Identification of outcomes reported for hospital antimicrobial stewardship interventions using a systematic review of reviews

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Background: Randomized trials of hospital antimicrobial stewardship (AMS) interventions aimed to optimize antimicrobial use contribute less to the evidence base due to heterogeneity in outcome selection and reporting. Developing a core outcome set (COS) for these interventions can be a way to address this problem. The first step in developing a COS is to identify and map all outcomes. Objectives: To identify outcomes reported in systematic reviews of hospital AMS interventions. Methods: Cochrane Database of Systematic Reviews, MEDLINE and Embase were searched for systematic reviews published up until August 2019 of interventions relevant to reducing unnecessary antimicrobial use for inpatient populations in secondary care hospitals. The methodological quality of included reviews was assessed using AMSTAR-2, A (revised) MeaSurement Tool to Assess systematic Reviews. Extracted outcomes were analysed using deductive and inductive thematic analysis. A list of overarching (unique) outcomes reflects the outcomes identified within the systematic reviews. Results: Forty-one systematic reviews were included. Thirty-three (81%) systematic reviews were of critically low or low quality. A long list of 1739 verbatim outcomes was identified and categorized under five core areas of COMET (Core Outcome Measures in Effectiveness Trials) taxonomy: ‘resources use’ (45%), ‘physiological/clinical’ (27%), ‘life impact’ (16%), ‘death’ (8%) and ‘adverse events’ (4%). A total of 421 conceptually different outcomes were identified and grouped into 196 overarching outcomes. Conclusions: There is significant heterogeneity in outcomes reported for hospital AMS interventions. Reported outcomes do not cover all domains of the COMET framework and may miss outcomes relevant to patients (e.g. emotional, social functioning, etc.). The included systematic reviews lacked methodological rigour, which warrants further improvements.
Funding Information
  • University of Aberdeen

This publication has 58 references indexed in Scilit: