In updating its recommendations for HIV screening, the USPSTF strongly recommended that primary care clinicians screen all pregnant women as well as all adults and adolescents with identifiable risk factors for HIV infection (both were grade A recommendations) (1). It made, however, no recommendation for or against screening patients who are neither pregnant nor at high risk for HIV infection (a grade C recommendation), leaving the decision of whether to screen those individuals to the discretion of the primary care clinician. A grade C recommendation indicates that the USPSTF has found at least fair evidence that screening can improve health outcomes but has concluded that the balance of benefits and harms associated with screening is too close to justify a general recommendation (2).

In the clinical considerations that accompany the recommendations for HIV screening, the Task Force defined "high risk" not only in terms of the risk behaviors of the individual but also in terms of the risk and prevalence characteristics of the populations served in the clinical setting. The USPSTF explicitly encourages clinicians to consider the prevalence of HIV infection and the risk characteristics in the populations they serve when determining an appropriate screening strategy based on the recommendations. The consideration of risk in terms of clinical setting and prevalence warrants additional attention as the Task Force recommendations on HIV screening are implemented.

As Dr. Beckwith and his colleagues point out, routinely screening all asymptomatic primary care patients, including those with no identifiable risk factors for HIV infection, would probably increase the proportion of infected persons who are diagnosed in this country and could result in improved health outcomes for at least some of them. The Task Force, however, concluded that these potential benefits do not outweigh the burden on primary care practices or the potential harms of a general HIV screening program. The overall number of asymptomatic primary care patients not at high risk for HIV infection who would be identified as seropositive through general screening would be small. In fact, the USPSTF estimated that as many as 11 000 non–high-risk patients would have to be screened to prevent 1 clinical progression or death over 3 years (3). Given the competing demands, the limited duration of the average primary care visit, and the very small likelihood that any 1 clinician would provide health benefits to an otherwise-undetected seropositive patient, the physician should consider whether this time might be better spent providing other preventive services (such as Papanicolaou testing or smoking cessation counseling) that carry the potential to improve outcomes for more patients. The Task Force's grade C recommendation for non-high-risk patients allows the clinician to give lower priority to this service and to make the decision on an individualized basis, in collaboration with his or her patient.