a1 Medical Research Council Clinical Psychiatry Unit, Graylingwell Hospital, Department of Community Health, University of Nottingham
Suicide rates in towns with a Samaritan branch were compared with rates in matched control towns without a branch. Four methods of choosing controls are described, all of which have advantages over those used by Bagley (1968). No statistically significant differences between Samaritan and control towns were found. The scientific case for the effectiveness of the Samaritans in reducing the suicide rate is therefore seriously weakened.
c1 Address for correspondence: Mr C.Jennings, MRC Clinical Psychiatry Unit, Graylingwell Hospital, Chichester, West Sussex, PO19 4PQ.
2 Formerly of University of Nottingham, now at Health Research Unit, British Rubber Manufacturers' Association, Birmingham 1.