The Australian antiepileptic drug in pregnancy register: Aspects of data collection and analysis

doi:10.1016/j.jocn.2006.08.015

Journal of Clinical Neuroscience

Volume 14, Issue 10, October 2007, Pages 936-942

https://doi.org/10.1016/j.jocn.2006.08.015 Get rights and content

Abstract

Internal comparisons using the data of the Australian Register of Antiepileptic Drugs in Pregnancy as of November 2005, and comparisons with Australian population data, were carried out. It was found that (i) except for under-representation of mothers of Asian origin, the demography of the register population was reasonably representative of the Australian situation; (ii) except for more pregnancies terminated for foetal malformation, malformation rates were similar in retrospectively and non-retrospectively enrolled pregnancies; (iii) some 21% of foetal malformations would have been excluded by not including malformations first recognised after the neonatal period; and (iv) in practice, the comparator against which malformation rates were expressed made little difference to the rates found. It is desirable to have available such analyses of pregnancy register data before comparing the findings of different registers.

Introduction

Since reports in the 1970s of an association between maternal antiepileptic drug (AED) intake in pregnancy and foetal malformations (FMs),[1], [2] there have been a number of attempts to clarify the extent and detailed nature of the problem. The earlier attempts usually involved analyses of relatively small case series, sometimes series collected retrospectively and usually derived from a single institution’s records or from data not necessarily obtained for the purpose of the analysis later carried out. Subsequently, larger case series were assembled, based on data from cooperating institutions situated in one or sometimes several countries. More recently, formal national or supranational registers have been set up in several countries, including Australia, which are specifically intended to prospectively collect data concerning AED-treated pregnancies and their outcomes, particularly in relation to the occurrence of FMs. Publications reporting various findings from some of the registers, particularly those in Australia[3], [4], [5], [6], [7] and the United Kingdom,⁸ have appeared.

One problem in assessing the significance of the reported FM rates in the literature has been determining the malformation rates that exist in otherwise comparable but AED-unexposed pregnancies. Analyses of Australian data have used for comparative purposes FM rates in AED-unexposed pregnancies that had been included in the register. An analysis has been carried out to assess the risk of bias when such internal control material is used for comparative purposes in the Australian data. This analysis is the subject of a separate publication, and the matter is not pursued further in the present paper. As of November 2005, the Australian register had nearly 1000 pregnancies enrolled, and it seemed prudent at this stage to try to assess whether the approach employed for data collection and analysis for this register might have produced serious bias in conclusions drawn on the basis of study involving these data. In particular, the following issues seemed worth exploration:

1.
Are women with epilepsy (WWE) and other AED-treated women in the register representative of such persons in the wider Australian population, at least in relation to demographic aspects?
2.
What were the effects of retrospective inclusion of data from recently completed pregnancies?
3.
What were the effects of inclusion in the analysis of FMs not detected in the neonatal period but recognised by the end of the first year of postnatal life?
4.
What was the effect of expressing FM rates relative to various comparators?

Section snippets

Materials and methods

The Australian Register began collecting data in the latter half of 1999. The register was originally situated at St Vincent’s Hospital in Melbourne and was then under the ethics oversight of the Research Ethics Committee of that institution. In the latter part of 2005 the register site moved to the Monash Medical Centre, Melbourne, and its ethical oversight became the responsibility of the Ethics Committee of Monash University.

The register was set up as a computerised database primarily

Results

As of November 2005, 957 pregnancies had been enrolled in the Australian register. The outcomes of 73 of these pregnancies were not known (nine were lost to follow up, and the remaining pregnancies were still in progress). Because 19 of the pregnancies (2.17%) involved twins, 893 known foetal outcomes resulted from the 874 pregnancies with known outcomes. Pregnancies accumulated at a rate of approximately 175 per year. Since Australian Bureau of Statistics data indicated that one baby is born

Discussion

After the Australian AED in Pregnancy Register had been running for 6 years, it seemed sensible to try to assess the validity of its data collection processes and the approach to data analysis employed. Such an assessment might have indicated the need for changes in the collection or analytical approaches, or even the possibility that the whole activity should be abandoned because conclusions drawn from it would be too uncertain.

