Long-term effect of carbapenem preauthorization: an interrupted time-series study over 20 years

Preauthorization of carbapenems is considered effective in reducing carbapenem overuse; however, long-term evaluations are lacking.

We aimed to evaluate the long-term effects of judicious carbapenem restriction over almost two decades in a tertiary teaching hospital in Japan.

Interrupted time-series (ITS) analysis was applied to investigate changes in the level and trend of antimicrobial use density (AUD) of carbapenems by comparing the pre-intervention (2004 and 2005) and intervention (2006-2023) periods. Furthermore, this study explored the relationship between carbapenem AUD in hospitalized patients, the mortality rate of hospital-acquired bacteraemia, and the prevalence of bacteraemia caused by antibiotic-resistant nosocomial pathogens.

The ITS analysis demonstrated remarkably significant reductions in the level change of carbapenem AUD following the preauthorization (-0.367 per 100 patient days (95% confidence interval (CI), -0.131 to -0.603; P=0.002). In addition, the trend shift was -0.014 per 100 patient days (95% CI, -0.001 to -0.028; P=0.048). Following the implementation of carbapenem preauthorization, no increase in the 28-day mortality due to nosocomial bacteraemia was observed. Significant positive correlations were found between carbapenem AUD and the prevalence of meticillin-resistant Staphylococcus aureus (correlation coefficient (ρ) = 0.77 (P<0.001)), meropenem-resistant in Pseudomonas aeruginosa (ρ = 0.85 (P<0.001)), and Acinetobacter baumannii (ρ = 0.80 (P<0.001)).

Long-term implementation of carbapenem preauthorization consistently proved to be effective with no adverse consequences for hospital epidemiology in a 20-year ITS analysis.