morales 1986 and Italy 1988. Furthermore, in one trial (Detroit 2 1990), the rate of exclusion after randomisation was twice as high in the treatment compared with the control group. The main reason for postrandomisation exclusion was continuation of the pregnancy to 34 weeks and beyond.

The proportion of children lost to follow up at 18 to 36 months for the three trials that reported data were high (Detroit 2 1990, 40 of 136 children randomised, 29%: five deaths, one child diagnosed with trisomy 21, 25 were not assessed at 36 months; Thorp 1994, 293 of 414 children, 71%: 39 deaths, 254 declined 24 month follow up; Detroit 3 1996, 232 of 668 children, 35%: 46 children were born at greater than/or equal to 34 weeks' gestation, 51 deaths, 135 did not participate). The proportion of children lost to follow up at seven years for the Thorp 1994 trial was lower (115 of 414 children, 28%: 42 deaths, 73 children not followed up).

See Figure 1 for review authors' judgements about each methodological quality item presented as percentages across all included studies and Figure 2 for review authors' judgements about each methodological quality item for each included study.

Selective reporting (reporting bias)

All included studies were free from selective reporting.

Other potential sources of bias

The Italy 1988 study had more male infants in the treatment group (15) compared to the control group (5). In one study (Morales 1986), more women in the phenobarbital group (55%) were treated with corticosteroids antenatally compared to the control group (37%). It is uncertain if there is bias in the Rayburn 1986 study as the number of participants in the abstract are different to those reported in the published article. All other trials were free from other sources of bias (Detroit 1 1986; Detroit 2 1990; Detroit 3 1996; Kaempf 1990; Mexico 1998; Thorp 1994).

Effects of interventions

We identified nine trials involving 1752 women.

1. Analysis of all trials

Overall there was a significant reduction in the incidence of all grades of PVH (risk ratio (RR) 0.65, 95% confidence interval (CI) 0.50 to 0.83; nine trials; 1591 infants) and in severe grades of PVH (3 and 4) (RR 0.41, 95% CI 0.20 to 0.85; eight trials; 1527 infants) in infants exposed to prenatal phenobarbital using a random-effects model because of significant heterogeneity in the data. This result should be viewed with caution due to the high percentage of postrandomisation exclusions in many of the included studies. The Morales 1986 trial, the one quasi-randomised trial that did not have adequate allocation concealment, strongly influenced that result. When the meta-analysis was plotted with trials in chronological order, there was an apparent reduction in treatment effect with time apart for Mexico 1998. The largest study (Detroit 3 1996), contributed half of the infants and was assessed by the review authors as having low bias risk. This study found no effect of phenobarbital on the risk of PVH.

There was a significant increase in infant deaths

discharge (RR 3.68, 95% CI 1.14 to 11.87; three trials; 1080 infants) in infants exposed to antenatal phenobarbital compared with those not so exposed but no such difference in stillbirths or death of live born infants prior to discharge between the phenobarbital and control groups. We observed non-significant trends towards a reduction in the combined outcomes of 'early infant death or severe PVH' and 'death before discharge or severe PVH' for infants exposed to prenatal phenobarbital. No differences were seen in other measures of neonatal morbidity.

Three trials reported data on neurodevelopment at paediatric follow up (Detroit 2 1990; Detroit 3 1996; Thorp 1994). There was no significant difference in Bayley Mental Developmental Index (MDI) scores or in the Bayley Psychomotor Developmental Index (PDI) scores between children exposed to phenobarbital and those not so exposed (Detroit 3 1996; Thorp 1994). For the Thorp 1994 trial, losses to follow up at two years were 71% and results should be interpreted cautiously. There was a discrepancy between the two published reports on childhood follow up at two years (Thorp 1997; Thorp 1999). In Thorp 1999, at two years of age, children in the phenobarbital group had significantly lower Bayley MDI scores than children in the control group (mean MDI score 104, standard deviation 21 compared with mean MDI score 113, standard deviation 22). For the Thorp 1994 trial, losses to follow up at seven years were 28% and no significant difference was found between children in the treatment and control groups in developmental outcomes at seven years of age (Thorp 2003).

In the Detroit 2 1990 and Detroit 3 1996 trials, where the loss to follow-up rates at 18 to 36 month follow up were 29% and 35% respectively, no difference was found in the incidence of neurodevelopmental abnormalities at paediatric follow up at 18 to 36 months of age.

Maternal sedation was more likely in women receiving phenobarbital (RR 2.06, 95% CI 1.79 to 2.37; one trial; 576 women).

2. Pre-specified analyses based on quality of allocation concealment: analysis with the exclusion of the trial not meeting allocation concealment criteria (Morales 1986)

There was still a significant reduction in the risk of all grades of PVH (RR 0.76, 95% CI 0.65 to 0.89; eight trials; 1441 women) but not in severe PVH (RR 0.44, 95% CI 0.19 to 1.01; eight trials; 1379 women), using a random-effects model because of substantial heterogeneity (I² = 52%).

There was still a significant increase in infant deaths postdischarge (RR 3.68, 95% CI 1.14 to 11.87; three trials; 1080 infants) in infants exposed to antenatal phenobarbital compared with those not so exposed but no differences in stillbirths, death of live born infants prior to discharge or measures of neonatal morbidity. In the Detroit 2 1990 and Detroit 3 1996 trials no difference was found in the incidence of neurodevelopmental abnormalities at paediatric follow up at 18 to 36 months of age. There was no significant difference in Bayley MDI scores or in the Bayley PDI scores between children exposed to phenobarbital and those not so exposed (Detroit 3 1996; Thorp 1994). In the Thorp 1994 trial, significantly lower Bayley MDI scores from children in the phenobarbital group compared with the control group were reported at two-year paediatric follow up (Thorp 1997; Thorp 1999). For the Thorp 1994 trial, at seven years of age no differences were found between children in the treatment and control groups (Thorp 2003).

The increased risk of maternal sedation remained for women given prenatal phenobarbital (RR 2.06, 95% CI 1.79 to 2.37; one trial; 576 women).

3. Pre-specified analyses based on high-quality trials

This analysis excluded trials with inadequate allocation concealment (Detroit 1 1986; Detroit 2 1990; Italy 1988; Kaempf 1990; Mexico 1998; Morales 1986; Rayburn 1986) and trials known to have 20% or more postrandomisation exclusions for primary (PVH) outcome (Detroit 1 1986; Detroit 2 1990; Kaempf 1990; Rayburn 1986) or where the rate of postrandomisation exclusions was unclear (Italy 1988).

