Clinimetric properties and suitability of selected quality indicators for assessing antibiotic use in hospitalized adults: a multicentre point prevalence study in 24 hospitals in Germany

Abstract

Objectives: The capability to measure and monitor the quality of antibiotic prescribing is an important component of antibiotic stewardship (ABS) programmes. Several catalogues of consensus-based structure and process-of-care quality indicators (QIs) have been proposed, but only a few studies have tested and validated ABS QIs in practice tests. This multicentre study determined the clinimetric properties and suitability of a set of 33 process QIs for ABS that had earlier been developed and in part recommended in a German-Austrian hospital ABS practice guideline.

Methods: Two point prevalence surveys were conducted in a convenience sample of 24 acute care hospitals throughout Germany, and data of all screened adult inpatients with prescription of a systemic antibiotic at a given day (n=4310) were included in the study. For each QI, the following clinimetric properties were assessed: applicability, feasibility, performance, case mix stability and interobserver reliability.

Results: Eighteen QIs were considered sufficiently feasible, applicable and reliable, and had adequate room for improvement. The finally selected QIs primarily cover antibiotic therapy of common infections (bloodstream infection, pneumonia and urinary tract infection), while two of the QIs each address surgical prophylaxis and general aspects of antibiotic administration.

Conclusions: Practice tests may be important to test the suitability of consensus process-of-care QIs in the field of hospital ABS. The 18 selected QIs considered suitable enough for hospital ABS in this study should be regarded as priority QIs useful for internal quality control and assurance. More research and additional practice tests may be needed to confirm their suitability for external quality assessment schemes.