Hospital-resource utilization and tuberculosis.

As I sat down to write this editorial, I was paged by a colleague who had just learned that she may have pulmonary tuberculosis. She had a very strong exposure history in a highly endemic area of the world within the past year, a positive tuberculin skin test, and a “soft” but compatible history of recent symptoms. Her only available diagnostic test, a chest radiograph, obtained because of a nonproductive cough of 1 week duration, was suggestive of pulmonary tuberculosis. Based on this information, I arranged for her to be furloughed from work, ordered sputum specimens, and sent her home to complete the evaluation as an outpatient. One of her first two sputum smears was positive for acid-fast organisms, and antituberculosis therapy was initiated at home. Six months ago, while I managed a hospitalized diabetic, alcoholic patient with multidrug-resistant tuberculosis and a track record of poor compliance, the medical director of the patient’s managed-care company called to inform me that payment for the hospitalization would be denied effective 2 days after an episode of massive hemoptysis. He expressed his concern for the dilemma I faced, but informed me that the hospitalization would be denied nonetheless. I replied that the dilemma was his, not mine, and that I would not discharge the patient from the hospital. Each of us stood our ground, and the patient stayed in the hospital. The institution eventually recouped a level of reimbursement below the acute-care rate. The patient was discharged a week later, but died within a month, although not of hemoptysis. Opinions on the role of hospitalization for tuberculosis are divided. Physicians with whom I’ve shared these cases, an admittedly biased sample, have concurred with my decisions. I believe most physicians who treat tuberculosis regularly would have managed these patients in a similar fashion. Milliman and Robertson, Inc, a national consulting firm that provides guidelines to managed-care organizations on length of stay, indicates in the December 1997 edition of its Healthcare Management Guidelines that, for all diagnostic categories of tuberculosis, a 2to 4-day hospitalization is the optimal length of stay, with the exception of tuberculous meningitis, which is accorded 7 days.1 These guidelines are designed to “define recovery and care for patients who do as well as one hopes and have no complications.”2 The authors acknowledge that their recovery guidelines are “extremely optimistic and in some instances, unrealistic,” particularly for the most severe medical conditions. The qualifications in these guidelines notwithstanding, the denial I described previously is not an unusual occurrence. Our institution is informed regularly by managed-care organizations that admissions for the evaluation and initiation of therapy for tuberculosis will be denied on the basis that such patients can be evaluated and have therapy initiated in the outpatient setting. At the heart of the issue is the cost of hospitalization and resource utilization. To this debate, Griffiths and colleagues have injected science from their well-executed study published in the current issue of Infection Control and Hospital Epidemiology.3 They have evaluated the utilization of medical resources related to tuberculosis in a municipal hospital in a community with one of the highest endemic rates of tuberculosis in the United States. Cases were selected on