These analyses included the trials of Thorp 1994 and Detroit 3 1996, assessed as of higher methodological quality. Both trials were rated A for allocation concealment and had less than 20% exclusions for the primary outcome of PVH (Detroit 3 1996, 12%; Thorp 1994, 10% postrandomisation exclusions). However, both these trials had exclusions of over 20% for their childhood follow up and so the data for these assessments have not been included in this analysis.

No significant differences were seen in the risk of all grades of PVH (RR 0.90, 95% CI 0.75 to 1.08; two trials; 945 infants) or severe PVH (RR 1.05, 95% CI 0.60 to 1.83; two trials; 945 infants), stillbirths (RR 2.38, 95% CI 0.46 to 12.21; two trials; 1082 infants), death of live born infants prior to discharge (RR 1.42, 95% CI 0.91 to 2.20; two trials; 1075 infants) or infant deaths postdischarge (RR 3.25, 95% CI 0.80 to 13.19; two trials; 953 infants). Similarly, no differences were seen in other measures of neonatal morbidity.

The increased risk of maternal sedation remained for women given prenatal phenobarbital (RR 2.06, 95% CI 1.79 to 2.37; one trial; 576 women).

[top]

Discussion

Summary of main results

The apparent benefit of prenatal phenobarbital in preventing periventricular haemorrhage (PVH), shown in earlier studies and in the overall analysis is unlikely to be a true effect of the drug. This is suggested by the disappearance of the effect when only high-quality trials are examined. In addition, there has been an increase in the rate of prenatal corticosteroid use, which is known to reduce rates of PVH (Roberts 2006). Rates of prenatal corticosteroid use were around 30% in the earlier studies and have increased to 100% (Thorp 1994) and 60% (Detroit 3 1996) in some of the more recent trials, although were only 6% in the most recent trial reported (Mexico 1998). Phenobarbital may be less effective if corticosteroids have been given.

The Thorp 1994 trial was potentially confounded by the use of vitamin K as a co-intervention with phenobarbital. Since vitamin K alone has not been shown to affect the rate of PVH (Crowther 2010a), this is unlikely to confound the result.

Overall completeness and applicability of evidence

No data are available at present on the effects of phenobarbital when given to women prior to preterm birth at different gestational ages or birthweight strata. It is not recommended that phenobarbital be used in routine practice due to the poor quality of the current evidence and the questionable efficacy of phenobarbital in preventing PVH in the high-quality trial. Also only one trial (Detroit 3 1996) reported on the incidence of maternal sedation. Materal side effects must be assessed more completely before phenobarbital is accepted into routine practice.

Quality of the evidence

Poor-quality trials contribute excessively to the weight in the overall analysis due to the higher rate of adverse outcomes in those trials. Others have criticised the poor quality of trials assessing prenatal phenobarbital (Doyle 1996). The follow-up results of neurodevelopmental outcome are important, yet the results should be viewed with caution since the overall rate of follow up was suboptimal for all trials reporting data (losses to follow up: Detroit 2 1990, 24% of infants at 36 months; Thorp 1994, 71% at two years, and 28% at seven years; Detroit 3 1996, 35% at 18 to 22 months). In addition, there was an imbalance between the treatment and control groups in the ascertainment of women at risk in Detroit 2 1990.

Potential biases in the review process

The largest trial (Thorp 1994), which was also the best quality trial, had vitamin K as a co-treatment with phenobarbital. Evidence from the systematic review of trials evaluating prenatal vitamin K prior to preterm birth (Crowther 2010a) suggests that vitamin K does not influence the rate of PVH.

[top]

Authors' conclusions

Implications for practice

The evidence in this review does not support the use of prophylactic maternal phenobarbital administration to prevent periventricular haemorrhage in preterm infants or to protect from childhood neurological disability. Phenobarbital administration leads to maternal sedation.

Implications for research

If any future trials are carried out, they should be of high quality, examine the effects of prenatal phenobarbital at gestational ages with a high risk of periventricular haemorrhage, stratify for gestational age and ensure minimal exclusions after randomisation. Neurodevelopmental status at follow up should be measured as the most important outcome.

[top]

Acknowledgements

Professor Adrian Grant compiled the first version of this review. Professor Lex Doyle provided helpful comment on the 1997 version of this review.

Special thanks to Lynn Hampson who performed the literature search for this update and to Philippa Middleton who commented on all drafts of this update.

We acknowledge the significant contributions of Professor David Henderson-Smart to the development of the original protocol, identification and selection of trials for inclusion, data extraction and preparation of the text of the initial review and previous updates.

[top]

Contributions of authors

CA Crowther (CAC) and D Henderson-Smart (DHS) contributed to the development of the original protocol, identification and selection of trials for inclusion, data extraction and prepared the update in 2003.

For the 2008 update, CAC and DD Crosby (DDC) assessed the new studies for inclusion and performed the data extraction. DDC drafted the changes to the text and prepared subsequent drafts. CAC and DHS contributed to the drafts.

For the 2011 update, DDC drafted the changes to the text and prepared subsequent drafts. CAC contributed to the drafts and final version.

[top]

Declarations of interest

None known.

[top]

Differences between protocol and review

[top]

Published notes

[top]

Characteristics of studies

Characteristics of included studies

Detroit 1 1986

Methods

Randomised controlled trial.

Participants

63 women < 35 weeks' gestation in threatened preterm labour with a singleton or twin pregnancy.
Excluded if barbiturate allergy.

Interventions

500 mg phenobarbital intravenously over 30 minutes, then 100 mg orally every 24 hours (n = 31). The control group received no treatment (n = 32).

Outcomes

Primary outcome: IVH.
Other: mortality.

Notes

No information on antenatal steroid usage.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) High risk

Randomised using a card deck.

Allocation concealment (selection bias) Unclear risk

Not reported.

Blinding (performance bias and detection bias) High risk

No placebo used however the assessment of the primary outcome of IVH was blinded.

Incomplete outcome data (attrition bias) Unclear risk

27% (17/63) of women were lost to follow up due to withdrawing from the study, or continuation of the pregnancy to term. 4% (2/49) infants were excluded due to death within 24 hours of birth.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Low risk  

Detroit 2 1990

Methods

Randomised clinical trial.

Participants

160 women in preterm labour at < 35 weeks with or without ruptured membranes.

Interventions

10 mg/kg (maximum of 100 mg) phenobarbital intravenously over 30 minutes, then 100 mg daily until delivery or 34 weeks (n = 84). The control group received no treatment (n = 76).

Outcomes

Primary outcome: intracranial haemorrhage.
Outcomes of follow up at 24-36 months: physical growth, neurologic and developmental outcome.

