Citation

Saunderson T, Haynes R, Langford IH. J. Public Health Med. 1998; 20(3): 261-267.

Affiliation

School of Environmental Sciences, University of East Anglia, Norwich.

Copyright

(Copyright © 1998, Oxford University Press)

DOI

unavailable

PMID

9793890

Abstract

BACKGROUND: Given that a high proportion of undetermined deaths are thought to be suicides, the objective of the study was to compare the geography of the two verdicts, with emphasis on urban-rural differences, to determine whether the likelihood of a death being classified as suicide varies according to location. METHODS: Standardized mortality ratios (SMRs) were mapped for the local authority districts of England and Wales, using data aggregated over the period 1989-1992. Districts were classified according to population density, and SMRs calculated for each quartile. Finally, regression equations were obtained to model the relationship of suicide and undetermined death with a series of risk factors. RESULTS: Both male and female undetermined deaths showed a significant deficit in the most sparsely populated districts and a significant excess in the most densely populated districts. Female suicides were also significantly high in densely populated districts, but male suicides were highest in the most rural areas (SMR 110, 95 per cent confidence intervals 105-115). District variations in both verdicts were associated with the proportion of single-person households for both sexes, but male suicides and male undetermined deaths were associated with additional predictors, consistent with their distinct geographical distributions. CONCLUSIONS: Differences in suicide methods, the likelihood of communicating suicidal intent and, perhaps, variations in access to psychiatric services might be contributory factors to the observed differences between the verdicts. The results suggest that recorded suicides are a biased subset of actual suicides, particularly for males, and that the study of suicide verdicts alone is potentially misleading.

Language: en