Clinical studyThe Australian antiepileptic drug in pregnancy register: Aspects of data collection and analysis
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)
- et al.
Maternal epilepsy and abnormalities of the fetus and newborn
Lancet
(1972) - et al.
The Australian registry of anti-epileptic drugs in pregnancy: experience after 30 months
J Clin Neurosci
(2003) - 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) Congenital abnormalities and anticonvulsant drugs
Proc R Soc Med
(1970)- et al.
Foetal malformations and seizure control: 52 months data of the Australian Pregnancy Registry
Eur J Neurol
(2006) - et al.
Should valproate be taken during pregnancy?
Ther Clin Risk Manage
(2004) - et al.
Maternal valproate dosage and foetal malformations
Acta Neurol Scand
(2005) - 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)
Cited by (29)
Specific fetal malformations following intrauterine exposure to antiseizure medication
2023, Epilepsy and BehaviorFetal malformations in successive pregnancies in Australian women with epilepsy
2022, Epilepsy and BehaviorNeurological factors and Cesarean section in Australian women with epilepsy
2022, Epilepsy and BehaviorCitation 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.
The outcome of altering antiepileptic drug therapy before pregnancy
2020, Epilepsy and BehaviorCitation 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].
Cesarean section in Australian women with epilepsy
2018, Epilepsy and BehaviorCitation 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.
Predicting epileptic seizure control during pregnancy
2018, Epilepsy and Behavior