Notes

Sample size calculation done.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) High risk

Randomly assigned using a card deck.

Allocation concealment (selection bias) Unclear risk

Not reported.

Blinding (performance bias and detection bias) High risk

No placebo was used however the assessment of the primary outcome was blinded.

Incomplete outcome data (attrition bias) High risk

34 of 84 (40%) women randomised into the phenobarbital group were excluded due to cessation of labour, continuation of pregnancy beyond 34 weeks and incomplete infusion of the drug before delivery. However, only 16 of 76 (21%) women randomised to the control group were excluded for those reasons. 21/62 (34%) infants in the phenobarbital group and 19/74 (26%) infants in the control group were excluded from the follow up at 36 months (in total 40/136 (29%) infants at randomisation were excluded). Reasons for exclusion included death postdischarge, failure to keep at least 4 follow up appointments, diagnosis of trisomy and failure to keep the 36 month visit.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Low risk  

Detroit 3 1996

Methods

Randomised controlled trial.

Participants

610 women 24-33 weeks' gestation who were expected to give birth within 24 hours.

Interventions

10 mg/kg intravenously phenobarbital (n = 309) then, 100 mg phenobarbital or placebo orally until delivery or > 32 weeks or placebo (n = 301).

Outcomes

Primary outcome: intracranial haemorrhage, (only ascertained for liveborn babies born at < 34 weeks (311 in intervention group and 279 in control group), or early death (< 72 hours).
Secondary outcomes: maternal side effects, intracranial haemorrhage, periventricular leucomalacia, infant neurodevelopment. Outcomes of follow up at 18-22 months (corrected age): physical growth, neurologic and developmental outcomes.

Notes

Approximately 60% antenatal corticosteroid usage in each group. Trial stopped after 610 women recruited (of a calculated sample size of 1038) as probability of showing a statistically significant treatment effect was unlikely.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk

Randomly assigned by a pharmacist.

Allocation concealment (selection bias) Unclear risk  
Blinding (performance bias and detection bias) Low risk

Placebo used. Blinded primary outcome assessment.

Incomplete outcome data (attrition bias) Low risk

35/309 (11%) women in the phenobarbital group and 40/301 (13%) women in the placebo group were excluded postrandomisation due to continuation of the pregnancy beyond 33 weeks and delivery prior to initiation of the treatment (12% excluded).

Follow up at 18-22 months excluded 118/344 (34%) infants in the phenobarbital group and 114/324 (35%) infants in the placebo group due to delivery after 34 weeks, death before and after discharge from the neonatal intensive care unit, and failure to attend the 18 to 22 month follow-up visit (in total 232/668 (35%) infants at randomisation were excluded).

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Low risk  

Italy 1988

Methods

Randomised trial.

Participants

39 women at risk of imminent risk of preterm birth before 32 weeks' gestation. Excluded: congenital abnormalities, maternal allergy to phenobarb, maternal liver, kidney or renal disease, neonatal birthweight < 500 g.

Interventions

Phenobarbital IV 2 doses of 350 mg 10 minutes apart, each given over 10 minutes; if not born in 24 hours 250 mg/day in 2 doses (n = 21). Control group received no treatment (n = 18).

Outcomes

PVH, mortality, other neonatal morbidity.

Notes

Requires author clarification re exclusions, antenatal corticosteroid usage.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) High risk

Randomised using a card deck.

Allocation concealment (selection bias) Unclear risk

Not reported.

Blinding (performance bias and detection bias) High risk

No placebo was used however the primary outcome of PVH was blinded.

Incomplete outcome data (attrition bias) Unclear risk

No information on exclusions.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias High risk

The treatment group included 15 male infants whereas the control group only had 5 males.

Kaempf 1990

Methods

Randomised controlled trial.

Participants

151 women < 31 weeks' gestation in preterm labour or with preterm prelabour rupture of the membranes, or maternal or fetal complication requiring delivery.
Excluded if on barbiturates or cardio-respiratory disease, or porphyria.

Interventions

10 mg/kg phenobarbital intravenously (minimum 500 mg maximum 700 mg) over 30 minutes (n = 60).
Placebo involved 0.9% saline in equal volume over 30 minutes (n = 50).

Outcomes

Primary outcome: IVH.
Other: neonatal mortality, respiratory distress syndrome, Apgar scores.

Notes

Postnatal phenobarbital given if ventilated (91% phenobarbital group, 75% placebo group).

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk

Randomly assigned to either the treatment or control group by use of a computer-generated random number list.

Allocation concealment (selection bias) Unclear risk

Unclear if number list was concealed.

Blinding (performance bias and detection bias) Low risk

A placebo was used. The patient, obstetric, neonatal, nursing and radiology staff were unaware of the treatment allocation. The assessment of the primary outcome was blinded.

Incomplete outcome data (attrition bias) High risk

37% (32/86) of infants randomised to the phenobarbital group were lost to follow up, whereas only 20% (17/84) of those randomised to the placebo group were lost to follow up. Reasons for exclusion included incomplete infusion, delivery after 31 weeks' gestation and neonatal death prior within 24 hours of birth.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Low risk  

Mexico 1998

Methods

Randomised controlled trial.

Participants

100 women with a singleton pregnancy at 28-32 weeks' gestation at risk of preterm birth.
Excluded if taking 'hemostasia' drugs or if they have an autoimmune disease.

Interventions

Those in the phenobarbital group given phenobarbital 10 ug/kg diluted in 100 ul of 5% dextrose solution via a 30-minute infusion (n = 42). Those in the control group given "diluted distilled water dosaged at the same time" (n = 46). Both given at least 4 hours and more than 24 hours before the caesarean section.

Outcomes

Primary outcome: PVH.
Secondary outcome: neonatal mortality and morbidity.

Notes

Antenatal steroids only given to 5/88 (6%) women.
Reasons for exclusions: stillbirth (1), major congenital malformation (1), no head ultrasound (2), moved hospital (1), technical difficulties in blood sampling (7).
No sample size calculation given.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk

Not reported.

Allocation concealment (selection bias) Unclear risk

Not reported.

Blinding (performance bias and detection bias) Low risk

A placebo was used and the assessment of the primary outcome was blinded.

Incomplete outcome data (attrition bias) Unclear risk

12% (12/100) were lost to follow up.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Low risk  

Morales 1986

Methods

Randomised trial.

Participants

150 women < 32 weeks at imminent risk of delivery (active labour not responding to tocolysis, preterm prelabour rupture of the membranes, or maternal complication).
Consent obtained only from treatment group.

