Reducing volunteer bias: using left-over specimens to estimate rates of HIV infection among inmates in Ontario, Canada

AIDS. 1995 Jun;9(6):631-7.


Objective: To estimate the prevalence of HIV-1 infection among adult and young offenders admitted to remand facilities in the province of Ontario, Canada, by using a design that reduces volunteer bias.

Methods: A study using a modified anonymous HIV-surveillance design was conducted with urine specimens routinely collected from male and female entrants to all Ontario jails, detention and youth centres between February and August 1993. Information on sex, age, and history of injecting drug use was also collected. Urine was screened using a modified commercial HIV enzyme-linked immunosorbent assay kit and confirmed using a modified in-house Western blot assay.

Results: Data were obtained on 10,530 adult men, 1518 adult women, 1480 young male offenders, and 92 young female offenders. Urine specimens were available for 88% of new entrants. Of the entrants, 1% (n = 163) refused to have their urine used for research. Refusals were not associated with history of injecting drug use. Overall rates of HIV-1 infection were 1% for adult men, 1.2% for adult women, and 0% for young offenders. Both the rates of infection and prevalence of injecting drug use varied across facilities and geographic regions. Overall, 13% of adult men, 20% of adult women, 3% of young male offenders, and 2% of young female offenders reported a history of injecting drug use. Rates of infection were highest among self-reported injecting drug users. Rates of HIV were 3.6% for adult men and 4.2% for adult women who injected compared with 0.6 and 0.5%, respectively, for non-injecting drug users.

Conclusions: The use of unlinked left-over specimens is an important tool for measuring HIV-prevalence rates and should be encouraged. The results indicate that HIV rates are much higher among those entering prisons than in the general population. The pattern of HIV in Ontario prisons is similar to that reported in Europe and the United States. We are optimistic that these data will stimulate much needed efforts towards education and health promotion, and open the door to further research in Canadian prisons.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Acquired Immunodeficiency Syndrome / epidemiology*
  • Adult
  • Bias
  • Female
  • Humans
  • Male
  • Middle Aged
  • Ontario / epidemiology
  • Prevalence
  • Prisoners*