Transfusion‐transmitted kala‐azar in India

evaluated through large-scale clinical trials in blood bank settings. Although none of us were professionally involved in transfusion medicine during the time referenced in Dr. Bettigole’s letter, it is our understanding that consideration of the above factors led to the prospective follow-up studies of the late 1970s and early 1980s that produced data defining the incidence of and morbidity associated with non-A,non-B posttransfusion hepatitis and identified it as a disease for which a partial reduction in transmission would be worthwhile.2,3 The second issue raised by Dr. Bettigole has to do with the setting of cutoff values for unit acceptance. At first glance, it may appear that this consideration applies only to ALT testing, as this test generates a numerical value along a continuous scale rather than a positive or negative interpretation, as is the case with enzyme immunoassays. However, enzyme immunoassays also produce values on a continuous scale, which requires that cutoff value decisions be made. The difference is that such decisions are made by the manufacturer and the Food and Drug Administration rather than by blood centers. The aim of those decision makers is to strike a balance between the competing aims of optimizing both test sensitivity and specificity. For some tests, such decisions are appropriate at the time the test is implemented but should be modified as additional data are accumulated and as circumstances change. For example, many believe that the cutoff for blood donor screening for antibody to hepatitis C core antigen should be altered now that hepatitis C virus antibody testing is routinely p e r f ~ r m e d . ~ , ~ The issue of selecting optimal cutoff values also applies to ALT testing in the era of hepatitis C virus antibody screening. The data in our study can be used to reset the ALT cutoff value if such testing is to be continued. Our data show that setting the cutoff value at a higher level (i.e., twice the previous cutoff) will improve the tradeoff between sensitivity and specificity, although, in our opinion, not to a level that warrants retention of ALT testing.6 This cutoff change has been adopted, in the interim, by the many US blood banks that still continue ALT testing because of the European requirement for such testing of plasma manufactured into derivatives. Given the profound implications for donors and recipients and the impact on transfusion economics, we strongly believe that decisions regarding the implementation or elimination of donor screening tests should be made after a thorough analysis of test performance, predictive value, and cost-effectiveness. We hope that those making future test implementation decisions, both for previously identified and newly discovered agents, will carefully consider the full set of issues outlined in this letter.