Interventions

390 mg intravenously phenobarbital if in active labour, if not in labour 720 mg in 3 oral doses every 4 hours, then 60 mg orally every 6 hours (n = 75). Control group given no treatment (n = 75).

Outcomes

Primary outcome: IVH.
Other: respiratory distress syndrome, Apgar scores, need for resuscitation.

Notes

Antenatal steroids given "whenever possible".
Phenobarbital used postnatally to all babies (2.5 mg/kg every 12 hours for 4 days).

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) High risk

Randomly assigned by the last digit of the hospital chart number.

Allocation concealment (selection bias) High risk  
Blinding (performance bias and detection bias) High risk

No placebo was used however the assessment of the primary outcome was blinded.

Incomplete outcome data (attrition bias) Unclear risk

No exclusions specified.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias High risk

55% (41/75) of patients in the phenobarbital group were given corticosteroids compared with 37% (28/75) in the control group.

Rayburn 1986

Methods

Randomised controlled trial.

Participants

126 women prior to anticipated preterm delivery 26-34 weeks' gestation.
Excluded if on anticonvulsants, or if drug abusers, or major congenital abnormality of fetus.

Interventions

Phenobarbital (n = 31) or placebo (n = 33) 30 mg 3 times a day or 90 mg once daily at least 24 hours before birth.
All infants given a 6-day maintenance therapy of phenobarbital. Those in the placebo group were given a loading dose of 20 mg/kg.

Outcomes

Primary outcome: IVH.

Notes

Reasons for exclusions postrandomisation: gestational age > 34 weeks, delivered < 24 hours from entry, stopped medication, withdrew.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk

Randomly selected and assigned.

Allocation concealment (selection bias) Unclear risk

Not reported.

Blinding (performance bias and detection bias) Unclear risk

Placebo used. Unknown if primary outcome was blinded although unblinding likely in view of differential treatment of neonates.

Incomplete outcome data (attrition bias) Unclear risk

49% (62/126) excluded postrandomisation.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Unclear risk

Numbers of participants do not match between abstract and main published paper.

Thorp 1994

Methods

Randomised controlled trial stratified by gestational age.

Participants

353 women (414 fetuses) at 24-34 weeks' gestation at risk of spontaneous or indicated preterm birth. All received betamethasone and had ampicillin until known culture negative for GBS.

Interventions

Vitamin K 10 mg intramuscularly. Repeated in 4 days then 20 mg per day orally after 8 days. Phenobarbital 780 mg intravenously over 90 minutes or 720 mg orally in 3 doses over 8 hours, then 60 mg/6 hours intravenously or oral (n = 177).
Versus identical appearing placebo (n = 176).

Outcomes

Primary outcome, PVH; other, neonatal mortality and morbidity.
PVH assessed by cranial ultrasound on day 2-3 and day 7 by a radiologist blinded to treatment group. Secondary outcome: infant neurodevelopment. Developmental outcomes at 2 and 7 year follow up.

Notes

Power calculation based on reduction of all PVH 50% to 25% and severe PVH 20% to 5%, needed 200. Interim analysis expanded the sample size because the frequency of severe IVH was substantially lower in the study compared to previous studies. Stopped trial early because baseline risk lower and no difference found.

Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk

Block randomisation.

Allocation concealment (selection bias) Unclear risk

Randomisation completed by pharmacist.

Blinding (performance bias and detection bias) Low risk

Placebo identical to treatment. Phenobarbital levels not followed in treatment group. Blinded assessment of primary outcome.

Incomplete outcome data (attrition bias) High risk

22/203 (11%) babies from the placebo group and 20/211 (9%) babies from the treatment group were lost to follow up initially because they were delivered at or after 34 weeks' gestation (in total 42/414 (10%) infants were excluded). 141/203 (69%) children from the placebo group and 152/211 (72%) children from the treatment group were excluded at 2 years (in total 293/414 (71%) infants randomised were excluded). Reasons for exclusion were antenatal and postnatal demise and mothers declining to be involved in follow up. 52/203 (26%) children in the placebo group and 63/211 (30%) children in the treatment group were lost to follow up at 7 years (in total 115/414 (28%) children randomised were excluded). Reasons for exclusion included antenatal and postnatal demise, and death since discharge.

Selective reporting (reporting bias) Low risk

No indication of selective reporting.

Other bias Low risk

All patients received betamethasone and ampicillin.

Footnotes

GBS: GBS: group B streptococci
IM: intramuscular
IV: intravenous
IVH: intraventricular haemorrhage
PVH: periventricular haemorrhage

[top]

References to studies

Included studies

Detroit 1 1986

* Shankaran S, Cepeda EE, Ilagan N, Moriana F, Hassan M, Bhatia R et al. Antenatal phenobarbital for the prevention of neonatal intracerebral hemorrhage. American Journal of Obstetrics and Gynecology 1986;154:53-7.

Shankaran S, Ilagan N, Cepeda EE, Mariona F, Bedard MP, Poland RL et al. Antenatal phenobarbital for prevention of intracerebral hemorrhage: preliminary observations. Pediatric Research 1984;18:346A.

Detroit 2 1990

* Shankaran S, Cepeda E, Muran G, Mariona F, Johnson S, Kazzi N et al. Antenatal phenobarbital therapy and neonatal outcome I: Effect on intracranial hemorrhage. Pediatrics 1996;97:644-8.

Shankaran S, Cepeda E, Kazzi N, Ostrea E, Poland R, Bedard M et al. Antenatal phenobarbital administration significantly reduces intracranial hemorrhage in <1250g neonates. Journal of Perinatal Medicine 1990;18:104.

Shankaran S, Cepeda E, Kazzi N, Ostrea E, Poland R, Bedard M et al. Antenatal phenobarbital administration significantly reduces intracranial hemorrhage in <1250g neonates. Pediatric Research 1990;27:225A.

Shankaran S, Woldt E, Nelson J, Bedard M, Delaney-Black V. Antenatal phenobarbital therapy and neonatal outcome II: neurodevelopment outcome at 36 months. Pediatrics 1996;97:649-52.

Detroit 3 1996

* Shankaran S, Papile L, Wright L, Ehrenkranz R, Mele L, Verter J. Neurodevelopmental outcome following antenatal phenobarbital (PB) exposure in a multicenter randomized trial. Pediatric Research 1997;41:177A.

McCain G, Donovan E, Gartside P. Preterm infant behavioral and heart rate responses to antenatal phenobarbital. Research in Nursing and Health 1999;22:461-70.

