Abstract
Twenty-eight metropolitan health departments in the USA reporting more than 200 cases of tuberculosis annually were surveyed to determine the standard of practice in the diagnosis and treatment of pulmonary tuberculosis. Twenty-seven departments responded adequately for inclusion in this report. All programs used Mantoux skin tests, and only 3 programs ever used any other skin test. Of the programs, 26% followed the American Thoracic Society recommendation to collect 4-6 sputum samples in evaluating the bacteriologic status of patients suspected of pulmonary tuberculosis. Forty-five percent of the programs continued to follow some or all patients after completion of the designated course of antituberculosis chemotherapy. Two drug regimens (isoniazid with enthambutol or with rifampin) were used most frequently in the initial treatment. No program used parenteral agents in more than 11% of patients receiving initial treatment for pulmonary tuberculosis. Short-course chemotherapy regimens (6-9 mo.) and intermittent regimens (twice weekly) were used by only 41% of the programs; neither type of regimen was used for more than 10% of patients in any program. Fourteen percent of the programs performed routine laboratory biochemical tests to monitor for streptomycin-induced nephrotoxicity, 15% for isoniazid-induced hepatotoxicity and 31% for rifampin-induced hepatotoxicity. Eighty-one percent of the programs performed routine clinical testing for ophthalmologic complications from ethambutal; only 30% ever used dosages > 15 mg/kg. No program routinely hospitalized all new cases of tuberculosis for initial treatment, but the proportion of patients considered ill enough to require hospitalization varied from 1-80%. In contrast to previous findings for tuberculosis chemoprophylaxis, considerable variation existed in the treatment and monitoring of patients with pulmonary tuberculosis in major metropolitan programs in the USA.