Identification of people attending outpatients: congruity between hospital records and response to a postal screen

A comparison of information obtained from a postal screen of people on the electoral register about attendance at outpatient clinics with data extracted from hospital records found agreement between the two sources for 87 per cent of people. This rose to 90 per cent after an interview follow-up of those whose replies were unclear. Given the complexities of defining outpatient attendances clearly, these levels of agreement seemed good. Fewer consultations were omitted by people when a three rather than a 12-month study period was used, but the particular three-month period used (January-March) may have contributed to this difference.