For the Australian register, which is based entirely on voluntary

Acknowledgement

We wish to thank our colleagues, both medical and paramedical, the ethics committees and advisory board, and our sponsors, the Epilepsy Society of Australia, Epilepsy Australia and Epilepsy Action NSW, Sanofi-Aventis, UCB Pharma, Janssen Cilag, Biogen, Pfizer, Mayne Pharma, Novartis and GSK.

References (17)

B.D. Speidel et al.
Maternal epilepsy and abnormalities of the fetus and newborn
Lancet
(1972)
F.J. Vajda et al.
The Australian registry of anti-epileptic drugs in pregnancy: experience after 30 months
J Clin Neurosci
(2003)
F.E. Vajda et al.
Critical relationship between sodium valproate dose and human teratogenesis: results of the Australian register of anti-epileptic drugs in pregnancy
J Clin Neurosci
(2004)
S.R. Meadow
Congenital abnormalities and anticonvulsant drugs
Proc R Soc Med
(1970)
F. Vajda et al.
Foetal malformations and seizure control: 52 months data of the Australian Pregnancy Registry
Eur J Neurol
(2006)
M. Eadie et al.
Should valproate be taken during pregnancy?
Ther Clin Risk Manage
(2004)
F.J.E. Vajda et al.
Maternal valproate dosage and foetal malformations
Acta Neurol Scand
(2005)
J. Morrow et al.
Malformation risks of anti-epileptic drugs in pregnancy: A prospective study from the UK Epilepsy and Pregnancy Register
J Neurol Neurosurg Psychiatry
(2006)

There are more references available in the full text version of this article.

Cited by (29)

