Clean bill of health to both PPD antigens
False positives not a worry, study concludes
For years, reports have popped up in the medical literature warning about clusters of false positives thought to have resulted from one or both of the two approved purified protein derivative skin-test products. Usually, the reports fingered Aplisol (from Parkdale, formerly Parke- Davis, of Morris Plains, NJ) as too sensitive and apt to give false-positive results. Tubersol (made by Connaught Labora tories in Swiftwater, PA), sometimes was said to be too specific, resulting in false-negative readings.
Four years and one long, two-part study later, the Centers for Disease Control and Prevention in Atlanta has thoroughly dispelled any uncertainties about the trustworthiness of the two products.
Conclusions from the second part of the study found Tubersol and Aplisol have equally high specificity, says Elsa Villarino, MD, MPH, chief of the therapeutic and diagnostic studies section of the CDC's Division of Tuberculosis Elimination. That good news follows findings from the first part of the study, which concluded both products scored equally high in the arena of sensitivity.
To select the population for the specificity portion, researchers used a screening questionnaire designed to yield subjects unlikely to have been exposed to TB. The subjects were injected intracutaneously with Tubersol and Aplisol; and to provide a basis for comparison, they also were injected with PPD-S1, the antigen the Food and Drug Administration maintains as the standard and provides to PPD manufacturers. Some subjects received a fourth antigen, PPD-S2, which the FDA intends eventually to substitute for PPD-S1.
About 1,500 volunteers took part, from six sites around the country, including Denver; Marion County, IN; Seattle-King County, WA; Atlanta; Tucson, AZ; and San Diego. Results were distributed equally across all categories for all sites, except for San Diego, Villarino says. Readers there found about the same number of positive results as did those at other sites, but they recorded a greater proportion of results between 1 and 9 mm.
After puzzling over the discrepancy, researchers decided what probably had happened, Villarino says. The San Diego readers work in a university-run clinic that performs about 10,000 skin tests every year; because many of the tests assay for anergy, the readers were skilled at palpating even the smallest suggestion of an induration.
That underscores an important consideration for all PPD skin-testing, Villarino adds. "Since it takes a highly trained person to read this test, the likelihood of error varies," she says. "Other tests have much better rates of accuracy because they use more standardized, less subjective measures."
Plus, even when the factor of reader error is ruled out, the rate of false positivity for PPD skin tests stands at about 4% to 5%, Villarino says. The rate of false negativity is higher still, at about 10% (because people whose immune systems have been compromised by HIV, other forms of illness - or even TB itself - are less likely to mount a strong response to the PPD antigen).
Thus, until a better test for TB infection is found, perhaps the best way to avoid false positives is not to test people unnecessarily, she says. Unnecessary testing only increases the chance a positive reading won't be a true one, she adds. Meanwhile, her study shows there is one thing clinicians don't have to worry about - which skin test is better.