Shankaran S, Papile L, Wright L, Ehrenkranz R, Mele L, Lemons J et al. Neurodevelopmental outcome of premature infants after antenatal phenobarbital exposure. American Journal of Obstetrics and Gynecology 2002;187:171-7.

Shankaran S, Papile L, Wright L, Ehrenkranz R, Mele L, Lemons J et al. The effect of antenatal phenobarbital therapy on neonatal intracranial haemorrhage in preterm infants. New England Journal of Medicine 1997;337:466-71.

Shankaran S, Papile L, Wright L, Ehrenkranz R, Verter J, Mele L. Does antenatal phenobarbital prevent neonatal intracranial hemorrhage in preterm infants: randomized controlled trial. Pediatric Research 1996;39(4 Pt 2):1452.

Italy 1988

De Carolis S, De Carolis MP, Caruso A, Oliva GC, Romagnoli C, Ferrazzani S et al. Antenatal phenobarbital in preventing intraventricular hemorrhage in premature newborns. Fetal Therapy 1988;3:224-9.

Kaempf 1990

Kaempf JW, Porreco R, Molina R, Hale K, Pantoja AF, Rosenberg AA. Antenatal phenobarbital for the prevention of periventricular and intraventricular hemorrhage: a double blind, randomized, placebo-controlled, multihospital trial. Journal of Pediatrics 1990;117:933-8.

Mexico 1998

Arroyo-Cabrales L, Garza-Morales S, Hernandez-Pelaez G. Use of prenatal phenobarbital in the prevention of subependymal/intraventricular hemorrhage in premature infants. Archives of Medical Research 1998;29:247-51.

Morales 1986

Morales WJ, Koerten J. Prevention of intraventricular hemorrhage in very low birth weight infants by maternally administered phenobarbital. Obstetrics & Gynecology 1986;68:295-9.

Rayburn 1986

* Rayburn W, Donn S, Piehl E, Compton A. Antenatal phenobarbital and bilirubin metabolism in the very low birth weight infant. American Journal of Obstetrics and Gynecology 1988;159:1491-3.

Rayburn W, Donn S, Compton A, Piehl E. Oral phenobarbital given antenatally to reduce neonatal intraventricular hemorrhage, a comparison between maternal and umbilical cord serum levels at delivery. Journal of Perinatology 1989;9:268-70.

Rayburn W, Donn S, Compton A. Maternal phenobarbital therapy and intraventricular hemorrhage in the low birth weight infant.. In: Proceedings of the 6th Annual Meeting of the Society of Perinatal Obstetricians; 1986 Jan 30-Feb 1; San Antonio, Texas, USA. 1986:108.

Thorp 1994

De Lia JE, Billman D. Antepartum vitamin K and phenobarbital for preventing intraventricular hemorrhage in the premature newborn; a randomized, double-blind, placebo-controlled trial [letter]. Obstetrics & Gynecology 1994;83(6):1067-8.

Thorp J, O'Connor M, Belden B, Etzenhouser J, Hoffman E, Jones P. Effects of phenobarbital and multiple-dose corticosteroids on developmental outcome at age 7 years. Obstetrics & Gynecology 2003;101(2):363-73.

* Thorp J, Parriott J, Ferrette-Smith D, Meyer BA, Cohen GR, Johnson J. Antepartum vitamin K and phenobarbital for preventing intraventricular hemorrhage in the premature newborn: a randomized, double-blind, placebo-controlled trial. Obstetrics & Gynecology 1994;83(1):70-6.

Thorp J, Yeast J, Cohen G, Poskin M, Peng V, Hoffman E. Does in-utero phenobarbital lower IQ? Follow up of the intracranial hemorrhage prevention trial. American Journal of Obstetrics and Gynecology 1997;176(1 Pt 2):S117.

Thorp JA, Blandon R, Jones AMH, Brown MJ, Clark R, Etzenhauser J et al. Antenatal/postnatal steroids and developmental outcome at age 2 years. American Journal of Obstetrics and Gynecology 2001;184(1):S50.

Thorp JA, Caspers DR, Cohen GR, Zucker ML, Strope BD, McKenzie DR. The effect of combined antenatal vitamin K and phenobarbital therapy on umbilical blood coagulation studies in infants less than 34 weeks' gestation. Obstetrics & Gynecology 1995;86:982-9.

Thorp JA, Etzenhouser J, OConnor M, Jones A, Jones P, Belden B, Hoffmann E. Effects of phenobarbital and multiple-dose antenatal/postnatal steroid on developmental outcome at age 7 years [abstract]. American Journal of Obstetrics and Gynecology 2001;185(6 Suppl):S87.

Thorp JA, Ferrette-Smith D, Gaston L, Johnson J, Caspers D, Yeast JD et al. Antenatal vitamin K (VK) and phenobarbital (PH) for preventing intracrancial hemorrhage (ICH) in the premature newborn: a randomized double blinded placebo controlled trial. American Journal of Obstetrics and Gynecology 1995;172:253.

Thorp JA, Ferrette-Smith D, Gaston LA, Johnson J, Yeast JD, Meyer B. Combined antenatal vitamin K and phenobarbital therapy for preventing intracranial hemorrhage in newborns less than 34 weeks' gestation. Obstetrics & Gynecology 1995;86:1-8.

Thorp JA, Gaston L, Ferrette-Smith D, Caspers D, Wickstrom E, Pal M et al. Mode of delivery and prediction of severe intracranial hemorrhage (ICH): a randomized double blinded placebo controlled trial. American Journal of Obstetrics and Gynecology 1995;172:289.

Thorp JA, McKenzie DR. Antenatal phenobarbital therapy and neonatal outcome. Pediatrics 1997;99:751-3.

Thorp JA, Neimark M, Poskin M. Maternal oxygen desaturation with intravenous magnesium sulphate therapy. Obstetrics & Gynecology 1997;89(6):963-6.

Thorp JA, O'Connor M, Jones AM, Hoffman EL, Belden B. Does perinatal phenobarbital exposure affect developmental outcome at age 2? American Journal of Perinatology 1999;16(2):51-60.

Thorp JA, Parriott J, Ferrette-Smith D, Holst V, Meyer BA, Cohen GR et al. Antepartum vitamin K and phenobarbital for preventing intraventricular hemorrhage in the premature newborn: a randomized double blinded placebo controlled trial. American Journal of Obstetrics and Gynecology 1993;168:367.

Thorp JA, Zucker M, Strope B, Claflin K, Callenbach JC, Shaffer S et al. The effect of antenatal vitamin K (VK) and phenobarbital (PH) on cord blood coagulation studies at birth: a randomized double blinded placebo controlled trial. American Journal of Obstetrics and Gynecology 1995;172:327.