Specific fetal malformations following intrauterine exposure to antiseizure medication
2023, Epilepsy and Behavior
To investigate in the Australian Pregnancy Register of Antiepileptic Drugs patterns of fetal malformation associated with intrauterine exposure to particular currently available antiseizure medications taken by women with epilepsy.
There was statistically significant evidence (P < 0.05) of an increased hazard of fetal malformation associated with exposure to valproate, carbamazepine, topiramate, zonisamide, and with conception after assisted fertilization, but a reduced hazard in the offspring of women who continued to smoke during pregnancy. Valproate exposure was associated with malformations in a wide range of organs and organ systems, carbamazepine and topiramate with hydronephrosis, topiramate also with hypospadias, zonisamide with spina bifida and assisted fertilization with heart and great vessel maldevelopment.
Prenatal valproate exposure appears to interfere with the development of many if not all, fetal tissues. It seems likely that prenatal exposure to carbamazepine and topiramate, and possibly exposure to zonisamide, but also some process related to in vitro fertilization, may more selectively affect the normal development of particular fetal tissues or organs.
Fetal malformations in successive pregnancies in Australian women with epilepsy
2022, Epilepsy and Behavior
To utilize data from the Australian Register of Antiepileptic Drugs in Pregnancy (APR) to determine the hazard of fetal malformation in the subsequent pregnancy or pregnancies in women with epilepsy following a pregnancy associated with a fetal malformation, and to identify factors relevant to the hazard.
There was a 7.4% initial pregnancy fetal malformation rate. The subsequent pregnancy malformation rate was 4.2% if there was no initial pregnancy malformation, but 21.2% if there was an initial pregnancy malformation (O.R. = 6.1448, 95% C.I. 2.3396, 16.1386). For pregnancies where antiseizure medication (ASM) therapy was unchanged between pregnancies (N = 196), the initial pregnancy malformation rate was 10.2%, but 30.0% in the subsequent pregnancy if there was a malformation in the initial pregnancy, and 2.35% if there was none (O.R. = 17.7857, 95% C.I. 4.4847, 70.5361). A cohort comprising 24% of the women with fetal malformations in their initial pregnancies seemed to be intrinsically vulnerable to fetal malformation during successive pregnancies: when their seizure disorder type had been recorded all had genetic generalized epilepsies, compared with a 45.8% generalized epilepsy rate in women with initial but not subsequent pregnancy malformations (P = 0.0121).
If fetal malformation had occurred in an initial ASM-treated pregnancy there was a significantly increased hazard of fetal malformation in the subsequent pregnancy, particularly if the woman involved had a genetic generalized epilepsy.
Neurological factors and Cesarean section in Australian women with epilepsy
2022, Epilepsy and Behavior
Citation Excerpt :
The present paper revisits the matter in a larger population of pregnancies from the APR with the intention of examining more closely the possible roles of various factors relating to epilepsy and neurological disorder in accounting for the higher Cesarean section rate in Australian women with epilepsy, relative to that in the general female population. Details concerning the APR, including its recruitment and data collection policies and ethics oversight have been published in several places in the literature [7,8]. The primary purpose of the APR is to permit ongoing investigation of the role of antiseizure medication in relation to fetal malformation, but information is also collected regarding other aspects of the pregnancies involved.
To analyze the records of the pregnancies of 2283 Australian women with epilepsy in the Australian Register of Antiepileptic Drugs in Pregnancy database to identify neurological factors relevant to the Cesarean sections carried out in these pregnancies.
The Cesarean section rate in Australian women overall increased by an average of 0.59% annually over 20 years, from 26.0% to its calculated 2020 value of 37.3%. For the operations in women with epilepsy, the corresponding figures were 0.71% annually, and 34.4% and 48.7%. The average annual rate of increase for pre-labor operations was 0.89% to a 2020 value of 39.1%, the annual rate for operations during labor showing no statistically significant change. Multivariate regression analysis identified a number of characteristics of women with epilepsy that were statistically significantly associated with an increased likelihood of Cesarean section, but of these only seizures continuing to occur in the third trimester and having chronic illness, in particular migraine, were neurological ones. In 70 migraine-affected women, the Cesarean section rate was 51.4%, compared with 39% in the remaining pregnancies (P < 0.05).
Having seizures in the final trimester of pregnancy and having chronic neurological illness, especially migraine, favored Cesarean section being carried out in Australian women with epilepsy, but did not adequately account for the increasing rates of occurrence of the operation over the past 20 years.
The outcome of altering antiepileptic drug therapy before pregnancy
2020, Epilepsy and Behavior
Citation Excerpt :
The APR contains data concerning the pregnancies of a cohort of Australian women who were taking AEDs, mostly but not exclusively for epilepsy, or who had epilepsy that was not being treated with such drugs, at least in the earlier term of pregnancy. More detailed information regarding the APR and its recruitment policies and data handling practices has been published previously [2,3]. After such women planning pregnancy, or already pregnant, learn of the Register's existence through various means, they may choose to be included in its database, which was estimated to contain information on about 8.7% of the relevant pregnancies that had occurred in Australia [4].
We investigated the outcome of altering antiepileptic drug (AED) therapy in the year before pregnancy on 2233 occasions in Australian women in the 20-year period of functioning of the Raoul Wallenberg Australian Pregnancy Register (APR).
