Use of anti-epileptic drugs during pregnancy and risk of spontaneous abortion and stillbirth
Population based cohort study
Use of anti-epileptic drugs during pregnancy and risk of spontaneous abortion and stillbirth: population based cohort study by Bodil Hammer Bech and colleagues (BMJ 2014;349:g5159)
Objective—To determine whether use of antiepileptic drugs during pregnancy may increase the risk of spontaneous abortion or stillbirth.
Design—Population based cohort study.
Setting—Register based study in Denmark, 1997-2008.
Participants—983 305 pregnancies identified in the Danish medical birth register and the Danish national hospital discharge register from 1 February 1997 to 31 December 2008 were linked to the Danish Register of Medicinal Product Statistics to obtain information on use of antiepileptic drugs.
Main outcome measures—Risk ratio of spontaneous abortion and stillbirth after use of antiepileptic drugs during pregnancy, estimated by using binomial regression adjusting for potential confounders of maternal age, cohabitation, income, education, history of severe mental disorder, and history of drug misuse.
Results—Antiepileptic drugs were used in a total of 4700 (0.5%) pregnancies. 16 out of 100 pregnant women using antiepileptics and 13 out of 100 pregnant women not using antiepileptics experienced a spontaneous abortion. After adjusting for potential confounders pregnant women using antiepileptics had a 13% higher risk of spontaneous abortions than pregnant women not using antiepileptics (adjusted risk ratio 1.13, 95% confidence interval 1.04 to 1.24). However, the risk of spontaneous abortion was not increased in women with an epilepsy diagnosis (0.98, 0.87 to 1.09), only in women without a diagnosis of epilepsy (1.30, 1.14 to 1.49). In an analysis including women with at least two pregnancies with discordant antiepileptic drug use (for example, use in the first pregnancy but not in the second), the adjusted hazard ratio for spontaneous abortion was 0.83 (0.69 to 1.00) for exposed pregnancies compared with unexposed pregnancies. Stillbirth was identified in 18 women who used antiepileptic drugs (unadjusted risk ratio 1.29, 0.80 to 2.10).
Conclusion—Among women with epilepsy and when analysing the risk in antiepileptic drug discordant pregnancies in the same woman, we found no overall association between the use of antiepileptic drugs during pregnancy and spontaneous abortions. Therefore unmeasured confounding may explain the slight increased risk for spontaneous abortion with any antiepileptic drug use (among women both with and without epilepsy). We found no association between antiepileptic drug use during pregnancy and stillbirth, but the statistical precision was low.
Why do the study?
The use of drugs in pregnancy is surprisingly common, with estimates of prescription drug use during the first trimester ranging from 35-69% in developed countries.
Epilepsy is the most common neurological disorder likely to require medical treatment during pregnancy. Anti-epileptic drugs (AEDs) are used to control seizures in patients with epilepsy, and are increasingly prescribed to patients with migraine, chronic pain, and some psychiatric disorders. Doctors who prescribe them during pregnancy must balance the risk of unwanted drug effects against the risk of poorly controlled seizure activity or relapse of chronic illness.
AEDs are associated with pregnancy complications including pre-eclampsia, bleeding, preterm birth, intrauterine growth retardation, and fetal malformations. But we know less about the effects of AED use on the risk of spontaneous abortion (miscarriage) or stillbirth. Animal studies suggest there might be a dose-dependent increased risk of fetal death. However, human studies are contradictory. Some studies have reported a positive association, whereas others have found no link. The authors sought to assess the risk of spontaneous abortion and stillbirth associated with the use of AED in pregnancy.
What did the authors do?
They designed a retrospective linked analysis of multiple Danish population based health registers. Denmark maintains several national registries, which collect data including births, deaths, hospital discharges, and other medical records. This allows retrospective analysis of large data sets. The authors identified data on birth outcomes, use of AEDs, diagnostic, and demographic information, using the Danish system of unique identification numbers.
The authors screened the Danish medical birth and national hospital discharge registers using WHO international classification diagnostic (ICD) codes to identify the outcomes of all clinically recognised pregnancies with an estimated date of conception between 1 February 1997 and 31 December 2008. They recorded miscarriage and stillbirth (defined as birth of a dead fetus after 22 completed gestational weeks). They excluded molar and ectopic pregnancies.
Use of anti-epileptic drugs
The authors cross referenced the Danish Register of Medicinal Product Statistics to identify women filling prescriptions for AEDs. On the basis of this pharmacy refill data, they calculated cumulative and average daily doses of anti-epileptic drugs redeemed before and during pregnancy.
Diagnostic and demographic data
The authors used the Danish national hospital discharge register to identify women with ICD codes for epilepsy, who had been diagnosed before the end of the index pregnancy. They screened the Danish psychiatric central research register to identify women with concurrent psychiatric illness, and obtained demographic data, including maternal age, co-habitation, income, and education status, from Statistics Denmark.
