EDITOR—Hippisley-Cox et al lament their inability to identify teenagers with repeat pregnancies and, by inference, repeat abortions.1 The Trent regional hospital admissions database apparently contains all details of hospital admissions for residents in Trent, whether treatment was provided in Trent or not. Does this database contain the NHS number, the only unique identification number that is almost universally held among the population of England and Wales?2 If so, record linkage analysis could have been used to identify repeat pregnancies and repeat abortions, especially since Hippisley-Cox et al compared their data with those of the Office for National Statistics, of which the NHS central register is a part. The goal of record linkage is to link quickly and accurately records corresponding to the same person or entity. A record linkage system exists in England that recently included the matching of general practice records with hospital and vital records to prepare a file for analysing referral, prescribing, and outcome measures.3

This study highlights a deficiency in the data collection system for legal abortion in England and Wales. Abortion statistics are compiled from completed abortion notification forms (form HSA4, revised 1991), which have no requirement for a unique personal identification number. Although it has a section for the number of previous legal abortions experienced by each woman, the accuracy of the data depends on the extent of truthful disclosure by women of their past experience of induced abortion and accurate reporting by abortion service providers. The information may not be verifiable from hospital records if a woman has attended different hospitals.

Record linkage analysis showed underreporting of experience of induced abortion when the study methodology depended on self-reporting.4 A mandatory requirement for the NHS …