Hospital care of tuberculosis patients referred for directly observed therapy in New York City: identifying factors in prolonged stays.

Individuals hospitalized and treated for tuberculosis (TB) who were then enrolled into TB directly observed therapy at four study hospitals in New York City (NYC) were identified. Review of hospital medical records determined whether the hospitalizations were warranted and whether lengths of stay were prolonged. Most hospitalizations were appropriate but over 70 percent of cases analyzed had prolonged stays. Of these, almost half were to document bacteriologic response to anti-TB treatment. Some were prolonged due to misunderstanding of state recommendations. Focused educational efforts could significantly reduce lengths of stay and save up to $9.7 million annually in NYC hospitalization costs.