BACKGROUND AND AIMS
Community-acquired pneumonia (CAP) is a common cause of antibiotic prescription, hospital admission, and mortality.1 Treatment according to guidelines has shown to be beneficial in CAP.2 However, adherence to the recommendations given in guidelines is highly variable.3 The authors have previously conducted a quality improvement project to increase guideline adherence;4 the project itself was a success, but the impact on key outcomes had not yet been assessed.
The aim of this study was to estimate the impact of a stewardship programme on patient management and outcomes in patients admitted with CAP.
MATERIALS AND METHODS
The authors conducted a before-after study comparing the odds for key outcomes in CAP before and after the implementation of an 8-month stewardship programme at three Danish hospitals. Comparisons were made using logistic regression models. The outcomes considered were antibiotics for ≤7 days, intravenous antibiotics for ≤3 days, antibiotics according to the guideline, length of stay ≤5 days, in-hospital mortality, and 30-day mortality. As stability within 72 hours of admission is a strong confounder, all analyses were performed on the overall cohort as well as a subset of patients stable within 72 hours. Univariable and multivariable analyses were performed, as well as a sensitivity analysis on a propensity score-matched cohort. The variables used for adjustment and matching were age, sex, comorbidities, pneumonia severity, antibiotics before admission, multilobular infiltrates, low oxygen saturation at admission, and a positive smoking history.
In total, 771 patients were eligible for participation in the present study. Of these, 423 were admitted in the baseline period and 348 in the follow-up period. The study cohort was a classic CAP cohort; the mean age was 72 years, and the sex ratio was balanced. Severe CAP with a CURB-65 score of 3–5 was observed in 17% of cases. Median durations of intravenous antibiotic treatment, total antibiotic treatment, and length of hospital stay were 2.6, 10.0, and 4.0 days at baseline, respectively.
The adjusted odds ratios (OR; 95% confidence interval [CI]) for antibiotics for ≤7 days were 1.85 (1.34–2.55) for the overall cohort and 2.14 (1.43–3.22) for the stable subgroup. The adjusted OR (95% CI) for correct empiric antibiotics were 1.94 (1.42–2.65) in the overall cohort and 1.78 (1.19–2.66) in stable patients. For all other outcomes, the effect sizes indicated no significant changes. Results for the propensity score-matched cohort were comparable to the results from the multivariable analyses.
The stewardship programme led to an increase in patients treated with the correct antibiotics according to the guideline, as well as a higher number of patients treated for 7 days or less. Significant changes were not observed for intravenous antibiotic treatment for 3 days or less, nor length of hospital stay of 5 days or less. However, the median durations of intravenous antibiotics and hospital stay were already satisfactory at baseline and noticeably better than reported in previously conducted studies. Hence, the improvement potential was not great. Significant changes regarding in-hospital or 30-day mortality were not detected. This was unsurprising, as the study was not powered to detect significant changes in mortality.