Excluded studies

None noted.

Studies awaiting classification

None noted.

Ongoing studies

None noted.

[top]

Other references

Additional references

Crowther 2010a

Crowther CA, Crosby DD, Henderson-Smart DJ. Vitamin K prior to preterm birth for preventing neonatal periventricular haemorrhage. Cochrane Database of Systematic Reviews 2010, Issue 1. Art. No.: CD000229. DOI: 10.1002/14651858.CD000229.pub2.

Deeks 2001

Deeks JJ, Altman DG, Bradburn MJ. Statistical methods for examining heterogeneity and combining results from several studies in meta-analysis. In: Egger M, Davey Smith G, Altman DG, editor(s). Systematic reviews in health care: meta-analysis in context. London: BMJ Books, 2001.

Doyle 1996

Doyle L. Antenatal phenobarbitone and neonatal outcome. Lancet 1996;348:975-6.

Goddard 1987

Goddard-Finegold J, Armstrong DL. Reduction in incidence of periventricular intraventricular hemorrhages in hypertensive newborn beagles pretreated with phenobarbital. Pediatrics 1987;79:901-6.

Higgins 2009

Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2 [updated September 2009]. The Cochrane Collaboration, 2009. Available from www.cochrane-handbook.org.

Horbar 1992

Horbar J. Prevention of periventricular-intraventricular haemorrhage. In: Sinclair JC, Bracken MB, editor(s). Effective care of the newborn infant. Oxford: Oxford University Press, 1992.

RevMan 2008

Review Manager (RevMan) [Computer program]. Version 5.0. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2008.

Roberts 2006

Roberts D, Dalziel S. Antenatal corticosteroids for accelerating fetal lung maturation for women at risk of preterm birth. Cochrane Database of Systematic Reviews 2006, Issue 3. Art. No.: CD004454. DOI: 10.1002/14651858.CD004454.pub.

Thorp 1997

Thorp J, Yeast J, Cohen G, Poskin M, Peng V, Hoffman E. Does in-utero phenobarbital lower IQ? Follow up of the intracranial hemorrhage prevention trial. American Journal of Obstetrics and Gynecology 1997;176(1 Pt 2):S117.

Thorp 1999

Thorp JA, O'Connor M, Jones AM, Hoffman EL, Belden B. Does perinatal phenobarbital exposure affect developmental outcome at age 2? American Journal of Perinatology 1999;16(2):51-60.

Thorp 2003

Thorp J, O'Connor M, Belden B, Etzenhouser J, Hoffman E, Jones P. Effects of phenobarbital and multiple-dose corticosteroids on developmental outcome at age 7 years. Obstetrics & Gynecology 2003;101(2):363-73.

Vohr 2000

Vohr BR, Wright LL, Dusick AM, Mele L, Verter J, Steichen JJ et al. Neurodevelopmental and functional outcomes of extremely low birth weight infants in the National Institute of Child Health and Human Developmental Neonatal Research Network 1993-1994. Pediatrics 2000;105:1216-26.

Whitelaw 2007

Whitelaw A, Odd D. Postnatal phenobarbital for the prevention of intraventricular hemorrhage in preterm infants. Cochrane Database of Systematic Reviews 2007, Issue 4. Art. No.: CD001691. DOI: 10.1002/14651858.CD001691.pub2.

Wimberley 1982

Wimberley PD, Lou HC, Hejl M, Lassen NA, Friis-Hansen B. Hypertensive peaks in the pathogenesis of intraventricular haemorrhage in the newborn: abolition by phenobarbitone sedation. Acta Paediatrica Scandinavica 1982;71:537-42.

Other published versions of this review

Crowther 1995

Crowther CA, Alfirevic Z. Phenobarbital prior to preterm delivery. [revised 07 April 1994]. In: Enkin MW, Keirse MJNC, Renfrew MJ, Neilson JP, Crowther C (eds.) Pregnancy and Childbirth Module. In: The Cochrane Pregnancy and Childbirth Database [database on disk and CDROM]. The Cochrane Collaboration; Issue 2, Oxford: Update Software; 1995.

Crowther 1997

Crowther CA, Henderson-Smart DJ. Phenobarbital prior to preterm birth. In: Neilson JP, Crowther CA, Hodnett ED, Hofmeyr GJ, Keirse MJNC (eds). Pregnancy and Childbirth Module of The Cochrane Database of Systematic Reviews, [updated 03 June 1997]. Available in The Cochrane Library [database on disk and CDROM]. The Cochrane Collaboration; Issue 3. Oxford: Update Software; 1997. Updated quarterly.

Crowther 1999

Crowther CA, Henderson-Smart DJ. Phenobarbital prior to preterm birth for preventing neonatal periventricular haemorrhage. Cochrane Database of Systematic Reviews 1999, Issue 2.

Crowther 2003

Crowther CA, Crosby DD, Henderson-Smart DJ. Phenobarbital prior to preterm birth for preventing neonatal periventricular haemorrhage. Cochrane Database of Systematic Reviews 2003, Issue 3. Art. No.: CD000164. DOI: 10.1002/14651858.CD000164.

Classification pending references

None noted.

[top]

