Ethnic group variations in alcohol-related hospital admissions in England: does place matter? A cross-sectional study

Abstract

BackgroundAlcohol is a major cause of disability and mortality in England. This health burden is known to differ both between ethnic groups and geographically. Little is known about the distribution of alcohol-related hospital admissions between ethnic groups living in different parts of the country. Ethnicity data coding has been near complete (>90% of admissions) in English hospital data since 2009–10. This study exploits this to examine variations in alcohol-related hospital admissions between ethnic groups in England and whether these vary by geographical location. MethodsData for this study come from Hospital Episode Statistics, an administrative database that contains data for all people admitted to a NHS hospital in England, for 2010–11. Alcohol-related diagnoses were classified according to the International Classification of Diseases and were in the three broad categories of mental and behavioural disorders due to use of alcohol, alcoholic liver disease, and admissions due to the toxic effect of alcohol. Geographical variations were examined by dividing England into the four regions of London, the south, the north, and the midlands. Ethnic categories were based on the Office for National Statistics classification. Differences between ethnic groups were examined with age-standardised and sex-standardised rates across the three major reasons for an alcohol-related admission, with denominator data from the Office for National Statistics mid-2010 population estimates. Rates were calculated separately for admissions in which the alcohol-related disorder was coded as a primary diagnosis or as a comorbidity. Rates were examined separately across the four different areas of England to ascertain variations in admission rates for people in the same ethnic groups living in different areas. 95% CIs were calculated on the basis of the Poisson distribution to assess whether rates differed. FindingsIn 2010–11, there were 264 870 alcohol-related admissions. Of these, 61 511 (23%) had an alcohol-related disorder as the primary diagnosis and 203 359 (77%) as a comorbidity. Diagnoses associated with behavioural and mental health made up 183 732 (69%), alcoholic liver disease accounted for 55 059 (21%), and the toxic effects of alcohol 26 079 (10%) of these admissions. For alcohol-related primary diagnoses, regional admission rates varied widely for people of the same ethnic origin. White British people had higher admission rates in the north of England than in the south (178 per 100 000 population, 95% CI 177–179 vs 70, 69–71). London had the highest admission rates for white Irish, Caribbean, African, and Indian people (eg, Indian people had an admission rate of 99 per 100 000 [95% CI 94–103] in London compared with 48 [44–53] in the north, and white Irish people had an admission rate of 306 per 100 000 [293–320] in London compared with 186 [173–199] in the north of England). Similar patterns were noted when an alcohol-related diagnosis was coded as a comorbidity. InterpretationEthnic group variation in alcohol-related admission rates differ substantially with geographical location in England. There are, however, some limitations of this study: the available data do not allow us to examine aspects such as religious association, and some patients might have had their ethnic origin incorrectly coded. However, ethnicity coding is now regarded as accurate in most cases. Some of the differences might be caused by variations in socioeconomic status, which was not examined by this study. Interventions to address the growing burden of alcohol misuse may need to be tailored to take the heterogeneity of associated morbidity into account. Finally, any population-level interventions (such as a minimum unit price) should be assessed for their effect on these ethnic group inequalities. FundingNone.