Safety at work for railway staff in Britain.

The recent series of railway accidents raised questions about the safety of British Railways. We have studied trends in accidents to railway staff over a 33-year period, 1954-86, which time has covered the pre- and post-modernization of the railways, in conjunction with large reductions in both staffing and traffic volume. The Railway Inspectorate, which has the duty of investigating and reporting on accidents occurring on railway premises, defines three kinds of accident: train, movement and non-movement. The numbers of railway staff killed and injured in each kind of accident have been studied in relation to the number of staff employed and train-miles run. Trends in the three accident rates derived thereby show quite different patterns: for most of the period under study, train accident rates have shown a linear increase and movement accident rates a linear decrease, whilst non-movement accidents exhibited no particular trend at all. In the last 5 years, however, both movement and non-movement accident rates have decreased sharply, whilst train accident rates have increased above the generally rising trend. Explanations for some of these trends and changes are suggested, but the reasons for others remain obscure.