Setting: San Francisco, California.
Objectives: To identify the characteristics of persons in whom tuberculosis was diagnosed after death, and determine whether secondary cases of tuberculosis resulted from them.
Design: Retrospective review of all cases of tuberculosis reported in San Francisco from 1986 through 1995, combined with a prospective evaluation of the molecular epidemiology of tuberculosis.
Results: Four per cent of the reported 3102 tuberculosis cases were diagnosed after death. The rate of tuberculosis cases diagnosed after death was 1.63 per 100000 population. Age 43 years or older, male sex, white race, and birth in the United States were characteristics independently associated with a diagnosis of tuberculosis after death. During 1993-1995, injecting drug use was also independently associated with a diagnosis of tuberculosis after death (odds ratio 9.24, 95% confidence interval 1.77-39.38). Cases of tuberculosis diagnosed after death do not appear to be significant sources of undetected tuberculosis transmission causing new secondary tuberculosis cases in the community.
Conclusions: Health care providers in San Francisco, and probably other urban areas, should maintain a high index of suspicion for tuberculosis in ageing, white, US-born males, and injecting drug users.