A prospective study of mortality up to eight years after starting treatment for alcohol and drug problems in Stockholm County: 2000–2008

Abstract

Background and aim: Research indicates an association between substance use and premature death and that misuse of illicit drugs is more closely linked to mortality than alcohol misuse. Given that these studies often are based on homogeneous treatment populations, we sought to analyse long-term mortality among both alcohol and drug misusers in a representative treatment system sample by examining: (1) excess death ratios (SMR, standardised mortality ratio) in comparison with the general population and (2) risk factors for mortality within the sample.

Method: Prospective study (N = 1659; 28% women) interviewed when starting treatment in Stockholm County, 2000–2002, and followed-up with regard to mortality up to 8 years after baseline. Analyses were based on death certificates and intake interview data (demographics, social situation/support, ICD-10 alcohol/drug dependence, treatment experiences). The strength of the study is the prospective design, that we have been able to link mortality to interview data, and to reach a heterogeneous treatment population.

Results: (1) SMR was 5.7 (no sex difference). (2) Logistic regression showed that being older, male, retired and having reported living with a substance misuser were identified as risk factors for mortality within the sample. Housing organised by authorities and no dependence on alcohol/drugs were protective factors. The mortality risk did not differ between alcohol and drug-dependent cases. Neither was homelessness, living situation (3 years) nor education predictive of mortality.

Conclusions: No difference regarding mortality risk between treated alcohol and drug-dependent patients in Sweden is found when controlling for age.

• Mortality
• treatment
• alcohol and drug dependence
• Sweden

Notes

Notes

1. Under-reporting of substance-related deaths occur in statistics (Gossop et al., 2002; Lenke & Olsson, 1990; Marshall & Taylor, 1994; Romelsjö et al., 1993).

2. Electronic source: Interested reader can refer about the quality of cause of death register by the National Board of Health and Welfare at: www.socialstyrelsen.se/register/dodsorsaksregistret/bortfallochkvalitet, last viewed 2 December 2010.

3. This sub-study was added after the original sampling of recruitment sites and the exclusion of these 102 respondents does not have an effect on the representativeness of the analysed sample.

4. As a reference, in 2007 about 5% of all deaths in Sweden were unnatural in character (Socialstyrelsen, 2009).

5. The analysis also included other measures of social support (not presented – ADM within the participant's social circle, serious conflicts with family members/intimates), but these had no statistically significant effects and were excluded from the models.