Because the outcomes of interest are binary—either the event happens or it does not happen—the authors used a generalised linear regression model with a log-link to return risk ratios of spontaneous abortion and stillbirth in the group of women exposed to AED compared with those who were not. This type of model allows the researcher to adjust the outcome according to explanatory variables that are known to affect risk, such as maternal age, cohabitation, and income.
What did they find?
The study included data from 983 305 pregnancies, of which 109 800 (11%) resulted in spontaneous abortion and 3222 (0.3%) in stillbirth. The authors excluded from their analysis patients with missing data for any variable and those who had terminated their pregnancies (18%).
Women who were prescribed an AED (n=3981 (0.5%)) were more likely to have a history of a psychiatric disorder (including severe psychiatric disorders), substance misuse, and use antipsychotics and antidepressants than women who were not prescribed AED. Moreover, women in this group were more likely to terminate a pregnancy. The researchers adjusted for these factors in their regression analysis.
More women prescribed AEDs had a spontaneous abortion of their pregnancy (adjusted relative risk or risk ratio, 1.13; 95% confidence interval, 1.04 to 1.22). This means a statistically significant increase (the 95% confidence interval does not cross 1) in the risk of spontaneous abortions in women exposed to AEDs, with a 95% chance that the magnitude of that risk lies between 4% and 22%.
When the researchers excluded women with severe psychiatric disorders or those co-prescribed anti-psychotics or insulin from the analysis, the risk remained the same. Extending the time period in which women might have been exposed from 30 days to six months before conception did not change the risk. However, if women had stopped filling the prescription for their AED one year before conception, the researchers found no association between AED prescription and spontaneous abortion.
Adjusted sub-analyses undertaken by indication showed that women who were prescribed AEDs for epilepsy were not at increased risk of spontaneous abortion (risk ratio 0.98; 95% confidence interval 0.87 to 1.09) whereas patients prescribed AED for other reasons were (risk ratio 1.30; 95% confidence interval 1.14 to 1.49). This was true for each of the five most frequently prescribed AEDs (carbamazepine, clonazepam, lamotrigine, oxcarbazepine, and sodium valproate). Furthermore, and regardless of indication or type of AED, the risk of spontaneous abortion was higher in patients prescribed higher doses of AED.
There was no increase in the unadjusted (crude) relative risk of stillbirth in women prescribed AED (risk ratio 1.29; 95% confidence interval 0.80 to 2.10). However, only 18 women who had taken AEDs experienced stillbirth, so the authors caution that these figures lack precision.
Strengths and limitations
By including all clinically recognised pregnancies in a population based sample, this analysis reduced the chance of selection or tertiary referral bias of patients with clinical manifestations of rare diseases. We know this type of bias might have confounded previous prospective cohort or cross-sectional studies from academic centres. Although medication use in pregnancy is common, the use of specific treatments and birth defects are rare, necessitating a very large sample size to determine associations.
The main strength of this study is its sample size. This allowed the researchers to include a number of co-variates known to affect birth outcome, whilst testing for associations with the five most commonly used AED. Yet despite studying over 800 000 births, the small number of stillbirths meant the researchers were unable to draw conclusions about this outcome. This demonstrates the problem of studying rare events.
This study design has several limitations. Firstly, studies that rely on diagnostic coding for their source data risk case ascertainment bias because of miscoding and missing data. Ascertainment bias, sometimes referenced as sampling bias, refers to the potential error in measuring the true frequency of observed events because of the way the data was collected. In this study, it was caused by data registry collection (sampling). On the basis of the positive predictive value of 97% cited by this study authors’ for the code “spontaneous abortion,” about 120 of the spontaneous abortions reported might have been miscoded.
Secondly, retrospective cohort studies can only analyse data that has already been collected. In this study, AED use was assessed by pharmacy refill, a technique that does not permit any assessment of adherence to treatment. Differences in adherence represent a potential explanation for the risk differential between women with different indications for treatment. For example, patients with epilepsy and a low seizure frequency might have a low perceived need for AED, and patients with migraine or chronic pain might have a high perceived need to continue taking their drugs.
Finally, as the authors acknowledge, because the study only identified patients with hospital based clinically recognised pregnancies, it is highly likely that a proportion of spontaneous abortions occurring in early pregnancy will have been missed. This might account for the difference in spontaneous abortion rate reported here (10%) and the more frequently cited rates of 20-30% reported from prospective cohort series.
What does this study mean?
