Background Significant improvements in quality of inpatient care would be expected to improve objective outcomes, such as length of stay (LOS) and mortality. The management of community acquired pneumonia was reviewed on both sites of this Trust over a 33 month period during which time training and feedback was provided on management, monitoring 5 aspects of clinical care:- documentation of oxygenation; documentation of CURB65 scores; blood cultures (if obtained) taken before antibiotics; antibiotics given in accordance with Trust guidelines; antibiotics given within 6 hrs.

Methods 4922 patients discharged between July 2010 and March 2013 had been coded as being admitted with pneumonia. The notes of 52 patients were unobtainable. Patients were excluded where there was no evidence of pneumonia on admission CXR, where the admitting clinicians had not diagnosed pneumonia or when the pneumonia was not considered to be community acquired. A monthly ‘composite quality score’ (CQS) score was derived by averaging the percentage of the 5 parameters that were achieved for each patient/admission.

Results 2842 (58%) patients remained for analysis. Significant improvements were noted in all the parameters being monitored with quarterly CQS scores rising from 68% to 93% and, more specifically, quarterly scores for antibiotic delivery within 6 hrs increasing from 40% to 78%. Despite this no improvement at all was seen in LOS (quarterly LOS range: 10.8–13.4 days; no time trend). A gradual fall in mortality occurred, but only at one of the two sites (absolute fall of 2.8% per yr cf. no change). Of note was an unexpected and progressive increase in quarterly admissions (35% per year) at the site with falling mortality, but no change at the other.

Conclusions Demonstrable improvement in parameters of clinical care result from education and feedback. However, a subsequent lack of change in outcomes (LOS, and, at one site, in mortality) despite such clear improvements would be unexpected. There may either be more appropriate clinical parameters on which to focus or these outcomes may have multifactorial influences amongst which improvements in current levels of clinical care may actually have a minor role.