Human Immunodeficiency Virus (HIV) Repeat Testing and Seroconversion

Data Collection

Enhanced surveillance data between 2005 and 2012 in the Web-enabled Infectious Disease Surveillance System (WebIDSS) was examined for repeat assessment for bloodborne and sexually transmitted infections. Reported risk behaviours were tabulated. Records with repeat laboratory assessments for HIV^{Footnote 1}were reviewed for data quality and consistency. HIV seroconversion was defined as a negative serology test result followed by a positive serology test result. Data were extracted June 2016.

Analysis and Results

Repeat HIV Testing

Follow up testing is voluntary, and not all inmates require follow up. A total of 15,737 inmates had repeat assessments over a period of seven years (2005-2012).

Reported HIV Risk Behaviours

Inmates were asked about risk behaviours “since the last assessment”. Questions were not specific as to whether the risk occurred in prison or in the community. Roughly 40% of the records had risk information available for analysis (see Table 1).

±-Due to missing data denominator ranged between 6,057 – 6,681

Half of all inmates reassessed reported receiving a tattoo (53%) and twenty percent (20%) reported injecting drugs. Among those who reported injection drug use 50% reported sharing needles or other injection drug equipment (i.e., spoons and cookers).

HIV Seroconverters

A total of 7,166 inmates had repeat HIV laboratory tests available for analysis. Of these 17 were identified as seroconverters for HIV (positive laboratory results preceded by a documented negative test). The mean time between first negative test and subsequent positive laboratory result was 3.3 years (median 3.0 yrs, range 73 days to 7.0 yrs).

HIV Incidence Rate

Based on the total number of days under observation in this open cohort, the incidence of HIV was estimated at 1 case per 1,000 inmates at risk per year^{Footnote 2}.

Risk Factors for HIV Seroconversion

Risk ratios for the exposed and unexposed groups were determined and the relative risk (RR) was calculated. Relative Risk was used to estimate the population attributable fraction (proportion of new cases attributed to that risk factor). These data are summarized in Table 2.

±-Population Attributable Fraction

Because the numbers are small (due to missing data the number of HIV seroconverters included in the bivariate RR is n=11) the confidence intervals around each point estimate are wide. None reach statistical significance. However it is interesting to note the magnitude of the RR estimates for risky sexual behaviour, consistent with known transmission pathways.

Community Exposure

Preliminary investigation into the 17 HIV seroconverters indicated that 9 or 53% had spent some time in the community, ranging from 2% to 90% of the time interval under observation. Offenders may reengage in risk behaviours and could acquire HIV via community exposure.

Summary

Offenders in CSC seroconvert to HIV positive status at an estimated rate of 1 per 1,000 inmates at risk per year. Not all inmates are at risk, and testing uptake is high. Roughly half of the seroconverters spent time in the community. Understanding the attribution of risk behaviours in prison versus community requires more study.