News of the Day ... In Perspective09/26/2006
CDC breaks precedent in recommending universal HIV screening
The U.S. Centers for Disease Control and Prevention (CDC) has announced a new recommendation to test all Americans for human immunodeficiency virus (HIV), irrespective of risk status.
Testing would be done when patients present for urgent or emergency care, or even during a routine physical. Consent would be included in a clinic or hospital’s standard consent form, although patients would be allowed to opt out.
The AMA endorses the CDC recommendation. The ACLU worries that routine testing will in practice become mandatory testing.
The CDC estimates that 250,000 Americans carry the HIV virus without knowing it, a prevalence of less than 1 in 1,000 overall. Early diagnosis would permit “prompt care” with $10,000 worth of drugs per year.
In the past, the CDC has used a 1% (1 in 100) prevalence as general guidance for routine as opposed to targeted screening.
Applying a little evidence-based medicine to the recommendation, one can easily calculate that in a low-risk population, a high percentage of positive screening tests will be false positives, subjecting patients to intense anxiety, severe marital problems, and potentially toxic therapy.
If the test has a specificity of 99.99%, only about half of positive tests would be true positives in a population with a 1 in 10,000 risk of disease. It is likely that the accuracy of the screening test in practice will not be as high as has been asserted.
In a population of 1,000,000 individuals with a disease prevalence of 1 in 10,000, there would be 100 patients with the disease, and 999,900 without. Of those with the disease, essentially all 100 would have a positive test. Of those without the disease, 0.01% or about 100 would have a falsely positive test. Thus, of the 200 positive tests, only half would be true positives.
For persons with a lower risk than 1 in 10,000, or with a less specific tests, the percentage of false positives would be even higher.
Joel Kauffman, Ph.D., suggests that many suicides or divorces would be prevented if the informed consent for testing included this fact. .
A study of the benefits of screening asymptomatic individuals in a population assumed to have an HIV risk of 1%, much higher than the population as a whole, calculated an increased [population] life expectancy of 3.92 days (or 2.92 quality-adjusted days). If benefits to sexual partners are included, an increased life expectancy of 5.48 days (4.70 quality-adjusted days) was calculated. In this study, the specificity of the entire series of screening tests was taken to be 99.9994%, with a range from 99 to 100%. The number of false positives was calculated to be negligible after all positives were subjected to further tests.