Abstract

OBJECTIVES: This study aimed to assess the feasibility, sensitivity, and specificity of a systematic search of the NHS central register for twins of the same sex. DESIGN, SETTING, AND SUBJECTS: Computerised searches by alphanumeric NHS number, sex, and date of birth of 1.6 million patients registered with general practitioners in Kent, followed by validation by postal questionnaire sent to 66 pairs of children, 129 pairs of adult twins, and one set of triplets. MAIN OUTCOME MEASURES: These were as follows: confirmed twin status, previous involvement in research, and willingness to participate in future research based on a twin register. RESULTS: The procedure was most efficient at identifying twins among children, in males born from 1930 onwards, and females born before 1940. Altogether 2397 male sets and 1684 female sets were identified as probable twins, triplets, or quadruplets with identical surnames, same dates of birth, and consecutive NHS numbers. Among a further 4004 pairs of adult females with the same dates of birth (1940-79), consecutive NHS numbers, but different (married) surnames, an estimated third are twins. Response to the postal questionnaire was 60% (197/327), including only one singleton. Only 10% of responders had previously participated in medical research, but 63% (65/103) of complete sets who responded expressed a willingness to participate in a twin register. CONCLUSIONS: A systematic search of the NHS central register could identify large numbers of British twins. The procedure is efficient, reasonably sensitive, and highly specific, if supplemented by additional information from birth records for adult females born after 1939. The potential exists to create an important new resource for twin studies in Britain.