Hospital morbidity associated with the natural history of heroin use

To assess changes in hospital morbidity associated with heroin use from pre-initiation, recreational to dependent use, and changes following treatment.

Analysis of hospital admission data assessed over six, 6-month periods (before and after heroin initiation; two dependent heroin use periods; and post-oral and post-sustained release naltrexone treatment).

A sequential cohort of 139 patients for whom date of initial heroin use, regular heroin use, and two periods of heroin dependence (prior to treatment with oral and implant naltrexone) were known.

The West Australian Data Linkage System was used to assemble hospital admission data. Admissions were analyzed as follows: "all cause" admissions and then subgrouped as "mental health (MH) other," "MH opioid related," "MH non-opioid drug related," and "opioid overdoses."

The database identified 130 (94 percent) of the participants with 76 (59 percent) being male. The mean length of follow-up from first heroin use was 9.4 (SD 4.9) years. Significant increases in morbidity were not found in the periods pre-first and post-first heroin use, but were found from pre-heroin use to dependent use for "all cause," "MH opioid related, " and "opioid overdose" admissions. A significant decline in these measures was observed from the first dependent period to the post-implant period. "MH opioid related" admissions declined from the first to the second period of dependent heroin use.

Findings suggest that the movement to dependent heroin use increases hospital morbidity; that morbidity in the presence of treatment does not necessarily increase; and the treatment with a sustained release preparation may be more effective than oral naltrexone in arresting morbidity.