Background:The clinical value of sputum culture in suspected lower respiratory tract infection (LRTI) remains contentious. The quality of samples submitted significantly impacts their clinical usefulness. Methods: Using pre-defined criteria we prospectively analysed the appropriateness of sputum samples submitted from consecutive patients with suspected LRTI attending two acute hospital units over ten weeks. We then provided an education package for staff on when and how to collect appropriate sputum samples, and repeated the evaluation. Results: Our intervention reduced sample numbers from 347 to 133, simultaneously increasing the proportion of appropriately sent samples from 40.5 to 60.2% (p=0.001) and reducing cost. Appropriate sampling was associated with a higher yield of pathogens (relative risk 1.51, 95% confidence intervals 1.03-2.21, p=0.03). The rate at which sputum samples appeared to alter clinicians' management remained low and constant at 18% pre-and post-intervention. Conclusion: A simple educational intervention can significantly increase appropriateness of sputum sampling, reducing workload and cost. The role of sputum sampling in the management of lower respiratory tract infection (LRTI) has been controversial for some time. Evidence suggests that sputum sampling has low sensitivity, specificity and impact on treatment decisions. [1][2][3][4] Ideally antibiotic treatment should be tailored to the causative pathogen but at presentation the aetiology is rarely known, and patients therefore commonly receive empirical, often broad-spectrum antibiotics. KeywoRdSIn theory, a pathogen-directed approach to treatment has many advantages such as narrowing the therapeutic index of antibiotics, thus reducing adverse events and the risk of antimicrobial resistance. In addition, it allows the identification of unusual or unexpected pathogens, antibiotic-resistant organisms or those of epidemiological or public health significance. In practice the diagnostic usefulness of sputum cultures is limited by use of prior antibiotics, 5 or the patient's inability to produce sputum. Generally, when sputum is obtained, the proportion of 'high quality' samples ranges from 25-60%, though higher rates have been reported. [6][7][8][9][10][11] The variable yield in sputum cultures is also influenced by the transport, processing and interpretation of samples.With the resultant low diagnostic yield and minimal impact on therapeutic decisions, routine sputum cultures have been discouraged. The Infectious Diseases Society of America/American Thoracic Society guidelines recommend that sampling in hospitalised adult patients be restricted to those able to provide a high quality pretreatment specimen and where quality performance measures for sputum collection, transport and processing of samples can be met, particularly in patients with severe community-acquired pneumonia (CAP).12 The British Thoracic Society similarly recommend that in hospitalised adults, samples should only be sent from patients able to expectorate pu...

No abstract