Determinants of non-adherence to antibiotic treatment guidelines in hospitalized adults with suspected community-acquired pneumonia: a prospective study

Abstract

Background: Antimicrobial resistance (AMR) is a global health threat with millions of deaths annually attributable to bacterial resistance. Effective antimicrobial stewardship programs are crucial for optimizing antibiotic use. This study aims to identify factors contributing to deviations from antibiotic treatment guidelines in hospitalized adults with suspected community-acquired pneumonia (CAP).

Methods: We conducted a prospective study at Haukeland University Hospital's Emergency Department in Bergen, Norway, from September 2020 to April 2023. Patients were selected from two cohorts, with data on clinical and microbiologic test results collected. We analysed adherence of antibiotic therapy to guidelines for the choice of empirical treatment and therapy duration using multivariate regression models to identify predictors of non-adherence.

Results: Of the 523 patients studied, 479 (91.6%) received empirical antibiotic therapy within 48 h of admission, with 382 (79.7%) adhering to guidelines. However, among the 341 patients included in the analysis of treatment duration adherence, only 69 (20.2%) received therapy durations that were consistent with guideline recommendations. Key predictors of longer-than-recommended therapy duration included a C-reactive protein (CRP) level exceeding 100 mg/L (RR 1.37, 95% CI 1.18-1.59) and a hospital stay longer than two days (RR 1.22, 95% CI 1.04-1.43). The primary factor contributing to extended antibiotic therapy duration was planned post-discharge treatment. No significant temporal trends in adherence to treatment duration guidelines were observed following the publication of the updated guidelines.

Conclusion: While adherence to guidelines for the choice of empirical antibiotic therapy was relatively high, adherence to guidelines for therapy duration was significantly lower, largely due to extended post-discharge antibiotic treatment. Our findings suggest that publishing updated guidelines alone is insufficient to change clinical practice. Targeted stewardship interventions, particularly those addressing discharge practices, are essential. Future research should compare adherence rates across institutions to identify factors contributing to higher adherence and develop standardized benchmarks for optimal antibiotic stewardship. Trial registration NCT04660084.

Keywords: Antibiotic stewardship; Antibiotic therapy duration; Antimicrobial resistance (AMR); C-reactive protein (CRP); COPD; Community-acquired pneumonia (CAP); Empirical antibiotic therapy; Guideline adherence; Hospital discharge practices.

Fig. 1

Flowchart of included patients. LRT Lower Respiratory Tract

Fig. 2

Guideline Adherence in Empirical Antibiotic Treatment Across Patient Groups by CRB-65 Score and the presence of Penicillin Allergy. Percentages of guideline adherence to empirical antibiotic treatment across patient groups stratified by CRB-65 score and the presence of penicillin allergy. Groups are defined as follows: (1) patients with penicillin allergy and CRB-65 ≥ 3, (2) patients with penicillin allergy and CRB-65 ≤ 2, (3) patients without penicillin allergy and CRB-65 ≥ 3, and (4) patients without penicillin allergy and CRB-65 ≤ 2. Each bar represents the proportion of patients within each group who received guideline-adherent (blue) or non-adherent treatment (orange). The total number of patients within each group (N) is noted within each bar

Fig. 3

Antibiotic Treatment Durations. Panel A Total duration of antibiotic therapy among patients who received between 24 h and 20 days of treatment, comparing those included (n = 341) and not included (n = 107) in the analysis. Panel B Total antibiotic treatment duration in patients with CRB-65 scores of 0–2 compared to those with scores > 2, with lines indicating the cutoffs used to define guideline adherence. Panel C Relative contribution of in-hospital versus post-discharge treatment durations

Fig. 4

Impact of Length of Stay, CRP Levels, and Time Since Guideline Change on Adherence to Antibiotic Treatment Duration. A–C The impact of three continuous variables—maximum C-reactive Protein (CRP) level, length of stay, and time since the guideline change—on the probability of receiving antibiotic treatment durations longer than recommended. For A and B the dotted line represents the cutoff chosen when converting the variables to binary for the Poisson regression model. For C the dotted line represents the start of patient inclusion