The association between adherence to national antibiotic guidelines and mortality, readmission and length of stay in hospital inpatients: results from a Norwegian multicentre, observational cohort study
Peer reviewed, Journal article
MetadataVis full innførsel
Background: Clinical antibiotic prescribing guidelines are essential in defining responsible use in the local context. Our objective was to investigate the association between adherence to national antibiotic prescribing guidelines and patient outcomes across a wide range of infectious diseases in hospital inpatients. Methods: Over five months in 2014, inpatients receiving antibiotics under the care of pulmonary medicine, infectious diseases and gastroenterology specialties across three university hospitals in Western Norway were included in this observational cohort study. Patient and antibiotic prescribing data gathered from electronic medical records included indication for antibiotics, microbiology test results, discharge diagnoses, length of stay (LOS), comorbidity, estimated glomerular filtration rate (eGFR) on admission and patient outcomes (primary: 30-day mortality; secondary: in-hospital mortality, 30-day readmission and LOS). Antibiotic prescriptions were classified as adherent or non-adherent to national guidelines according to documented indication for treatment. Patient outcomes were analysed according to status for adherence to guidelines using multivariate logistic, linear and competing risk regression analysis with adjustments made for comorbidity, age, sex, indication for treatment, seasonality and whether the patient was admitted from an institution or not. Results: In total, 1756 patients were included in the study. 30-day-mortality and in-hospital mortality were lower (OR = 0.48, p = 0.003 and OR = 0.46, p = 0.001) in the guideline adherent group, compared to the non-adherent group. Adherence to guideline did not affect 30-day readmission. In linear regression analysis there was a trend towards shorter LOS when LOS was analysed for patients discharged alive (predicted mean difference − 0.47, 95% CI (− 1.02, 0.07), p = 0.081). In competing risk analysis of LOS, the adherent group had a subdistribution hazard ratio (SHR) of 1.17 95% CI (1.02, 1.34), p = 0.025 for discharge compared to the non-adherent group. Conclusions: Adhering to antibiotic guidelines when treating infections in hospital inpatients was associated with favourable patient outcomes in terms of mortality and LOS.