Sir—As demonstrated by Pezzotti et al. combining information on HTV and AIDS is useful in estimating the incidence of HTV infection, and especially, in quantifying and characterizing asymptomatic people with prevalent HTV infection and people with AIDS. 1 Unfortunately, as the authors point out, a substantial proportion of HTV-infected people may be unknown to HTV surveillance.

As part of a US Centers for Disease Control and Prevention evaluation of AIDS surveillance, the Louisiana Department of Health measured the completeness of reporting of people with an HTV diagnosis to the Health Department. A list of all health care facilities licensed by the state was created. Facilities were categorized as: hospitals (n= 213), outpatient clinics (n= 48), and private physician's offices (n= 4643). A 15% random sample was drawn from hospitals (n= 32), 25% of clinics (n= 12), 15% of private physicians specializing in infectious diseases (n= 9), and 2% of physicians in all other specialities (n= 70). At each facility in the sample, diagnostic codes for HTV infection have been used to screen patients at the facilities since 1993, to determine which patients were most likely to meet HTV reporting criteria. 2 Then, the medical records of each of the potentially HTV-infected patients were reviewed to confirm whether they were eligible to have been reported. Of the patients, 493 were eligible to have been reported with HTV infection from hospitals, 2276 from clinics, and 159 from private physicians' offices. This information was matched with the HTV/AIDS surveillance registry.