In Focus
A web-based tool for modelling the residual risk of HIV transfusion transmission
An open-source web-based model and software tools on which others can build
Part of the work of the TTID WP is to provide resources and expertise for understanding the current situation regarding known and emerging threats to the safety of the blood supply.
The risk of HIV in the blood supply is a known threat. An emerging issue is the risk because of antiretrovirals used to treat or prevent HIV infection. The expanding availability of pre-exposure prophylaxis (PrEP) in many countries raises questions on the impact on blood safety. Where blood donations are screened by nucleic acid testing (NAT) and antigen/antibody tests, HIV transmission by blood transfusion has become very rare. The remaining (residual) risk arises mainly from donations made early in infection, during the infectious window period, when a donor is infectious but before the infection becomes detectable. The increasing use of antiretroviral drugs to prevent HIV infection (pre-exposure prophylaxis or PrEP) poses a challenge since HIV breakthrough infections exhibit different dynamics in viral and serological markers than infections occurring in the absence of antiretroviral drugs.
Figure 1. Conceptual model of the risk of transfusion transmission associated with PrEP use.
The risk can be conceptualised as a series of sequential probabilities combined to estimate the residual risk (Figure 1). First, an individual must be using PrEP. Second, the person must seek to donate blood. Third, during the donor selection process the person must fail to disclose PrEP use (where this use would disqualify a donor) or other deferrable risks. Fourth, the person must have acquired HIV infection for reasons such as inadequate adherence in taking PrEP or true failure of the PrEP, for example in the unlikely situation where the HIV infection is with a drug-resistant strain. If this combination of events occurs, then the infectious window period and the likelihood that the blood products from the donation have enough HIV to cause transfusion-transmitted infection become important.
We developed a web-based modelling tool for estimating the residual risk of HIV transmission through transfusion, allowing for the estimation of the infectious window period, separately in the absence of antiretroviral drugs and for the scenario of PrEP breakthrough infection. In the web-based model, daily oral PrEP formulations and long-acting injectable PrEP formulations are modelled separately, since these impact HIV disease progression and biomarkers differently. The web-based tool and additional resources can be found at https://residualrisk.org [see Footnote].
Transfusion transmission risk does not depend simply on whether HIV is detectable at a particular point in time. Rather, it is the overlap between the period during which an infected individual can transmit HIV and the period during which available screening assays fail to detect infection (Figure 2). This modelling approach is derived from the framework developed Weusten and colleagues1, which allows one to estimate the likelihood of an infection being missed by NAT screening given a particular viral load and the performance characteristics of the specific NAT assays (limits of detection), as well as NAT pooling strategy. The risk of transmission of HIV by transfusion in blood products additionally relies on the quantity of virus transfused, see Belov and colleagues2.
Figure 2. The infectious window period
In a recent review with colleagues3, the implications of undisclosed HIV treatment and prevention with antiretrovirals for blood safety and an approach to modelling risk, are discussed. There are several unknowns, but it is known that the presence of these drugs during early infection can suppress viral replication and therefore alter the appearance and evolution of HIV RNA, p24 antigen, and anti-HIV antibodies on which blood donation screening relies. Other residual risk models consider only NAT, since viral RNA normally becomes detectable well before antibodies. On PrEP, however, a donation may test negative on both. We therefore added serological screening to the model, with antibody detection delayed by weeks to months in the case of oral or injectable PrEP breakthrough infection. As with other model parameters, the user can specify distributions from which these PrEP impact parameters are sampled during simulations.
This tool is implemented as an open-source web-based model and software tools on which others can build. The web tool provides an interface that is easier to use and an accessible way to explore the impact of different assumptions on the risk of HIV transfusion transmission in different settings.
Footnote: Development of the web-based tool was sponsored by Vitalant Research Institute. I thank Brian Custer for his leadership and guidance, and additionally acknowledge the contributions of Michael Busch, Marjorie Bravo, Vivian Avelino-Silva and Renata Buccheri; Artur Belov provided access to transmission model posteriors. Members of the U.S. TTIMS program contributed to conceptualisation of the base model.
References
- Weusten J, Vermeulen M, van Drimmelen H, Lelie N. Refinement of a viral transmission risk model for blood donations in seroconversion window phase screened by nucleic acid testing in different pool sizes and repeat test algorithms. Transfusion. 2011;51:203-15.
- Belov A, Yang H, Forshee RA, Whitaker BI, Eder AF, Chancey C, et al. Modeling the risk of HIV transfusion transmission. J Acquir Immune Defic Syndr. 2023;92:173-9.
- Buccheri R, Grebe E, Custer B. The potential impact of undisclosed use of HIV treatment and prevention antiretrovirals on blood safety. Transfus Med Rev. 2026. doi:10.1016/j.tmrv.2026.151045.
