Objective: To determine the relative frequencies of primary and reactivation tuberculosis in the urban homeless.
Design: Prospective evaluation of homeless tuberculosis patients.
Setting: Central Los Angeles, Calif.
Patients: Thirty-four homeless patients with culture-proven tuberculosis.
Interventions: IS6110-based restriction fragment length polymorphism (RFLP) analysis was performed on Mycobacterium tuberculosis isolates. If results were inconclusive, pTBN12-based RFLP analysis was performed.
Main outcome measure: Clustering of M tuberculosis isolates. A cluster consisted of two or more isolates with indistinguishable RFLP patterns.
Results: Twenty-four of 34 homeless patients had clustered isolates in six clusters.
Conclusions: The minimum percentage of cases due to primary tuberculosis in the homeless was estimated to be 53%, compared with the traditional estimate of 10% in the general population. The results suggest that primary tuberculosis caused the majority of tuberculosis cases in this population of the urban homeless in central Los Angeles.