Benzodiazepine use in pregnancy and major malformations or oral clefts
BMJ 1999; 319 doi: https://doi.org/10.1136/bmj.319.7214.918 (Published 02 October 1999) Cite this as: BMJ 1999;319:918
This article has a correction. Please see:
Induced abortions should be included
- Ester Garne, specialist in paediatrics,
- Ulf Bergman, visiting professor
- Eurocat, University of Southern Denmark, 5000 Odense C, Denmark
- Department of Clinical Pharmacology, University of Southern Denmark
- Bushey, Hertfordshire WD2 2NN
- Department of Obstetrics and Gynaecology, Birmingham Women's Hospital, Birmingham B15 7TG
EDITOR—Dolovich et al have reported their meta-analysis concerning use of benzodiazepines during the first trimester of pregnancy and the risk of oral clefts or major congenital malformations.1 In the cohort studies no increased risk of congenital malformations was found, while case-control studies showed an increased risk of both oral clefts and major congenital malformations. The authors excluded studies from the meta-analysis if only stillbirths or abortions were included.
The paper missed important information about the criteria for inclusion in the meta-analysis. During the past 10-15 years prenatal diagnosis of congenital malformations has increased significantly in many countries and is often followed by termination of pregnancy if severe malformations are diagnosed.2 The Eurocat project is a European epidemiological programme surveying congenital malformations. The Eurocat registries are population based and based on multiple sources of information; they include information about live births, deaths of fetuses with gestational age ≥20 weeks, and induced abortions after prenatal diagnosis of malformations.
During 1990-4 there were 13 registries following strict Eurocat methodology and where induced abortion after prenatal diagnosis of congenital malformations was noted. The 13 registries covered 1 284 599 births. For this period the contribution of induced abortion in the registries ranged from 4.0-7.6 per 1000 births in the three French registries to 0.7-3.1 per 1000 births in the other European countries.2 The table shows the proportion of induced abortion among cases of major congenital malformations. If studies concerning risk factors during pregnancy include only information from live births and stillbirths they will miss the large number of cases diagnosed prenatally in which termination of pregnancy is subsequently carried out. This is not discussed in the paper.1
Selected major congenital malformations detectable by prenatal ultrasound investigation (data from 13 Eurocat registries 1990-42)
In studies from the 1970s and ‘80s the problem is of only minor importance For future studies of risk assessments during pregnancy it is important also to include terminations of pregnancies. These terminations often occur in the more severe cases, and prenatal diagnostic methods are more often performed in risk pregnancies, in anxious mothers taking benzodiazepines, and in mothers with epilepsy. If only live births are studied the conclusions might be wrong because today live births occur in a selected group of all pregnancies.
References
Pooled results are sensitive to zero transformation used
- Christopher Cates, general practitioner
- Eurocat, University of Southern Denmark, 5000 Odense C, Denmark
- Department of Clinical Pharmacology, University of Southern Denmark
- Bushey, Hertfordshire WD2 2NN
- Department of Obstetrics and Gynaecology, Birmingham Women's Hospital, Birmingham B15 7TG
EDITOR—In 1994 Shapiro suggested that meta-analysis to combine the results of observational studies should be abandoned because the individual studies are open to particular bias related to their study design.1 Dolovich et al have reported their meta-analysis of the use of benzodiazepines during the first trimester of pregnancy and the risk of oral clefts or major congenital malformations.2 The differences they found between the results of the cohort and case-control studies suggest that it may be bias that is being pooled rather than true effects. The authors themselves express reservations that most cases reported were derived from only three studies, so have the patients with oral cleft been counted more than once in the pooling process?
The statistical method used in this paper cannot be relied on, as the pooled results are sensitive to the adjustment made for trials in which the number of malformations was zero. For example, the largest cohort study (that by Bergman et al, 1992) is calculated as showing an odds ratio of 1.21, based on 0/1354 exposed patients having oral cleft compared with 62/102 985 not exposed. This result suggests that the zeros have been adjusted to 1 for the calculation (as odds of zero cannot be turned into an odds ratio). If the zero was counted as 0.5, however, the odds ratio for the study would fall to 0.61.
Even more bizarre is the study of epileptic patients (that by Robert et al, 1986), in which 0/4 compared with 1/144 is transformed into an odds ratio of 10.63 on the basis of the zero transformation. Surely 0/4 is exactly the expected finding if the odds ratio was 1 and there was no difference between the exposed and non-exposed groups
This paper seems to lend weight to Shapiro's suggestion that meta-analysis should be confined to controlled clinical trials.
Quality of primary studies must influence inferences made from meta-analyses
- Khalid S Khan, lecturer (khalid.khan{at}bham-womens.thenhs.com),
- Catherine Wykes, specialist registrar,
- Harry Gee, consultant
- Eurocat, University of Southern Denmark, 5000 Odense C, Denmark
- Department of Clinical Pharmacology, University of Southern Denmark
- Bushey, Hertfordshire WD2 2NN
- Department of Obstetrics and Gynaecology, Birmingham Women's Hospital, Birmingham B15 7TG
EDITOR—In a meta-analysis of cohort and case-control studies on the association of malformations with use of benzodiazepines in pregnancy, Dolovich et al recommend the use of antenatal diagnostic tools for detecting malformations in clinical practice.1 For future research they say that further case-control studies are needed. They have disregarded the importance of the quality of primary studies in arriving at these conclusions.
In their analysis Dolovich et al stratify pooling of data according to the quality of the study. Higher quality studies (cohort studies) showed lack of an association, while poor quality studies (case-control studies) showed an association. This is apparent from the graphic presentation of odds ratios in the paper. To emphasise this point we have performed logistic regression analysis, with fetal malformation (both major malformation and oral cleft) as the binary dependent variable.
A logistic model was built to assess the impact of the quality of the study on the strength of the association between exposure to benzodiazepines and malformation. The value of the coefficient for the interaction term between “study quality” (cohort v case-control) and “exposure” (exposure v non-exposure) provided an estimate of the bias in poor quality studies.2 Our analysis showed that the case-control studies exaggerated the estimate of the odds ratio for the association between malformation and exposure to benzodiazepines by 55% (95% confidence interval 28% to 72%, P=0.0006) compared with cohort studies.
It has been well established that poorer quality literature tends to exaggerate the estimate of the strength of an association, and in this circumstance it is more appropriate to base inferences on higher quality evidence.3 Our conclusion on the basis of meta-analysis of higher quality evidence is that there is no association between maternal exposure to benzodiazepines and fetal malformation; further research should aim at generating high quality evidence in the form of cohort studies rather than case-control studies.