3 August 2004

Mortality risk among new onset injection drug users

Addiction (2004) 99:946-954



Vlahov D, Wang C, Galai N, Bareta J, Mehta SH, Strathdee SA, Nelson KE.



Dear Colleagues,

This important and elegant piece of research demonstrates several crucial features of drug use in a group of 256 'early' injecting drug users over a 12 year period. Subjects were recruited from a variety of sources, mostly word-of-mouth in Baltimore, USA, all with a history of less than 2 years since first injecting drugs. The study's end point was death yet several other important findings on the natural history of drug use are also revealed.

At recruitment 70% were male, 94% were African-American and mean age was 30. HIV rate was 22%. There were 90% currently injecting, 25% more than once daily. Just 25% had ever had drug addiction treatment, only 2% on methadone maintenance treatment (MMT). The authors state that 'needle sharing and shooting gallery use were not uncommon'.

With 69 known deaths among the 256 during the 12 year study period, the overall mortality rate was 3.3 per 100 patient years. Thus on average, over 3% of the sample died each year of the twelve. Yet the rate was not at all even, showing a peak at 6 to 8 years which was around 8 times the 'expected' US mortality. At 2, 4, 10 and 12 years from recruitment, the mortality figures were about 4 times that expected for the same sex/age controls. These figures in turn were about 2 to 4 times the actual Baltimore City mortality data which appear to be worse than elsewhere in the country.

These findings are surprising as there is neither an early nor a late peak of mortality as some had suspected, but a higher risk of death around 6 to 8 years from initial injecting with levelling out again after that. Even if the HIV cases are excluded, the peaks and other trends persist, but with less accentuation.

Follow-up rates at 5 and 10 years were ~75% and ~60%. The researchers found at the 5 (and 10) year follow up (respectively) that 54% (48%) were still injecting, 5.6% (5.7%) heroin alone; 15.7% (11.3%) cocaine alone; 79% (70%) alternating cocaine and heroin. Important to note is that fully half of the subjects had ceased injecting. About 9% were in treatment, 3% (6%) on MMT, 6% (3%) abstinence based. Self reported health rating of 'good' or better was reported by 82% at 5 years and 94% at 10 years. Needless to say, these figures exclude the large numbers who had dropped out or died.

It is a tragic reality of the American health care system that it took ten years for methadone treatment to become available for 6.3% of the sample (initially 2% and 3.1% at 5 years). Knowing that more than 75% of the 256 were heroin users, the uptake of methadone and abstinence based treatments seems very low and probably reflects the lack of treatment available in Baltimore as well as the low socio-economic status of many of the subjects in this study. The death rates of MMT patients is generally less than 1% per annum. If only the other good citizens of Baltimore had acted on the impressive economic good sense of funding methadone and other treatments for addictions, many of the deaths reported here would have been avoided (and household insurance premiums would probably have been lower).

comments by Andrew Byrne ..