Good evidence begets good policy: or so it should be

Abstract

Policy analyses that seek to place blood safety interventions into a long-term health outcomes context are not new. Evidence-based decision making asks for more rather than less information, including evaluation of the entire set of costs and benefits that can be expected from a given intervention. However, as the evidence base grows, the choices become more difficult and implementation of measures to mitigate threats more complex. Perhaps this problem is even more acute during a time when the cost of health care, austerity measures, and the implications of health care reform in the United States place an ever more exacting demand on the way decisions are made and where we collectively choose to focus. The article by Borkent-Raven and colleagues in this issue of TRANSFUSION evaluates the cost utility of infectious disease screens in the Netherlands.1 Nucleic acid testing (NAT) for human immunodeficiency virus (HIV), hepatitis C virus (HCV), and hepatitis B virus (HBV) is evaluated as well as human T-lymphotropic virus (HTLV) screening and hepatitis A virus (HAV) screening. The analyses the authors have conducted, including the methods and data sources they have used, are ambitious and deserve to be considered in both a focused blood infectious disease screening and a broader public health perspective. Using mathematical modeling techniques that capture the statistical life histories of blood recipients and the components they receive, the authors have shown that screening all blood donations for HTLV has very limited utility with a result of €45.2 million (US$59.3 million at current exchange rates) per quality-adjusted life-year (QALY) saved. This result is the most extreme finding but none of the interventions or scenarios analyzed costs less than €2.23 million (US$2.94 million) per QALY. As part of the analysis, the annual testing cost for each intervention in the Netherlands is also provided. These monetary amounts may have more traction for many. These data when considered alone are examples of another type of important health economic analysis methodology, budget impact analysis. Budget impact analysis will answer how much it costs to adopt an intervention, while cost-effectiveness will answer how much health improvement can be achieved per monetary unit spent. Both types of economic data should be used in decision making. The importance of these two different types of economic analyses study results has been discussed previously.2 In many developed-country settings, the budget impact of new safety interventions may be tolerable, but it is important to investigate if the interventions prevent disease and improve health. In the case of the cost-effectiveness results reported by Borkent-Raven and colleagues, the euros spent are not substantially contributing to overall health of Dutch society. Are these results applicable to other settings? Maybe. The cost-effectiveness of blood safety screening in the United States and elsewhere will be different for a number of reasons, including differences in disease epidemiology, donor eligibility rules, the cost of testing, and health care costs. While many health economists and policy analysts have offered appropriate cautions when comparing studies conducted at different times or in different locales, it is still worthwhile to assemble cost-effectiveness results for adopted blood safety screens to observe trends and to place the new findings in context. For example, studies conducted more than 10 years ago often did not include survival probabilities in the transfused population and rather relied on life tables from the general population. Virtually all studies published in the last half of the 2000s have corrected this error. The use of general population life table data will overestimate the cost-effectiveness of blood safety screens (smaller, more favorable ratio of cost per QALYs saved). Table 1 provides key results for two settings, the United States and the Netherlands. This summary table obscures details of each analysis. Furthermore, these data are not intended to be a complete listing of all studies published in both countries, but represent a short list of published peer-reviewed analyses focused on infectious disease screening of donated blood. A few important concepts are evident. First, it is not simply a case of the order of adoption that influences whether an intervention is relatively cost-effective. Second, for infections with multiple interventions, such as serology and NAT for HIV, HCV, and HBV, the increment of infections identified by the second set of interventions adopted will virtually never be cost-effective. A willingness to reevaluate and potentially eliminate adopted interventions when considering new interventions for the same pathogen should be embraced by everyone. Third, the actual cost-effectiveness of any intervention is probably worse than reported in the table because modeled rates of transfusion-transmitted infections tend to be higher than observed rates of actual transmission, and most analyses assume 100% transmissibility by blood products. Therefore, most of the results presented in the table should be considered the best case cost-effectiveness ratio for each of the interventions evaluated. Are cost-effectiveness studies truly valuable in helping to inform policy decisions in blood safety? The evidence would suggest the answer is no. Advancements in the methodologies used to conduct analyses and the sophistication of analyses have not appeared to change the decisions that are made with respect to blood safety interventions. Efforts to understand the role of cost-effectiveness analysis in blood safety have been undertaken by several authors.3-7 The list of analyses that have had little impact on blood safety decision making is long. The obvious and often stated lesson is that thresholds for acceptable costs per QALY gained that may be applicable in other health care settings simply do not (and may never) apply in blood safety. This observation is likely a reflection of a broader societal desire to prevent transfusion-transmitted infection and appears to be applicable regardless of type of health care system (single-payer government-funded, market-driven, or somewhere in between). Even in developing country settings, acceptable cost-effectiveness ratios for blood safety exceed those of other health care interventions. There are many reasons why cost-effectiveness and budget impact analyses do not have more influence on the blood safety interventions that are adopted. At least for the United States, there is probably one primary reason—we do not use this information because we do not have to. If, as with pharmaceuticals in many settings,8 blood reimbursement decisions were linked to some combination of budget impact and cost-effectiveness, it is likely different blood safety intervention choices would be made in which the explicit inclusion of cost-effectiveness analysis results would be clear. We should be critical of the status quo, particularly with respect to potential improvements in the overall health of the transfused population as resources are expended seeking to prevent the transmission of rare infections. It is true that we must remain vigilant and strive to prevent transfusion transmission of clinically significant infections to recipients, but the decisions on how to screen for any one infection or even what infections to target appear to be based on the availability of screening tests with regulatory approval. This situation is made more complex by the recognition that availability of tests may be guided more by manufacturers' market and profit potential rather than public health need. Screening to prevent Babesia in the United States, a clear transfusion-transmitted infection,9 is long overdue, and targeted screening in highly endemic areas would likely be cost-effective if an approved assay were available. Cost-effectiveness and related studies remain relevant. Even if they do not specifically contribute to decision making in blood safety, they do serve the important function of being a reality check. Frequently, results show that we probably could find better uses for available blood safety resources. Even if we are not compelled to choose interventions that provide the most gain in terms of QALYs saved, publication of these studies prevent us from claiming we did not know the minimal health gains for blood recipients for each dollar or euro spent. Blood safety at nearly any cost appears to be alive and well. It remains to be seen how long this economically irrational state will persist. None.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.