The feasibility of conducting occupational epidemiology in the UK

A postal survey was carried out of 1,000 UK companies to collect informatio n about employee biographical and work history records. The overall respons e rate was 46%. All companies collected surname, forenames, address, date o f birth and National Insurance number - information needed for cross-sectio nal studies. Other biographical details such as maiden name and National He alth Service number were collected less often, which could increase the cos t and difficulty of tracing ex-employees. Seventy per cent reported destroy ing their records within 10 years of an employee leaving, rising to 82% for companies with fewer than 100 employees. The destruction of employee recor ds creates problems for historical cohort studies and case-control studies, and may hamper ex-employees trying to claim benefit for occupational-relat ed illness. If the scope of future occupational epidemiology is to be impro ved, guidelines for the collection and retention of the data required must be developed and industry encouraged to participate.