Antibiotic treatment and the diagnosis of Streptococcus pneumoniae in lower respiratory tract infections in adults.

OBJECTIVE

To analyze the possible influence of antibiotic treatment on the results of different diagnostic tests for the diagnosis of lower respiratory tract infections with Streptococcus pneumoniae.

MATERIAL AND METHODS

A prospective cohort of 159 unselected adult immunocompetent patients admitted to Silkeborg County Hospital in Denmark with community-acquired lower respiratory tract infections underwent microbiological investigations with fiber-optic bronchoscopy with bronchoalveolar lavage, blood and sputum culture and urine antigen test for type-specific polysaccharide capsular antigens of S. pneumoniae.

RESULTS

When stratified for antibiotic treatment prior to microbiological sampling, three different groups of patients with documented or probable infection with S. pneumoniae could be identified. The first group comprised 14 patients who were culture positive in one or more culture tests, where most (11/14) did not receive any antibiotic treatment within 24 hours of sampling. The second group consisted of nine patients with a positive urine antigen test where 8/9 and 9/9 received antibiotic treatment 24 and 48 hours, respectively, prior to urine sampling. Only a single patient was positive in both systems, making a total of 22 patients with documented pneumococcal infection. As a positive culture test was dependent on the absence of antibiotic treatment, whereas a positive urine antigen test depended on antibiotic treatment within 48 hours, the two tests were complementary in the diagnosis of infection with S. pneumoniae. The third group of patients with probable pneumococcal infection were identified as 26% and 20% of the remaining 137 patients with unknown or known non-pneumococcal etiology, respectively, who received recent antibiotic treatment within 2-4 weeks of diagnostic sampling. By comparison, 0% (p < 0.01) with documented pneumococcal infection received antibiotic treatment in weeks 2-4 prior to microbiological sampling. As such a further eight patients should be expected to have infection with S. pneumoniae but would test negative in both culture tests and the urine antigen test because of antibiotic treatment within weeks 2-4 prior to sampling.

CONCLUSION

The diagnosis of infection with S. pneumoniae is very dependent on whether or not recent (within 2-4 weeks) or immediate (within 48 hours) antibiotic treatment has been given prior to microbiological sampling of patients. The results suggest an optimized diagnostic strategy with, if possible, sampling for culture prior to antibiotic treatment, while sampling for pneumococcal antigens should wait 24-48 hours for antibiotic treatment.