Data and analyses

1 Phenobarbital versus control - all studies

Outcome or Subgroup Studies Participants Statistical Method Effect Estimate
1.1 All periventricular haemorrhage (PVH) 9 1591 Risk Ratio (M-H, Random, 95% CI) 0.65 [0.50, 0.83]
1.2 Severe (grades 3 and 4) PVH 8 1527 Risk Ratio (M-H, Random, 95% CI) 0.41 [0.20, 0.85]
1.3 Perinatal mortality 8 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
  1.3.1 Stillbirths 2 1082 Risk Ratio (M-H, Fixed, 95% CI) 2.38 [0.46, 12.21]
  1.3.2 Death of live born infants prior to discharge 8 1667 Risk Ratio (M-H, Fixed, 95% CI) 1.02 [0.76, 1.38]
  1.3.3 Infant deaths post-discharge 3 1080 Risk Ratio (M-H, Fixed, 95% CI) 3.68 [1.14, 11.87]
1.4 Infant death prior to discharge or severe PVH 2 184 Risk Ratio (M-H, Random, 95% CI) 0.40 [0.07, 2.32]
1.5 Early infant death or any PVH 3 849 Risk Ratio (M-H, Fixed, 95% CI) 0.97 [0.78, 1.22]
1.6 Early infant death or severe PVH 2 181 Risk Ratio (M-H, Random, 95% CI) 0.41 [0.07, 2.35]
1.7 Low Apgar score at 5 minutes 5 1297 Risk Ratio (M-H, Fixed, 95% CI) 1.23 [0.96, 1.59]
1.8 Use of mechanical ventilation 6 804 Risk Ratio (M-H, Random, 95% CI) 1.02 [0.86, 1.20]
1.9 Respiratory distress syndrome 3 359 Risk Ratio (M-H, Random, 95% CI) 1.01 [0.71, 1.43]
1.10 Pulmonary air leak 6 818 Risk Ratio (M-H, Fixed, 95% CI) 0.76 [0.45, 1.31]
1.11 Patent ductus arteriosus 4 583 Risk Ratio (M-H, Fixed, 95% CI) 0.95 [0.64, 1.42]
1.12 Cerebral palsy at 18-36 months follow up 2 517 Risk Ratio (M-H, Fixed, 95% CI) 0.71 [0.40, 1.28]
1.13 Cerebral palsy at 7 year follow up 1 299 Risk Ratio (M-H, Fixed, 95% CI) 0.77 [0.33, 1.76]
1.14 Other neuromotor impairment at 36 month follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.67 [0.13, 3.49]
1.15 Developmental delay at 18-22 months follow up 1 401 Risk Ratio (M-H, Fixed, 95% CI) 1.09 [0.89, 1.35]
1.16 Visual impairment at 18-36 months follow up 2 517 Risk Ratio (M-H, Fixed, 95% CI) 1.41 [0.37, 5.35]
1.17 Hearing impairment at 18-36 months follow up 2 517 Risk Ratio (M-H, Fixed, 95% CI) 0.82 [0.31, 2.16]
1.18 Bayley Mental Developmental Index at 18-24 months 2 543 Mean Difference (IV, Random, 95% CI) -4.21 [-11.89, 3.48]
1.19 Bayley Psychomotor Developmental Index at 18-24 months 2 538 Mean Difference (IV, Fixed, 95% CI) -1.82 [-4.71, 1.07]
1.20 Wechsler Intelligence Scale for Children at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  1.20.1 Full scale 1 299 Mean Difference (IV, Fixed, 95% CI) 0.30 [-2.92, 3.52]
  1.20.2 Performance 1 299 Mean Difference (IV, Fixed, 95% CI) 1.20 [-2.23, 4.63]
  1.20.3 Verbal 1 299 Mean Difference (IV, Fixed, 95% CI) -0.60 [-3.84, 2.64]
1.21 Wide Range Achievement Test at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  1.21.1 Reading 1 299 Mean Difference (IV, Fixed, 95% CI) 0.70 [-2.57, 3.97]
  1.21.2 Spelling 1 299 Mean Difference (IV, Fixed, 95% CI) -0.50 [-3.45, 2.45]
  1.21.3 Mathematics 1 299 Mean Difference (IV, Fixed, 95% CI) -1.40 [-4.61, 1.81]
1.22 Achenbach Teacher's Report Form at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  1.22.1 Total 1 299 Mean Difference (IV, Fixed, 95% CI) -1.20 [-3.55, 1.15]
  1.22.2 Internal 1 299 Mean Difference (IV, Fixed, 95% CI) -0.60 [-2.82, 1.62]
  1.22.3 External 1 299 Mean Difference (IV, Fixed, 95% CI) -1.50 [-3.45, 0.45]
1.23 Child Behavioral Check List total at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  1.23.1 Total 1 299 Mean Difference (IV, Fixed, 95% CI) 1.00 [-1.28, 3.28]
  1.23.2 Internal 1 299 Mean Difference (IV, Fixed, 95% CI) 2.00 [-0.10, 4.10]
  1.23.3 External 1 299 Mean Difference (IV, Fixed, 95% CI) -0.60 [-2.92, 1.72]
1.24 Height (cm) at 18-24 months follow up (mean, standard deviation) 1 436 Mean Difference (IV, Fixed, 95% CI) 0.00 [-0.75, 0.75]
1.25 Height < 5th centile at 18-24 month follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.49 [0.17, 1.42]
1.26 Weight (kg) at 18-24 month follow up (mean, standard deviation) 1 436 Mean Difference (IV, Fixed, 95% CI) -0.10 [-0.38, 0.18]
1.27 Weight < 5th centile at 18-24 month follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.22 [0.03, 1.79]
1.28 Head circumference (cm) at childhood follow up (mean, standard deviation) 1 436 Mean Difference (IV, Fixed, 95% CI) -0.20 [-0.53, 0.13]
1.29 Head circumference < 5th centile at childhood follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.67 [0.18, 2.52]
1.30 Maternal sedation 1 576 Risk Ratio (M-H, Fixed, 95% CI) 2.06 [1.79, 2.37]
 

