Removing hepatitis C antibody testing for Australian blood donations: A cost‐effectiveness analysis

Abstract Background and Objectives The risk of transfusion‐transmitted hepatitis C virus (HCV) infections is extremely low in Australia. This study aims to conduct a cost‐effectiveness analysis of different testing strategies for HCV infection in blood donations. Materials and Methods The four testing strategies evaluated in this study were universal testing with both HCV antibody (anti‐HCV) and nucleic acid testing (NAT); anti‐HCV and NAT for first‐time donations and NAT only for repeat donations; anti‐HCV and NAT for transfusible component donations and NAT only for plasma for further manufacture; and universal testing with NAT only. A decision‐analytical model was developed to assess the cost‐effectiveness of alternative HCV testing strategies. Sensitivity analysis and threshold analysis were conducted to account for data uncertainty. Results The number of potential transfusion‐transmitted cases of acute hepatitis C and chronic hepatitis C was approximately zero in all four strategies. Universal testing with NAT only was the most cost‐effective strategy due to the lowest testing cost. The threshold analysis showed that for the current practice to be cost‐effective, the residual risks of other testing strategies would have to be at least 1 HCV infection in 2424 donations, which is over 60,000 times the baseline residual risk (1 in 151 million donations). Conclusion The screening strategy for HCV in blood donations currently implemented in Australia is not cost‐effective compared with targeted testing or universal testing with NAT only. Partial or total removal of anti‐HCV testing would bring significant cost savings without compromising blood recipient safety.


INTRODUCTION
Hepatitis C virus (HCV) is a bloodborne virus of global public health concern. In Australia, the main transmission route is the sharing of needles among people who inject drugs. Globally, transmission via the reuse or inadequate sterilization of medical equipment and the transfusion of unscreened blood and blood products are important considerations. Australian Red Cross Lifeblood (Lifeblood) is responsible for the collection and distribution of blood and blood products in Australia. Collections are made from voluntary non-remunerated donors in Australia. HCV antibody (anti-HCV) testing is currently used in parallel with HCV RNA nucleic acid testing (NAT) to detect current or past HCV infections. With approximately 1.6 million blood donations collected annually [1], no transfusion-transmitted HCV infections have occurred with current testing [2].
A previous international modelling study demonstrated that the additional blood safety provided by anti-HCV testing is minimal with universal NAT [3]. However, universal donor testing with both anti-HCV testing and NAT continues not only in Australia but also in other developed countries [4,5]. The current screening strategy is considered as a 'belt and braces approach' primarily to mitigate the remote risks of test failure. While the rate of HCV infection in first-time donors is higher than repeat donors [6], approximately 90% of Australian blood donations are contributed by repeat donors [1]. In addition, plasma for further manufacture collections for plasmaderived medicinal products has been steadily increasing and now outnumbers whole blood collections. The fractionation process includes pathogen reduction steps that substantially reduce the HCV residual risk, so anti-HCV testing in this context is unlikely to provide any clinically relevant safety benefit.
The shift in the management of chronic hepatitis C provides further justification for reconsidering HCV donation testing strategies.
The use of dual testing was adopted at a time when direct-acting antiviral (DAA) therapy was not available, and the majority of those infected by HCV would progress to chronic hepatitis C. In the past, testing donors for HCV reduced the incidence of severe conditions such as liver failure, as well as avoiding significant costs associated with managing these conditions [7]. With the advent of DAA, which has demonstrated a cure rate of over 95% [8], in the vast majority of cases, diagnosed HCV infection can be cured.
Given the context of finite healthcare resources and the limited incremental risk-benefit contributed by anti-HCV testing, a targeted anti-HCV testing strategy or no anti-HCV testing may be favourable over current testing. Removing or targeting anti-HCV will lower the total costs of HCV screening for blood donations, but how this will affect the residual risk of HCV infection and the long-term costs and health outcomes requires investigation. Cost-effectiveness analysis is such a tool that can incorporate all available evidence and assist decision-making in the transfusion medicine [9,10]. Previous studies have investigated the cost-effectiveness of adding NAT to anti-HCV testing in blood donation [11][12][13], but the cost-effectiveness of removing anti-HCV testing or applying anti-HCV testing dependent on donor risks, remains unknown. Therefore, this study aims to conduct a cost-effectiveness analysis of different testing strategies for HCV infection in Australian blood donations.

