Introduction: Referrals with increased risk behaviours who test negative for blood borne viruses (BBV) may be suitable organ donors. However, infections contracted during the window period of detection (the time between infection and serology tests becoming positive) can pose risk to recipients. Where infection risks are low, remaining on the waiting list may be more harmful than risk of BBV transmission. Australia has highly effective harm reduction programs, and may have lower BBV rates than international settings. We aimed to estimate the contemporary prevalence and incidence of HIV among increased risk groups in Australia, and thence the absolute risk of window period infection.
Materials and Methods: We performed a systematic review and meta-analysis. Studies reporting Australian HIV prevalence or incidence (confirmed serological status) in increased risk groups 2000-2017 were eligible, including men who have sex with men (MSM), intravenous drug users (IVDU), prisoners, commercial sex workers (CSW), percutaneous BBV exposure, or those with high risk sexual partners (known blood borne virus or increased risk behaviour). Pooled prevalence and incidence rates were estimated from available data. Where there were insufficient incidence studies, incidence was estimated from pooled prevalence data. We assumed the days since HIV infection followed an exponential distribution to estimate the probability of window period infection and subsequently the residual risk. From previous literature, we assumed the window period for ELISA was 22 days and eclipse period for nucleic acid testing (NAT) was 7 days.
Results and Discussion: We examined 10,496 Medline records and 61 other records. Studies were excluded at abstract (9856) or full text level (210) for reasons including self-reported or modelled data, overlapping study cohorts, or inappropriate risk group.  We included 32 studies (353,846 participants), with most studies in MSM, IVDU and prisoners. There were no relevant Australian data for percutaneous non-IVDU exposure to BBV. Pooled incidence was directly estimated from available MSM studies only. The risk of window period HIV infection was low among all risk groups (Table 1), with the highest risk among MSM (6.3 per 10,000 donors with negative HIV serology testing). Residual risk was reduced further with negative NAT than ELISA alone (2.0 per 10,000 with negative HIV ELISA and NAT). The residual risk of HIV transmission among Australian IVDU and prisoners was substantially lower than reported in similar international studies. Local data among CSW and high risk partners was scarce, but also suggested low residual risk.

Conclusions: Accurate estimates of local infection risk allows informed decisions when considering increased risk potential donors. Given the low absolute residual risk of HIV infection, increased consideration of donors with increased risk behaviours but negative testing may be one avenue to expand the donor pool.