[Mortality of drug users attending public treatment centers in Italy 1998-2001: a cohort study].

OBJECTIVES

to describe the overall and cause-specific mortality among heroin users attending Public Treatment Centers (PTCs) in Italy and to estimate the impact of heroin use on mortality in the general population.

SETTING AND PARTICIPANTS

A cohort of 10,376 patients (8881 men and 1495 women) enrolled over a period of 18 months between september 1998 and september 2000 and followed-up through 31st of March 2001 (VEdeTTE study).

RESULTS

190 deaths occurred during the study period (153 men and 37 women): 70 deaths were due to overdose (36.8%), 38 to AIDS (20.0%), 30 to violence (15.8%). The direct standardized overall mortality rate per 1000 person/years is 12.0 (CI 95% 5.4-18.6): 12.7/1000 p-y (CI 95% 4.9-20.5) among males and 8.4/1000 p-y (95% CI 4.7-12.2) among females. This study confirms that overdose is the leading cause of death in heroin users (mortality rate 2.6/1000 p-y (95% CI 0.8-4.5) among males and 4.0/1000 p-y (95% CI 0.9-7.2) among females. AIDS mortality rates are 2.6/1000 p-y, 95% CI 0.6-4.6 among males and 1.8/1000 p-y (95% CI 0.4-3.1) among females. The mortality rate for all the other causes is 6.0/1000 p-y (95% CI 0.0-14.0) among males and 2.3/1000 p-y (95% CI 0.9-3.6) among females. The standardized mortality ratios for all causes in comparison to age and gender matched general population show the excess particularly important for females (SMR 6.7; 95% CI 5.7-7.8 for males and SMR 22.8; 95% CI 16.5-31.5 for females). The population attributable fraction highlights that 14.4% (95% IC 10.9-18.5) of deaths in people aged 30-34 in Italy in 2000 could be attributed to heroin addiction; the fraction decreases to 10.7% (95% CI 6.9-15.6) at age 25-29 and to 12.8% (95% CI 9.9-16.2) at age 35-39.

CONCLUSIONS

Mortality observed in this cohort is lower than that observed in previous studies, mainly due to reduction of AIDS and overdose mortality. The excess mortality over matched population is confirmed. Study population is older than in other studies (mean age at enrollment 31.1; DS 6.2); and the observation time is mainly spent in treatment. The mortality attributable faction shows that almost the 13% of deaths around the third decade of age can be attributed to drug dependence even though it is important to take into consideration the assumptions about drug addiction prevalence in the general population.