Strategies for testing blood donors for West Nile virus

The recognition that West Nile virus (WNV) was transmissible by transfusion led to the rapid implementation of routine donor testing for viral RNA in the United States and Canada. Recent analyses have indicated that this intervention, like others used to enhance blood safety, does not meet currently accepted thresholds for cost-effectiveness. As reported by Kuehn,1 this raises the question of the use of alternate strategies for WNV screening. In her article, she reviewed the current status of WNV testing in the United States, with particular reference to two cost-effectiveness studies. In the first, Custer and coworkers2 analyzed a number of testing strategies involving temporal and geographic variations in the use of minipool and single-donation testing including the absence of testing, whereas Korves and associates3 analyzed strategies varying from the absence of testing, testing only for immunocompromised patients, seasonal testing, and testing methods as used currently.3 Custer and coworkers concluded that, although not the most cost-effective approach, the current mix of year-round minipool testing and targeted single-donation testing offered the best way of assuring blood safety. Korves and coworkers, on the other hand, favored those strategies that minimized or eliminated testing. Kuehn interviewed both authors and found that Korves had concluded that the financial benefits of preventing transfusion transmitted WNV did not justify the cost, whereas Custer had pointed out that the issue was more appropriately seen as one of the balance of risks and benefits. Although it is certainly of interest and relevant to question the reasons for the acceptability of disproportionately high cost-effectiveness ratios for blood safety, it is more important to ask whether alternate strategies for WNV screening are, in fact, viable. Strategies that appear intuitive may not be appropriate from the perspectives of regulation, cost, or logistics. WNV infection is seasonal in nature, although the season in the US is longer than many presume, with RNA-positive donations detected as early as May 1st and as late as December 1st. Although seasonal testing for WNV from June 1 to November 30 has been introduced in Canada, it is unlikely to be acceptable in the United States. This is because the WNV-free period is short—indeed, the virus is thought to be a risk to humans year-round in some areas. In addition, blood donors are often mobile and may be exposed in one area, but give blood at another, distant location. During the off-season, donors in Canada are tested if they have traveled to the United States, but a simple geographic exclusion of this type is not practical within the United States. Korves and colleagues suggested that safety could be achieved by relying on identification of suspected WNV infection with a donor questionnaire in times or places without testing, but it is clear that questioning donors about symptoms associated with WNV infection was of little value and therefore has now been eliminated as a regulatory requirement.4 Although the FDA does not actually require testing for WNV RNA at this time, the agency does have control over the way that testing is performed through their oversight of the IND process currently governing testing and through the licensure and use of approved testing systems. It is not clear that the FDA would permit modifications to the current program of year-round testing without appropriate measures to safeguard against break-through recipient infection. As stated in the perspective of Kuehn, FDA’s concern is the safety of the blood supply and as such the focus is risk versus benefit, not cost versus benefit. It is also important to examine the underlying assumption that reducing testing will materially reduce costs. Reagents themselves are priced according to usage volume and a reduction in usage will not generate a directly equivalent reduction in cost. It is difficult to predict the volume of reagents required for any given year; the trend of WNV infection in donors has not been monotonic. One must also consider the consequences of under- as well as overestimating the necessary number of tests. Also, testing requires a substantial overhead in terms of resources and personnel: overhead that is fixed in the short term and cannot be readily turned on or off. Similar arguments can also apply to geographically restricted testing. Because the length of the season and the location of outbreaks are not amenable to an accurate forecast, the capacity for maximal testing must be retained. Trigger strategies have been developed that appear to ensure safety,5,6 but the actual logistics of starting and stopping testing are complex and could lead to errors because computer systems in place in most blood systems cannot be modified to switch from “marker on” to “marker off.” The consequences of inappropriate testing (or more correctly, the absence of appropriate testing) and inappropriate release of an untested unit to a susceptible recipient must be considered. Indeed, this point has been made by AuBuchon:7“In blood collection agencies, the tightly controlled computer systems that are optimized to prevent erroneous release of potentially infectious components have difficulty accommodating testing for WNV only during a portion of the year, thus forcing use of the less cost-effective approach of year round testing.” Although the concept of testing only for immunocompromised patients does have a precedent in the case of cytomegalovirus (CMV), there are considerable differences. The underlying prevalence of CMV antibodies in the donor population, which is greater than 50 percent, makes universal testing impossible. Additionally, populations at risk for CMV disease are much better characterized than are those for WNV. Indeed, in the study by Pealer and associates,8 of the 14 transfused patients investigated as a direct result of WNV disease, only 8 had primary diagnoses that would have been formally associated with immune compromise. Logistically, maintenance of a dual-tiered blood inventory, WNV tested and untested, is yet another source of potential errors. In addition, it is now clear that, within the hospital environment, selective transfusion protocols create operational difficulties and frequently fail to meet their objectives. Patients who should receive specially treated or selected blood components do not necessarily get them.9 Consequently, Korves and coworkers’ suggested strategy neither would identify all patients at risk, nor would it assure that at-risk patients would receive WNV-tested components. Both Custer and Korves agree that uniform testing at the single-donation level is least cost-effective, but Custer ultimately favors the current approach of testing in minipools with focused individual testing where and when a trigger strategy justifies such action. The nationwide and time-independent “surveillance” system with blood donations in a minipool approach provides an effective and sensitive early warning of the emergence of human WNV infection in a given area—information that may not otherwise be available. In a recent analysis, Stramer and coworkers have shown that, of 1332 potentially infectious blood donations tested during the years 2003 to 2005, 22 percent, or 294 viremic units, were detectable only by single-donation testing.10 This targeted testing serves as a bridge strategy between no testing and very expensive year-round individual-donation testing throughout the country, and it may serve as a model of testing for new agents if the disease epidemiology is clearly related to specific geographic regions. This approach may not be the least costly but it does achieve blood safety and public health goals. The effectiveness of the current strategy is underlined by Montgomery and coworkers’11 analysis of the impact of WNV testing over the period 2003 to 2005. During the first of these years, when testing was restricted in many blood centers to minipools, a total of six cases of apparent transfusion-associated transmission of WNV occurred. In the subsequent 2 years, only one such case was identified. That case occurred in an area with a very high incidence of WNV infection, but before the blood system in that area had completed the implementation of their single-donation “triggering” system. Although the size of the epidemic was smaller in 2004 and 2005 than in 2003, it is nevertheless likely that the strategy of focused single donation testing has been associated with the absence of transmission.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.