Measuring the prevalence of disease in middle-aged British men

Objective: To identify and validate a measure of disease prevalence in a to wn or other subpopulation for the purpose of estimating variations in morbi dity and health care needs between groups. Design: A national prospective study of cardiovascular disease based in 24 British towns. Subjects: 7,735 men aged 40-59 years sampled from 24 general practice age-sex registers between 1978 and 1980. Method: A questionnaire administered at recruitment and a postal questionna ire five years later (98% response) asked if the subject recalled ever havi ng had a diagnosis by a doctor of any of 12 listed conditions. In addition, the men were asked about regular medication, occupation, lifestyle and per ceived health status. All-cause mortality data were collected during 13.5 y ears of follow-up. Results: Prevalence of recalled doctor diagnoses varied between the towns; for cardiovascular disease, it correlated significantly with other health i ndices and personal characteristics and with standardised mortality ratios for the town. Consistency of reporting over time and mortality outcome afte r 13.5 years also supported the validity of patient recall. Conclusions: A simple standardised questionnaire using patient recall of co mmon conditions diagnosed by a doctor provides a plausible, valid measure o f variations in disease burden between communities.