Alternative HCV testing strategies
In this study, four HCV testing strategies were proposed for comparison: 1. Universal testing with both anti-HCV and NAT (status quo).

Decision-analytical modelling
The decision-analytical model for assessing the cost-effectiveness of alternative HCV testing strategies consisted of a decision tree model ( Figure 1) and a Markov model ( Figure 2) implemented in TreeAge Pro 2021 [14]. The decision tree started with one of the four alternative testing strategies. Following each testing strategy, there was a chance that the transfusion recipient was infected by HCV, which was determined by the residual risk estimate from a separate analysis (see Table 1, [6]). Once the recipient developed acute hepatitis C, it was assumed that the condition would either clear spontaneously or progress to chronic hepatitis C. For blood transfusion recipients who have achieved blood safety (i.e., no infection transmitted) or whose acute infection cleared spontaneously, it was assumed that they would survive or die of any causes in the following years.

Risk of transfusion-transmitted HCV infection
In this study, the residual risk refers to the risk that a donation from an HCV-infected donor is not detected by testing, leading to a transfusion recipient becoming infected with HCV. The residual risk for different testing strategies was estimated based on an in-house purpose-built risk model [6]. The more conservative mid-estimates were used in the baseline analysis (Table 1).

Transition probabilities
The proportions of first-time donations and transfusible component donations were taken from Lifeblood 2020 donation data ( Table 2, [6]).
The spontaneous viral clearance rate in acute infections was derived from a systematic review where the proportion achieving clearance within 12 months following infection was 0.36 [15]. The transition probabilities associated with chronic hepatitis C disease progression were sourced from previous modelling studies [16][17][18][19]. The mortality rates for non-cirrhotic stages and compensated cirrhosis were based on the mortality rates of blood transfusion recipients, who have higher mortality than the general population (Table S1). Elevated mortality rates were assigned for 'decompensated cirrhosis' and 'hepatocellular carcinoma' health states using published estimates [23]. The mortality rates following liver transplant were taken from a project that investigated the future health and economic burden of hepatitis C in Australia [24]. As there is no universal testing for HCV among the general Australian population, we assumed that for non-cirrhotic stages, only 1% (probability 0.01) of blood transfusion recipients would get tested for HCV each year.

Resource use and costs
This cost-effectiveness analysis was conducted from a healthcare system perspective where only direct costs of providing testing and treatment of HCV infection were considered. The costs of testing HCV in blood donations were the rolled-up costs including the costs of tests and labour. As NAT is a multiplex test that also tests for human immunodeficiency virus (HIV) and hepatitis B virus in blood donations, it should be noted that the costs of NAT listed in Table 2 are not just for hepatitis C alone. In the baseline analysis, the total costs of NAT were used as the costs of NAT for HCV. An alternative assumption that the costs of NAT for HCV were a third of the total costs of NAT was tested in the sensitivity analysis. The costs of managing different stages of chronic hepatitis C were sourced from an Australian study that assessed the cost-effectiveness of treating people who inject drugs with DAA therapy [20]. All cost items were valued in 2020 Australian dollars.

Health outcome measure
The health impact of HCV infection was quantified using qualityadjusted life years (QALYs). QALYs are calculated by multiplying the utility weight associated with a health state by the number of years lived in that state. The utility weights range between 0 and 1, with 0 representing death and 1 representing full health. The utility weight for the Australian general population was used for those blood transfusion recipients who were not infected by HCV [25]. The utility weights for different stages of chronic hepatitis C were informed by a recent systematic review and meta-analysis of health utilities in patients with chronic hepatitis C, including people who are treated with DAA [26]. As the chronic hepatitis C condition deteriorates, the utility weight would decrease accordingly. Those infected but achieving SVR were assumed to experience improved quality of life, but the utility weight would still be lower than for the general population. We also applied a disutility of 0.1 to the utility weights to all health states to account for the impact of health conditions that required transfusion.

Sensitivity analysis
To account for the model parameter uncertainty, we conducted a sensitivity analysis to assess the impact of varying parameter values on model outputs. Moreover, we conducted a threshold analysis to determine the parameter values required for the testing strategies to become cost-effective.

