Setting: Maryland Department of Health and Mental Hygiene, Division of Tuberculosis (TB) Control.
Objectives: To assess the implications of antibiotic treatment of presumed community-acquired pneumonia (CAP) on delays in the diagnosis of TB, and to assess the frequency with which chest radiographs (CXRs) were utilized before a diagnosis of pneumonia or pulmonary TB was made.
Design: A nested case-control study within a prospective study conducted to assess factors associated with delays in the diagnosis of TB.
Results: Cases (n = 85; 54%) were patients who received antibiotics for non-TB diagnoses/indications prior to TB diagnosis, and controls (n = 73; 46%) were patients who had initially received TB therapy. Median health care delay for cases was 39 days vs. 15 days (P < 0.01) for controls. Median antibiotic delay was similar among all antibiotic classes. Of 54 patients who did not have a CXR at their first health care visit, 41 (79%) received empiric antibiotics, compared to 44/105 (42%) who had a CXR (P < 0.01). Only 31/54 (57%) patients initially diagnosed with CAP had a CXR at the time of diagnosis.
Conclusion: More widespread use of CXR when diagnosing CAP should reduce delays in diagnosing TB, and the unnecessary use of antibiotics.