Introduction Community-acquired pneumonia (CAP) is a common cause for hospital admission and carries a high mortality rate. Choosing the correct antibiotic can be challenging and “atypical” pathogens, such as Mycoplasma pneumoniae and Chlamydophila pneumoniae, are not eliminated by some first-line empirical agents. Identification of the infecting organism through microbiological sampling can help to tailor antibiotic therapy and substantially improve patient outcomes. Guidelines exist (NICE, BTS, trust) that recommend which patients should undergo microbiological sampling. We wished to determine whether these guidelines were followed in our trust in patients admitted with CAP.

Methods We reviewed the notes of adult patients admitted over a 15-week period (February–May 2014) with a clinical code of pneumonia. Further information was derived from hospital systems, including Telepath (used by the Microbiology department) and from ‘Advancing Quality’ data available for these patients.

Results 175 patients were identified with CAP. Blood cultures (BCs) were indicated in 89 patients according to trust guidelines (based on Systemic Inflammatory Response Syndrome (SIRS) criteria, CURB-65 score, immunocompromised; data sufficient in 147) and appropriately collected in 55 (61.8%). However, only 4 had positive (clinically significant) BC results. Sputum samples were sent for 31 patients (17.7%, n = 175) and only 19% had significant bacterial growth. All patients transferred to intensive care (ITU; 3.4%) were screened appropriately for urinary pneumococcal antigen (UPA) and urinary legionella antigen (ULA), with 1 positive UPA result. 6 (of 10) UPA and 7 (of 12) ULA samples appeared to be sent inappropriately for non-ITU patients and were rejected by the laboratory. Serum samples were sent for Mycoplasma testing in 7 patients, despite this service being superseded by testing by polymerase chain reaction (PCR)-testing of nose/throat swabs, for which only 1 was sent for the entire cohort.

Conclusion While BC collection was frequent, compliance with local guidance was less than desired. Sampling to rule out atypical pathogens in CAP appeared to be low and, occasionally, inappropriate. Although positive detection rates were low, further work appears to be indicated to optimise patient outcomes in CAP by increasing awareness amongst clinicians of tests available and ensure appropriate sampling for atypical pathogen screening.