Baseline analysis
The expected costs and QALYs associated with 1000 blood donations under different testing strategies are presented in Table 3. As the residual risks are so low, the number of potential cases of acute hepatitis C and chronic hepatitis C approximates to zero in each scenario.
Thus, the costs of managing HCV infections had virtually no impact on the total costs (which were determined by the costs of testing), and the total QALYs contributed by uninfected blood transfusion recipients were the same for all four testing strategies. As a result, universal testing with NAT only is the preferred strategy in our analysis as it had the lowest testing cost for all three age groups.
Based on the actual number of blood donations in 2020 and 2021, the costs of testing blood donations for HCV using different strategies were calculated and are presented in Table 4. The costs of dual testing were estimated to be A$23 million in 2020 and increased as the volume of blood donations increased in 2021. If universal testing with NAT only were to be implemented, the total costs of testing T A B L E 2 (Continued)

Sensitivity analysis
Given that the background risks of HCV transmission are extremely low, varying residual risks and changing the value of parameters related to managing HCV infection had virtually no impact on the baseline results. We greatly increased the residual risk of other testing strategies in the threshold analysis. The results show that for the current practice to be cost-effective, the residual risk of other testing strategies would need to be at least one HCV infection per 2424 donations, which is over 60,000 times the baseline estimate for residual risk (1 in 151 million, Table 1).

DISCUSSION
To our knowledge, this is the first published study to assess the costeffectiveness of removing anti-HCV blood donation testing. We tested different scenarios where NAT was applied solely for repeat donations, fractionated donations or all blood donations. Given that the residual risks of acquiring HCV following blood transfusion are extremely low for each proposed testing strategy, our modelling predicts that almost no one will develop acute or chronic hepatitis C, incur substantial costs associated with managing hepatitis C and experience reduced quality of life. As a result, the cost-effectiveness of different testing strategies is almost completely determined by the cost of testing. Therefore, strategy 4 (the NAT-only testing) is clearly the optimal strategy due to its lower testing cost.
Although we are not aware of published studies assessing the impact of removing anti-HCV testing from blood donation screening, our finding that a single test is more cost-effective than dual testing is consistent with previous cost-effectiveness studies [11][12][13]. These studies assessed the cost-effectiveness of adding NAT to serological (antibody and antigen) testing for blood donations and all reported that adding the additional test would not be cost-effective. This is because using a serological test alone already reduces residual risks to a very low level. Adding NAT further lowers the residual risk but the additional reduction in viral transmission is minimal, while the additional cost of testing is significant. Similarly in our case, although removing the anti-HCV test would result in a slightly higher residual risk, the impact on total costs of managing HCV infections is negligible given an already very low background risk. The results from our threshold analysis also showed that the background risk needs to be elevated to a very high level for the dual testing strategy to be costeffective. It should be noted that our study did not include a scenario where anti-HCV alone was used for screening. The residual risk of using anti-HCV alone is many fold higher at 1 in 800,000, as estimated by the Lifeblood internal modelling. In addition, NAT also tests for hepatitis B and HIV. A strategy with anti-HCV testing alone was, therefore, not considered a realistic option in the Australian setting. Cost-effectiveness analysis is one important assessment in considering blood safety risk management but risk-based decisionmaking includes other assessments, including stakeholder view, reputational risk and ethics [29,30]. Historically, blood operators have tended to risk-mitigate at any cost, but with risk-based decision-making principles, blood operators are moving to risk reduction. One recent example includes a transfusion-transmitted hepatitis C case in Germany that occurred with minipool testing that would likely have been prevented by ID-NAT [31]. Despite this, the authors concluded that it remains a very rare event, and the implication is, therefore, that the risk is considered tolerable. Our cost-effectiveness analysis clearly concludes that the screening strategy for HCV in blood donations currently implemented in Australia is not cost-effective compared with targeted testing or universal testing with NAT only. Partial or total removal of anti-HCV testing would bring significant cost savings without compromising blood supply safety. The purpose of risk management is not to eliminate risk but to use resources appropriately to minimize or accept the risk, and with overwhelming evidence of ineffectual resource use in our case, this provides a strong argument to cease or change anti-HCV testing.
Lifeblood to provide blood, blood products and services to the Australian community.
V.W., Q.C., V.C.H., C.R.S. and P.B. designed the research study; Q.C., S.T.F.S., J.A.K. and R.T.G. developed the economic models; Q.C. analysed data and wrote the first draft; all authors reviewed and edited the manuscript; V.W. supervised the research.