Women who took AEDs for epilepsy during pregnancy had no increased risk of spontaneous abortion or stillbirth. Women who were prescribed AED for a non-epilepsy indication had a 30% increased risk of spontaneous abortion, but no increased risk of stillbirth. The authors speculate that the differences seen according to indication may be as a consequence of the underlying disorder being treated or an unmeasured covariate.
Both groups of women demonstrated a dose-dependent increase in risk of spontaneous abortion. This suggests that if AED are to be continued in pregnancy, prescribers should opt for the lowest efficacious dose.
|Using antiepileptics (n=4700)||Not using antiepileptics (n=978 605)|
|Maternal age at conception (years):|
|<21||280 (6.0)||55 535 (5.7)|
|21-25||784 (16.7)||155 703 (15.9)|
|26-30||1484 (31.6)||333 391 (34.1)|
|31-35||1374 (29.2)||290 464 (29.7)|
|≥36||778 (16.6)||143 511 (14.7)|
|Missing||0 (0.0)||1 (0.0)|
|Yes||3244 (69.0)||754 687 (77.1)|
|No||1436 (30.6)||215 216 (22.0)|
|Missing||20 (0.4)||8702 (0.9)|
|First (lowest)||1130 (24.0)||193 300 (19.8)|
|Second||1280 (27.2)||195 440 (20.0)|
|Third||957 (20.4)||195 989 (20.0)|
|Fourth||778 (16.6)||196 379 (20.1)|
|Fifth (highest)||555 (11.8)||196 348 (20.1)|
|Missing||0 (0.0)||1149 (0.1)|
|<10||828 (17.6)||107 387 (11.0)|
|10-12||1782 (37.9)||303 302 (31.0)|
|>12||1970 (41.9)||539 611 (55.1)|
|Missing||120 (2.6)||28 305 (2.9)|
|Maternal psychiatric history:|
|Yes||1213 (25.8)||61 335 (6.3)|
|No||3487 (74.2)||917 270 (93.7)|
|Missing||0 (0.0)||0 (0.0)|
|Maternal history of substance misuse:|
|Yes||257 (5.5)||5263 (0.5)|
|No||4443 (94.5)||973 342 (99.5)|
|Missing||0 (0.0)||0 (0.0)|
|Maternal history of severe psychiatric disorder:|
|Yes||253 (5.4)||2439 (0.2)|
|No||4447 (94.6)||976 166 (99.8)|
|Missing||0 (0.0)||0 (0.0)|
|Concurrent drug use:|
|Any||4700 (100.0)||617 424 (63.1)|
|Antipsychotics||364 (7.7)||2740 (0.3)|
|Antidepressants||752 (16.0)||21 000 (2.1)|
|Insulin||42 (0.9)||4537 (0.5)|
|Induced abortion (reason):|
|Fetal disease||22 (0.5)||2,987 (0.3)|
|Other||1116 (23.7)||171 569 (17.5)|
|Combined||1138 (24.2)||174 556 (17.8)|
Limitations of retrospective cohort studies
- Very large sample sizes are needed if the outcome of interest is rare
- Case ascertainment bias because of miscoding
- Missing data
- Recall bias
- Observational only
1Colchester General Hospital, Colchester, Essex, 2Barts Health NHS Trust - Newham University Hospital, London
Correspondence to: email@example.com
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.
- Daw JR, Hanley GE, Greyson DL, Morgan SG. Prescription drug use during pregnancy in developed countries: a systematic review. Pharmacoepidemiol Drug Saf 2011;20:895-902.
- Meador KJ, Baker GA, Finnell RH, Kalayjian LA, Liporace JD, Loring D, et al, for the NEAD Study Group. In utero anti-epileptic drug exposure: fetal death and malformations. Neurology 2006;67:407-12.
- Ornoy A, Zvi N, Arnon J, Wajnberg R, Shechtman S, Diav-Citrin O. The outcome of pregnancy following topiramate treatment: a study on 52 pregnancies. Reprod Toxicol 2008;25:388-9.
- Kanafani ZA, Kanj SS, Cabell CH. Revisiting the effect of referral bias on the clinical spectrum of infective endocarditis in adults. Eur J Clin Microbiol Infect Dis 2010;29:1203-10.
- Thorpe PG, Gilboa SM, Hernandez-Diaz S, Lind J, Cragan JD, Briggs G, et al. National Birth Defects Prevention Study. Medications in the first trimester of pregnancy: most common exposures and critical gaps in understanding fetal risk. Pharmacoepidemiol Drug Saf 2013;9:1013-8.
- Buss L, Tolstrup J, Munk C, Bergholt T, Ottesen B, Grønbaek M, et al. Spontaneous abortion: a prospective cohort study of younger women from the general population in Denmark. Validation, occurrence and risk determinants. Acta Obstet Gynecol Scand 2006;85:467-75.
Cite this as: Student BMJ 2014;22:g6328