Journal of Health Care for the Poor and Underserved

263 Brief communication TUBERCULIN POSITIVITY AMONG THE HOMELESS To the Editor: The homeless suffer disproportionately from a variety of medical problems including tuberculosis.1 Various studies have found that the prevalence of clinically active tuberculosis among the homeless is 150 to 300 times that in the general population.2 Certain cities have reported prevalences of tuberculin positivity in the homeless between 22 percent and 50 percent.2 Poorly ventilated, crowded homeless shelters contribute to this increased prevalence.1 In addition, many shelter guests suffer from substance abuse disorders (including alcoholism) and human immunodeficiency virus (HIV), both of which are risk factors for tuberculosis.2 Unlike many large cities in the United States, Baltimore has not witnessed a major outbreak of tuberculosis in any of its homeless shelters (J. Scavatto and R. Rizzo, personal communication). We studied a convenience sample of Baltimore's homeless to examine the prevalence of tuberculin test positivity in a city with no known tuberculosis outbreaks in homeless shelters. Methods. The sample consisted of 84 persons who had spent at least one night during the previous year in a homeless shelter. Screening was performed on four separate occasions at a night shelter, Helping Up Mission, and a day shelter,Christopher's Place. In addition, all persons who presented to Baltimore's Health Care for the Homeless Clinic during July 1991, and were identified as having spent a night in a shelter, were screened. Every person included in the study voluntarily sought medical care. Persons who had been tested within the past six months or who had histories of positive tuberculin skin tests were not retested. Before the skin tests were administered, research personnel completed a modified version of the Baltimore City Health Department's tuberculosis referral form, which assessed demographics, tuberculosis symptoms, and risk factors. Each subject received 0.1 ml of 5 TU purified protein derivative (PPD) intradermally in the volar surface of either forearm approximately three to four inches below the elbow. Each subject also received written and verbal information regarding the transmission of tuberculosis. Journal of Health Care for the Poor and Underserved, Vol. 3, No. 2, Fall 1992 264___________________________________________________________ Research personnel or trained nurses read the skin tests using the pen method (marking up the edge of any induration from four directions and measuring the distance between marks). Skin tests were defined as positive or negative according to the criteria established by the Centers for Disease Control and American Thoracic Society.3 Since no persons reported HIV positivity, close contact with newly infected persons, or had chest x-rays compatible with old healed tuberculosis, researchers used the 10 mm cut-off for all subjects. Whenever possible, results were read 72 hours after skin testing. Because some participants did not return for the initial reading, research personnel trained shelter staff to interpret skin tests for later reading. Staff readings and reports at 72 hours of self-readings were accepted up to one month following skin testing. Participants who could not be found after one month were excluded from the analysis (n=27). Small incentives (cookies) were provided, when possible, at the time of testing. Subjects were informed that they would receive a pair of socks upon returning for skin test readings. Persons with positive PPDs were referred to the Baltimore City Health Department, Health Care for the Homeless Clinic, or their usual source of medical care. Bus tokens were provided to encourage compliance . Epi Info statistical software was used for analysis of the data.4 Risk factors examined included age, race, length of stay in shelter, and presence of ΗΓν risk factors (including intravenous drug use). Continuous variables were categorized into groups, and univariate relationships between tuberculosis status and risk factors were assessed. Odds ratios and p-values were calculated for all variables. The Mantel-Haenszel χ2 test was used to evaluate significance unless the expected value for any cell in a table was less than five, in which case a twotailed p-value for the Fisher exact test was used. Results. No data are available on nine persons excluded from screening because of previous skin testing. Ninety-three people were screened; Table 1 presents descriptive data on the 84 people who were included in...