REGIONAL VARIATIONS IN DEATHS FROM VOLATILE SOLVENT ABUSE IN GREAT-BRITAIN

Aims. To assess geographical variations in mortality and the relations hip of socio-economic correlates to deaths from volatile substance abu se (VSA) in Great Britain. Design. Analysis of the National Register o f deaths from VSA by linking the addresses (postcode) of the deceased to census enumeration districts and hence wards and counties. Setting and participants. All 775 deaths in Great Britain listed in the nation al register between 1985-91. Population counts from the 1991 census we re used as denominators. Measurement. The Poisson heterogeneity test w as used to test the null hypothesis that all standardized mortality ra tios were from the same population. The t-test was used to compare dif ferences in Townsend Deprivation scores between wards with and without VSA deaths. Multiple regression was used to assess the relationship b etween indices of deprivation and deaths from VSA. Findings. One-third of all VSA deaths occurred in six regions of Great Britain which acco unted for 17% of the at risk population. There was considerable variat ion in age standardized mortality ratios (ASMR) across regions. A ward level analysis indicated that wards with a VSA death had significantl y higher levels of deprivation. The Townsend deprivation index showed mean deprivation scores of 2.8 and 0.2, respectively, for wards with a nd without a registered death (p < 0.0001). Multiple regression at cou nty/region level found that 45% of the variation in ASMR was explained by the components of the Townsend deprivation index. Conclusions. Fac tors associated with deprivation are important in VSA deaths and this should be taken into account in planning public health measures to pre vent mortality.