Verification of quality assurance for blood culture surveillance using 6 years of data from the Japan Infection Prevention and Control Conference for National and Public University Hospitals.

The importance of blood culture has been widely recognized, and there is a need for monitoring to evaluate the accuracy of blood culture that reflects domestic healthcare systems. In this study, we assessed 6-year trends in blood culture quality assurance data. The Japan Infection Prevention and Control Conference for National and Public University Hospitals conducted yearly blood culture surveillance at 52 national public university hospitals from 2015 to 2020. Statistical analysis showed that comparison with the previous year showed significant differences in the number of blood cultures per 1000 patient-days in all years. The number of blood cultures per 1000 admissions was not significantly different in 2017 and 2018, but significant differences were shown in all other years. The multiple blood culture set rate was significantly different between non-pediatric inpatients and outpatients but not between pediatric inpatients and outpatients. The contamination rate did not differ significantly. For all parameters, significant differences were found when comparing 2015 and 2020. Our survey showed that although the sample number improved over time, even the most recent values for 2020 were lower than Cumitech's targets. It is difficult to assess whether these sample numbers are appropriate because target values have not been set for the various types of hospitals in Japan. Surveillance is a useful tool for monitoring quality assurance for blood culture. All parameters improved over the 6-year period, but it is necessary to establish a benchmark for evaluating optimization. We will continue to monitor quality assurance and work on setting benchmarks.