Cost-effectiveness of contact tracing versus screening to find syphilis cases: further study is needed.

Abstract

To the Editor: Oxman and Doyle1 are to be congratulated for their careful study, published in the January–February issue of this journal, addressing a topic on which data are sorely needed. However, because of a number of potential biases and an inappropriate choice of indicators of outcome effectiveness, we think the results must be considered with great caution and may not be representative of syphilis control programs in areas of historically high syphilis incidence. There are several reasons to think the program studied may not be representative of most of those in the United States, or at least of those in the southern states. The stated indices of Disease Intervention Specialist (DIS) effectiveness in the Multnomah County, Portland, Oregon program (contact index, 1.75; percent of contacts examined, 50%) would be considered low by most managers. The comparable indices for the state of South Carolina, for example, are 2.07 and 76%. This suggests that DIS effectiveness in the program studied may be unrepresentative of most STD programs. Additionally, the stated cost of testing per person tested (presumably including clinic and laboratory costs) of $2.88 seems far too low; the South Carolina state laboratory's charge to the Health Department for a rapid plasma reagin test alone was more than $4.50 in 1990, not counting confirmatory tests or clinic costs of drawing blood. Finally, although Oxman and Doyle state that in Portland fewer than 10% of patients with early syphilis were identified by early contact tracing, in South Carolina in 1995, 32% of patients were found that way. The authors' inclusion of both “volunteer” (walk-in) testing and true asymptomatic screening testing under the same category of “testing self-motivated individuals” is not appropriate. Clearly, case finding by testing volunteers is likely to be the most cost-effective method in any program because those patients are symptomatic and are likely to have syphilis; similarly, unlike true screening or contact intervention, program staff have no choice about testing such patients. The appropriate comparison in this study would be between DIS contact intervention and true screening of asymptomatic persons. Experience shows that asymptomatic screening usually yields a lower proportion of new syphilis cases than either testing volunteers or testing sexual contacts, and patients found by screening usually have a later stage of syphilis. If the cost effectiveness of screening asymptomatic persons had been measured separately, it would almost certainly have been higher. Including interviews of suspects and associates, as well as contacts of patients in the Multnomah County program, biases the cost effectiveness of DIS activities too high because the yield of new patients almost certainly will be lower than the yield from interviews of true sexual contacts. For this reason, suspects or associates are not routinely interviewed in most programs. Of greatest concern, however, is whether testing and DIS contact intervention are equally likely to bring to treatment those infected persons who probably will infect others, i.e., the “core” or high-frequency transmitters who should be our primary target. The authors show that in their program, contact intervention was no more likely than “testing” to find patients in their primary and secondary stages, but that simply may be the result of a weak program rather than evidence of intrinsic lower efficacy of the method. An indicator of a “core transmitter” might be the number of contacts named or the number of infected contacts per case. Patients in South Carolina from 1990 to 1995 more often named at least two contacts if they were found by contact intervention (57%) or as volunteers (42%) rather than if they were found by screening (38%). Oxman and Doyle state that contact tracing in their program yielded only 0.2 new cases per index case, strongly implying that they were not finding many core transmitters. That does not mean that other programs in the United States do no better. In summary, this study does not provide enough representative data to allow judgment on the relative cost effectiveness of contact intervention versus screening versus other community-based approaches in controlling syphilis transmission in the United States.