Testing for human immunodeficiency virus (HIV) most commonly is performed based on risk factors or in certain clinical situations where early detection clearly would be beneficial. Diagnosing patients in the acute or primary phase of HIV infection maximizes the effectiveness of secondary prevention and counseling aimed at reducing disease transmission. Ninety percent of patients with primary HIV infections are undiagnosed because they present with nonspecific symptoms that are similar to other viral illnesses such as influenza. Coco performed an analysis to estimate the cost per quality-adjusted life year (QALY) gained from primary HIV testing in all patients in the United States presenting with viral symptoms.
The analysis included three tests for primary HIV infection: (1) p24 antigen enzyme immunoassay (EIA), (2) HIV-1 RNA, and (3) third-generation HIV-1 EIA. The study population was a hypothetical cohort of approximately 3 million patients presenting in an ambulatory care setting with fever and other symptoms consistent with primary HIV. Patients were tested and scheduled for follow-up visits 14 days later; the authors projected that 69 percent would have kept this appointment. Patients with true-positive test results received HIV care consistent with current Department of Health and Human Services guidelines, including initiation of antiretroviral therapy at a CD4+ cell count of 350 per mm3 (350 × 106 per L). The time frame of the study was 39.9 years, which was the average life expectancy of the study population.
Data sources included a national estimate of the prevalence of primary HIV in symptomatic ambulatory patients; the measured prevalence of primary HIV at an urban urgent care center; reported sensitivities and specificities of the three tests; data from the Centers for Disease Control and Prevention on the high-risk sexual behavior of patients with HIV in the 12 months before diagnosis; and costs of HIV testing, counseling, and referrals from a comprehensive state-funded program in Massachusetts. Outcome measures included increased QALYs from starting antiretroviral therapy before significant immune impairment and HIV prevention in participants' sex partners. Benefits from immunizations and from cervical cancer and tuberculosis screening were not included in the analysis.
With the estimated national prevalence of primary HIV infection at 0.66 percent, the p24 antigen EIA was the most cost-effective test at $30,800 per QALY gained compared with no testing. In an urban urgent care setting with a reported HIV prevalence of 1 percent, the cost fell to $23,000 per QALY gained. Even in a hypothetical population with a primary HIV prevalence of less than 0.35 percent, the p24 antigen EIA met the commonly used societal cost-effectiveness threshold of $50,000 per QALY gained.
The author concludes that testing patients who present with nonspecific viral symptoms for primary HIV is likely to be cost-effective in a variety of clinical settings and could have a significant impact on reducing the number of new HIV infections in the United States.
editor's note: Telling a patient with cold symptoms that “it is just a virus” is the “bread and butter” of acute primary care visits. As an editorialist notes in the same issue ofAnnals of Family Medicine, this provocative study may be more useful as a basis for national policy-making than as a tool for individual physicians.1 Nonetheless, it should encourage physicians to consider risk factors for HIV infection when evaluating patients with nonspecific viral symptoms. —k.w.l.