Therapy had been altered in 358 instances (16%) in the months prior to the pregnancy (median interval: 18 weeks). Antiepileptic drug doses had been changed in 141 pregnancies (39.4%), being decreased in 94; drugs changed in 151 (42.2%); drugs withdrawn without replacement in 66 (18.4%) but resumed in 40 before pregnancy ended. The main drugs involved were valproate (34%), phenytoin (16.5%), topiramate (12.6%), and carbamazepine (11.4%). Antiepileptic drug doses were increased significantly more often (16.9% vs. 6.4%) when epilepsy before pregnancy was not controlled, and AED treatment ceased significantly less often (13.6% vs. 24.0%). The alterations were more often made in women with generalized epilepsies and in those whose seizure disorders were not fully controlled in the prepregnancy year, suggesting that avoidance of teratogenicity and achieving improved seizure control often motivated the changes.
Overall, the alterations did not result in improved rates of seizure freedom during pregnancy, as compared with pregnancies where therapy was unchanged; however, fetal malformation rates were lower 3.6% vs. 5.4%, but this difference did not attain statistical significance. The same trends regarding seizure control and malformations persisted after pregnancies involving valproate exposure were excluded.
In conclusion, this analysis of the APR cohort did not demonstrate that altering AEDs before pregnancy produced a significant improvement in seizure control and the reduction in fetal malformation rate that occurred was not statistically significant.
Cesarean section in Australian women with epilepsy
2018, Epilepsy and Behavior
Citation Excerpt :
These women were essentially self-selected from the Australian population of WWE, and the primary purpose of collecting their data was not the issue of CS. Nevertheless, birth details were recorded systematically in all instances [6,7]. The similarity in the rates for forceps and suction deliveries in the APR and ABM populations suggests that there probably was little or no selection bias in relation to the pattern of obstetric management in the APR material.
The literature suggests that cesarean delivery or birth is carried out more often in pregnant women with epilepsy (WWE) than in pregnant women in the general population. Data were utilized from the Australian Pregnancy Register (APR) for Women on Antiepileptic Medication to investigate this issue in Australia.
Over almost two decades, the mean CS rate in 1900 APR women was 39.2%, but was only 29.9% in women in the general population (relative risk (R.R.) = 1.31, 95% confidence interval (C.I.) 1.24, 1.39). Rates for forceps and suction-assisted delivery were similar in the two datasets. The 9.3% excess CS rate was almost entirely accounted for by operations carried out prior to the onset of labor. The rates for CS during labor were very similar. Only 11.0% of the WWE knew the indication for their prelabor CS, whereas 69.8% knew why theirs had been carried out during labor (odds ratio (O.R.) = 0.054; 99% C.I. 0.032, 0.089). Slightly older mothers and increased proportions of primipara probably made small contributions to the increased prelabor CS rate in the Australian WWE, but most of the excess could not be accounted for in the Register data.
Australian obstetricians may have tended to regard prelabor CS as a preferable course of action in managing delivery in WWE, even in the absence of other indications.
Predicting epileptic seizure control during pregnancy
2018, Epilepsy and Behavior
The objective of the study was to assess whether the type of seizure disorder present in the prospective mother with epilepsy, her use of antiepileptic drugs (AEDs) in early pregnancy, and her seizure control before pregnancy help predict her prospects for seizure freedom throughout pregnancy.
This paper is based on data accumulated in the Australian Pregnancy Register (APR) between 1998 and late 2016. Information was analyzed concerning epileptic seizure occurrence and AED therapy taken before and during pregnancy, using simple statistical and confidence interval (C.I.) methods, mainly relative risk (R.R.) calculations.
After excluding pregnancies lost to follow-up, and those that ended prematurely because of spontaneous abortion or stillbirth, 1939 pregnancies were available for study. Seizures had occurred during pregnancy in 829 (42.8%), and convulsive seizures in 385 (19.9%). Seizures of any type occurred in 78.4% of pregnancies where seizures had occurred in the previous year (active epilepsy) and in 22.3% of those associated with inactive epilepsy. Seizures of any type had occurred in 54.9% of pregnancies initially unexposed to AEDs and in 45.5% of those treated with AEDs throughout. The corresponding figures for convulsive seizures during pregnancy were 31.7% and 22.3%. There was statistically significant evidence that, in women with epilepsy (WWE), having a seizure disorder that was active in the prepregnancy year and one untreated in early pregnancy was associated with decreased prospects of seizure freedom during pregnancy. Decreased chances of seizure-free pregnancies in women with focal epilepsies and those treated with multiple AEDs were probably explained by greater frequencies of active seizure disorders in these patient categories.
Women with epilepsy who experience seizures in the year prior to pregnancy appear 3 or 4 times more likely to continue to have seizures during pregnancy than women whose seizures are fully controlled prior to pregnancy. Not taking AEDs in early pregnancy also increases the hazard for seizure occurrence in pregnancy.

View all citing articles on Scopus

View full text

Clinical studyThe Australian antiepileptic drug in pregnancy register: Aspects of data collection and analysis

Abstract

Introduction

Section snippets

Materials and methods

Results

Discussion

Acknowledgement

Lancet

J Clin Neurosci

J Clin Neurosci

Congenital abnormalities and anticonvulsant drugs

Proc R Soc Med

Foetal malformations and seizure control: 52 months data of the Australian Pregnancy Registry

Eur J Neurol

Should valproate be taken during pregnancy?

Ther Clin Risk Manage

Maternal valproate dosage and foetal malformations

Acta Neurol Scand

Malformation risks of anti-epileptic drugs in pregnancy: A prospective study from the UK Epilepsy and Pregnancy Register

J Neurol Neurosurg Psychiatry

Clinical study
The Australian antiepileptic drug in pregnancy register: Aspects of data collection and analysis