Pneumococcal Colonization Rates in Patients Admitted to a United Kingdom Hospital with Lower Respiratory Tract Infection: a Prospective Case-Control Study

Current diagnostic tests are ineffective for identifying the etiological pathogen in hospitalized adults with lower respiratory tract infections (LRTIs). The association of pneumococcal colonization with disease has been suggested as a means to increase the diagnostic precision. We compared the pneumococcal colonization rates and the densities of nasal pneumococcal colonization by (i) classical culture and (ii) quantitative real-time PCR (qPCR) targeting lytA in patients with LRTIs admitted to a hospital in the United Kingdom and control patients. A total of 826 patients were screened for inclusion in this prospective case-control study. Of these, 38 patients were recruited, 19 with confirmed LRTIs and 19 controls with other diagnoses. Nasal wash (NW) samples were collected at the time of recruitment. Pneumococcal colonization was detected in 1 patient with LRTI and 3 controls (P = 0.6) by classical culture. By qPCR, pneumococcal colonization was detected in 10 LRTI patients and 8 controls (P = 0.5). Antibiotic usage prior to sampling was significantly higher in the LRTI group than in the control group (19 versus 3; P < 0.001). With a clinically relevant cutoff of >8,000 copies/ml on qPCR, pneumococcal colonization was found in 3 LRTI patients and 4 controls (P > 0.05). We conclude that neither the prevalence nor the density of nasal pneumococcal colonization (by culture and qPCR) can be used as a method of microbiological diagnosis in hospitalized adults with LRTI in the United Kingdom. A community-based study recruiting patients prior to antibiotic therapy may be a useful future step.