Evaluation of appropriateness of antibiotic prescribing in primary healthcare institutions in China using proxy indicator

Abstract

Our objectives were to develop a set of proxy indicators (PIs) suited for assessing antibiotic use appropriateness in China's primary healthcare institutions (PHIs), and assess performance scores of these PIs while exploring factors that influence the antibiotic appropriateness. We selected potential PIs for the PHIs through a RAND-modified Delphi procedure, and assessed clinimetric properties, focusing on measurability, applicability, and potential for improvement. PIs with favorable clinimetric properties were used to evaluate antibiotic prescription appropriateness by calculating performance scores of each PI. Institutions were categorized into three clusters representing different levels of appropriateness. We used the chi-square test and an ordinal logistic regression model at PHI level to explore factors influencing antibiotic appropriateness. Eighteen PIs were developed through two rounds of online surveys and one face-to-face meeting involving 20 stakeholders. All PIs met the clinimetric properties criteria and were used to analyze 209,662 antibiotic prescriptions across 269 PHIs. The percentage of PHIs meeting the target ranged from 3.1% to 69.3%, with 6 PIs below 10%. The appropriateness of antibiotic prescriptions was significantly associated with percentages of patients' gender of the PHIs. The varied and suboptimal performance of the PIs indicated the need for diverse efforts to enhance the rational antibiotic use at PHI level. It was necessary to devise distinct sets of PIs for diverse settings in future endeavors. This work was supported by the National Natural Science Foundation of China (grant numbers 72374009, 81973294).