Field Methods and Response Rates in the Tecumseh Community Health Study

Abstract

Trained lay interviewers collected detailed histories of past and present illness and arranged appointments for physical examinations for 8600 persons of all ages to provide base-line data for a comprehensive longitudinal study of health in a natural community. Sampling techniques incorporated into the study design (which aimed ultimately at 100% coverage of the study population) permitted the inclusion of additional clinic procedures and inquiries and provided 10 ten% random samples of households for estimating population parameters at 10 different points during the 18-month canvass period. Of 9794 persons contacted, 88% were interviewed and examined; an additional 523 (5.4%) were interviewed but declined the examination. Based on one 10% sample, response was highest (98.8%) among children under 10 years of age and lowest (68.6%) for persons aged 60-69, with intermediate rates of 86% and 96%, for persons aged 20-29 and 40-49, respectively. The physical examination included chest X-ray, electrocardiograms, blood tests and other special procedures.