© 2002 American Thoracic Society
Improving tuberculin skin testing in hiv-infected individualsTo the Editor :We read with interest the article by Lalvani and colleagues (1) and the editorial by Barnes (2) concerning rapid detection of Mycobacterium tuberculosis infection by enumeration of antigen-specific T cells. This method of testing for latent tuberculosis infection (LTBI) is more sensitive, more specific, and more convenient than the tuberculin skin test (2). This scientific advance in diagnosing LTBI is especially useful for HIV-infected individuals, who are more at risk of acquiring M. tuberculosis infection (3). Certain subgroups of HIV-infected individuals are also more at risk of not returning a second time for interpretation of the tuberculin skin test (4).
In our retrospective study, we examined a population-based cohort of individuals in New Orleans ( The sample included 370 people. Most patients were male (71.5%) and African American (71.7%). The mean age was 35.8 years (SD, 10.52 years). More than one-third of the patients (36.2%) did not have any tuberculin skin tests recorded in their charts. After adjustment for CD4 cell count, multivariate analysis showed that male, African American, and subjects with no history of preventive or antiretroviral care were significantly associated with no tuberculin skin tests recorded in the medical charts. Receiving care at a public HIV outpatient clinic, as opposed to private care, inpatient care, or hospital outpatient care, was protective. The odds ratio for no tuberculin skin testing among those receiving care at a public HIV clinic compared with others was 0.22 (95% confidence interval [CI], 0.07–0.72]. HIV-infected patients receiving care in New Orleans' private facilities and hospitals need improved access to tuberculin skin testing. Providers also need to provide more tuberculin skin testing and care to males and African Americans. A more convenient test, such as the one described by Lalvani and colleagues may improve testing for LTBI in New Orleans. In our study, approximately a third of the patients who did receive a tuberculin skin test (31.9%) did not return for interpretation of the test. We recommend better tests to diagnose LTBI in HIV-infected individuals, such as the one described by Lalvani and colleagues.
HIV/AIDS Program Office, Louisiana Office of Public Health New Orleans, Louisiana REFERENCES
From the authors: Welch and More highlight the operational difficulties of the century-old tuberculin skin test (TST). Many people who should be tested are not, and many who are tested fail to return to have the test read. Welch and More have identified important risk groups in the United States who are not being offered TST, despite the fact that the U.S. has a more vigorous and comprehensive approach to identifying and treating latent tuberculosis infection (LTBI) than almost any other country (1). As Welch and More point out, the world awaits an improved diagnostic test for LTBI. Such a test should not only be more convenient, but also more accurate than the TST. The rapid enzyme-linked immunospot (ELISPOT) test that we have developed is more convenient, as it is a blood test that generates results by the next day and requires no return visit (2, 3). We have recently further simplified the format and procedure of the ELISPOT assay along the lines suggested by Barnes (4). The new streamlined assay could soon be suitable for deployment in routine hospital laboratories. Proving that a new test for LTBI is more accurate than the TST is difficult, as there is no gold standard against which to measure the accuracy of a new test. Given that time spent sharing room air with an infectious source case is a key determinant of M. tuberculosis transmission, an improved test for LTBI should correlate better with intensity of exposure to M. tuberculosis than the TST. We recently reported a study of 50 tuberculosis contacts in London that showed that our ELISPOT assay did indeed correlate more closely with the level of M. tuberculosis exposure than the TST, and unlike the TST, was independent of BCG vaccination status (5). However, before clinical decisions could be based on ELISPOT assay results, its relationship to M. tuberculosis exposure would need to be precisely defined in a larger study with sufficient power to generate statistically significant results with narrow confidence intervals. We have recently undertaken such a study in a major tuberculosis outbreak due to a single source case in a school in the United Kingdom, together with the local health authority. This large (n = 545), prospective, blinded trial compares the relationship of the ELISPOT assay and TST with M. tuberculosis exposure, as quantified precisely from school timetables. This should provide a definitive answer as to whether the ELISPOT assay could replace the TST for diagnosing LTBI in the 21st century.
University of Oxford, John Radcliffe Hospital Oxford, United Kingdom REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
13 years) diagnosed with HIV infection from June 1995 through June 1997. The purpose of the Centers for Disease Control and Prevention study was to assess whether the U.S. Public Health Service guidelines to prevent M. tuberculosis in HIV-infected persons were completed. Data were extracted from medical records for care provided from first HIV positive test through December 1998 or death.