Selection of hospital antimicrobial prescribing quality indicators: a consensus among German antibiotic stewardship (ABS) networkers

Purpose Simple, valid, and evidence-based indicators to measure the quality of antimicrobial prescribing in acute-care hospitals are urgently needed and increasingly requested by policymakers. The aim of this study was to develop new consensus quality indicators (QIs) for hospital antibiotic stewardship (ABS) and infection management which will be further evaluated for internal quality management and external quality assessment in Germany. Methods Based on an extensive literature review, the Austrian-German hospital ABS Guideline Committee and selected members of the German ABS Expert Network discussed and drafted a list of 99 potential indicators for hospitals that reflect structural prerequisites for ABS (35 items), ABS core activities (18 items), additional ABS measures (5 items), and process of care indicators (both generic and disease-specific-12 and 29 items, respectively). Questionnaires were mailed to German ABS experts and healthcare professionals with further education in ABS. Participants scored (on a nine-point Likert scale) relevance (clinical, ecological/resistance, economical/expenses) and presumed practicability (six categories: clarity of definition, effort to collect data, barrier to implementation, verifiability, suitability for external quality assessment, quality gap), taking into account their local work environment. The scores were processed according to the RAND/UCLA appropriateness method, and QIs were judged relevant if the median (clinical + ecological and/or economical) scores were >6. The indicators thus assessed to be potentially relevant were then filtered according to their practicability. Highly relevant QIs with borderline practicability scores and items with disagreements and overlapping areas were re-discussed in a final multidisciplinary panel consensus workshop convened in November 2012. Results Of the 340 questionnaires that were mailed, 75 questionnaires were completed and returned. Of 99 initially proposed items, 32 were excluded due to insufficient scores. Of the remaining 67 items, 21 structural and 21 process of care QIs were finally selected, including four QIs with high clinical and ecological but limited economical relevance, and three QIs with high clinical and economical but limited ecological relevance. Among the selected QIs, efforts to collect data and implementation barriers were scored as suboptimal in many cases. Conclusions A catalog of consensus structural and process of care ABS-QIs was established. These should undergo further pilot and feasibility studies in the German hospital healthcare sector. The panelists were most critical regarding resource use/complexity issues and presumed implementation barriers. How this may limit applicability of QIs remains to be determined.