2 Phenobarbital versus control - excluding trial with non-concealment at randomisation (C quality)

Outcome or Subgroup Studies Participants Statistical Method Effect Estimate
2.1 All periventricular haemorrhage (PVH) 8 1441 Risk Ratio (M-H, Fixed, 95% CI) 0.76 [0.65, 0.89]
2.2 Severe (grades 3 and 4) PVH 8 1379 Risk Ratio (M-H, Random, 95% CI) 0.44 [0.19, 1.01]
2.3 Perinatal mortality 7 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
  2.3.1 Stillbirths 2 1082 Risk Ratio (M-H, Fixed, 95% CI) 2.38 [0.46, 12.21]
  2.3.2 Death of live born infants prior to discharge 7 1517 Risk Ratio (M-H, Fixed, 95% CI) 1.13 [0.83, 1.55]
  2.3.3 Infant deaths post-discharge 3 1080 Risk Ratio (M-H, Fixed, 95% CI) 3.68 [1.14, 11.87]
2.4 Infant death prior to discharge or severe PVH 2 184 Risk Ratio (M-H, Random, 95% CI) 0.40 [0.07, 2.32]
2.5 Early infant death or any PVH 3 849 Risk Ratio (M-H, Fixed, 95% CI) 0.97 [0.77, 1.21]
2.6 Early infant death or severe PVH 2 181 Risk Ratio (M-H, Random, 95% CI) 0.41 [0.07, 2.35]
2.7 Low Apgar score at 5 minutes 5 1224 Risk Ratio (M-H, Fixed, 95% CI) 1.09 [0.75, 1.59]
2.8 Use of mechanical ventilation 7 806 Risk Ratio (M-H, Random, 95% CI) 1.02 [0.86, 1.20]
2.9 Respiratory distress syndrome 2 209 Risk Ratio (M-H, Random, 95% CI) 1.13 [0.77, 1.67]
2.10 Pulmonary air leak 5 668 Risk Ratio (M-H, Fixed, 95% CI) 0.79 [0.43, 1.44]
2.11 Patent ductus arteriosus 3 581 Risk Ratio (M-H, Fixed, 95% CI) 0.95 [0.64, 1.42]
2.12 Cerebral palsy at 18-36 months follow up 2 518 Risk Ratio (M-H, Fixed, 95% CI) 1.07 [0.58, 1.96]
2.13 Cerebral palsy at 7 year follow up 1 299 Risk Ratio (M-H, Fixed, 95% CI) 0.77 [0.33, 1.76]
2.14 Other neuromotor impairment at 36 month follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.67 [0.13, 3.49]
2.15 Visual impairment at 18-36 months follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.89 [0.16, 5.11]
2.16 Hearing impairment at 18-36 months follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.44 [0.02, 10.64]
2.17 Bayley Mental Developmental Index at 18-24 months 2 543 Mean Difference (IV, Random, 95% CI) -4.21 [-11.89, 3.48]
2.18 Bayley Psychomotor Developmental Index at 18-24 months 2 538 Mean Difference (IV, Fixed, 95% CI) -1.82 [-4.71, 1.07]
2.19 Wechsler Intelligence Scale for Children at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  2.19.1 Full scale 1 299 Mean Difference (IV, Fixed, 95% CI) 0.30 [-2.92, 3.52]
  2.19.2 Performance 1 299 Mean Difference (IV, Fixed, 95% CI) 1.20 [-2.23, 4.63]
  2.19.3 Verbal 1 299 Mean Difference (IV, Fixed, 95% CI) -0.60 [-3.84, 2.64]
2.20 Wide Range Achievement Test at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  2.20.1 Reading 1 299 Mean Difference (IV, Fixed, 95% CI) 0.70 [-2.57, 3.97]
  2.20.2 Spelling 1 299 Mean Difference (IV, Fixed, 95% CI) -0.50 [-3.45, 2.45]
  2.20.3 Mathematics 1 299 Mean Difference (IV, Fixed, 95% CI) -1.40 [-4.61, 1.81]
2.21 Achenbach Teacher's Report Form at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  2.21.1 Total 1 299 Mean Difference (IV, Fixed, 95% CI) -1.20 [-3.55, 1.15]
  2.21.2 Internal 1 299 Mean Difference (IV, Fixed, 95% CI) -0.60 [-2.82, 1.62]
  2.21.3 External 1 299 Mean Difference (IV, Fixed, 95% CI) -1.50 [-3.45, 0.45]
2.22 Child Behavioral Check List total at 7 year follow up 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
  2.22.1 Total 1 299 Mean Difference (IV, Fixed, 95% CI) 1.00 [-1.28, 3.28]
  2.22.2 Internal 1 299 Mean Difference (IV, Fixed, 95% CI) 2.00 [-0.10, 4.10]
  2.22.3 External 1 299 Mean Difference (IV, Fixed, 95% CI) -0.60 [-2.92, 1.72]
2.23 Height (cm) at childhood follow up 1 436 Mean Difference (IV, Fixed, 95% CI) 0.00 [-0.75, 0.75]
2.24 Height < 5th centile at childhood follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.49 [0.17, 1.42]
2.25 Weight (kg) at childhood follow up 1 436 Mean Difference (IV, Fixed, 95% CI) -0.10 [-0.38, 0.18]
2.26 Weight < 5th centile at childhood follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.22 [0.03, 1.79]
2.27 Head circumference (cm) at childhood follow up 1 436 Mean Difference (IV, Fixed, 95% CI) -0.20 [-0.53, 0.13]
2.28 Head circumference < 5th centile at childhood follow up 1 96 Risk Ratio (M-H, Fixed, 95% CI) 0.67 [0.18, 2.52]
2.29 Maternal sedation 1 576 Risk Ratio (M-H, Fixed, 95% CI) 2.06 [1.79, 2.37]
 

3 Phenobarbital versus control - high-quality trials only (A for allocation concealment and < 20% exclusions)

Outcome or Subgroup Studies Participants Statistical Method Effect Estimate
3.1 All periventricular haemorrhage (PVH) 2 945 Risk Ratio (M-H, Fixed, 95% CI) 0.90 [0.75, 1.08]
3.2 Severe (grades 3 and 4) PVH 2 945 Risk Ratio (M-H, Fixed, 95% CI) 1.05 [0.60, 1.83]
3.3 Perinatal mortality 2 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
  3.3.1 Stillbirths 2 1082 Risk Ratio (M-H, Fixed, 95% CI) 2.38 [0.46, 12.21]
  3.3.2 Death of live born infants prior to discharge 2 1075 Risk Ratio (M-H, Fixed, 95% CI) 1.42 [0.91, 2.20]
  3.3.3 Infant deaths post-discharge 2 953 Risk Ratio (M-H, Fixed, 95% CI) 3.25 [0.80, 13.19]
3.4 Early infant death or any PVH 1 668 Risk Ratio (M-H, Fixed, 95% CI) 1.04 [0.79, 1.37]
3.5 Low Apgar score at 5 minutes 2 1040 Risk Ratio (M-H, Fixed, 95% CI) 1.22 [0.82, 1.82]
3.6 Use of mechanical ventilation 1 372 Risk Ratio (M-H, Fixed, 95% CI) 0.94 [0.79, 1.12]
3.7 Pulmonary air leak 1 372 Risk Ratio (M-H, Fixed, 95% CI) 1.42 [0.41, 4.95]
3.8 Patent ductus arteriosus 1 372 Risk Ratio (M-H, Fixed, 95% CI) 0.95 [0.42, 2.13]
3.9 Maternal sedation 1 576 Risk Ratio (M-H, Fixed, 95% CI) 2.06 [1.79, 2.37]
 

[top]

Figures

Figure 1

Refer to Figure 1 caption below.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies (Figure 1).

Refer to Figure 2 caption below.

Figure 2

Methodological quality summary: review authors' judgements about each methodological quality item for each included study (Figure 2).

[top]

Sources of support

Internal sources

  • Discipline of Obstetrics and Gynaecology, The University of Adelaide, Australia
  • NSW Centre for Perinatal Health Services Research, University of Sydney, Australia

External sources

  • Australian Government. Department of Health and Ageing, Australia

This review is published as a Cochrane review in The Cochrane Library, Issue 1, 2010 (see http://www.thecochranelibrary.com External Web Site Policy for information). Cochrane reviews are regularly updated as new evidence emerges and in response to feedback. The Cochrane Library should be consulted for the most recent recent version of the review.