Impact of an antimicrobial stewardship programme on antibiotic utilization and resistance burden in patients with acute leukaemia: an 11-year longitudinal cohort study using interrupted time-series analysis

Abstract

Background: Antimicrobial resistance (AMR), driven by inappropriate and overuse of antibiotics, poses a significant threat, especially to patients with acute leukaemia.

Objectives: To evaluate the impact of antimicrobial stewardship programmes (ASPs) on antibiotic use and analyse temporal changes in bloodstream infections (BSI) caused by AMR organisms.

Methods: We performed a retrospective, interventional, longitudinal cohort study spanning an 11-year period. ASPs included optimizing antibiotic use, enhancing tracking and reporting systems and delineating leadership and accountability. A segmented regression model of interrupted time series was used to evaluate the trend of antibiotic consumption and BSI with AMR organisms after the interventions.

Results: A total of 3296 BSI episodes with 454 419 days of therapy (DOT) from 7754 patients were obtained. ASPs were significantly associated with an immediate reduction [-70.03 DOT/1000 patient-days (PD), P = 0.036] and a decreasing trend (-11.65 DOT/1000 PD per quarter, P < 0.001) in overall antibiotic use. The increasing incidence of BSI with AMR before ASP intervention was notably curbed and revealed a decreasing trend (slope change: -0.06 BSI/1000 PD per quarter, P = 0.002). The decreasing trend was more significant for Enterobacterales: ciprofloxacin-resistant and ESBL-producing isolates showed a slope change of -0.06 BSI/1000 PD and -0.08 BSI/1000 PD per quarter, respectively (all P < 0.05). However, Pseudomonas aeruginosa BSI increased.

Conclusions: Multidimensional ASPs effectively reduced both the immediate and trends in overall antibiotic usage even in patients with acute leukaemia. Additionally, there was a notable decrease in the incidence of BSI caused by AMR organisms, particularly among Enterobacterales.

Figure 1.

Consumption of antibiotics. (a) All antibiotics; (b) carbapenems; (c) ciprofloxacin; (d) glycopeptides; (e) fourth-generation cephalosporins; (f) third-generation cephalosporins; (g) piperacillin–tazobactam; (h) linezolid. Y-axis indicates DOT standardized by 1000 PD quarterly. Dotted vertical lines denote the commencement of antimicrobial stewardship interventions. PD, patient-days; Q, quarter.

Figure 2.

ITS analysis for antibiotic consumption. (a) All antibiotics; (b) carbapenems; (c) ciprofloxacin; (d) glycopeptides; (e) fourth-generation cephalosporins; (f) third-generation cephalosporins; (g) piperacillin–tazobactam; (h) linezolid. Y-axis indicates DOT standardized by 1000 PD quarterly. The two vertical dotted lines represent the roll-in period after antimicrobial stewardship interventions (fourth quarter of 2016). PD, patient-days; Q, quarter.

Figure 3.

ITS analysis for the incidence of BSI. (a) Incidence by resistant organisms; (b) incidence by non-resistant organism. Y-axis indicates the incidence rate standardised by 1000 PD quarterly. The two vertical dotted lines represent the roll-in period after antimicrobial stewardship interventions (fourth quarter of 2016). PD, patient-days; Q, quarter.

Figure 4.

ITS analysis for the incidence of BSI by Enterobacterales. (a) Enterobacterales; (b) ciprofloxacin-resistant isolates; (c) ESBL-producing isolates; (d) CRE. Y-axis indicates the incidence rate standardized by 1000 PD quarterly. The two vertical dotted lines represent the roll-in period after antimicrobial stewardship interventions (fourth quarter of 2016). BSI, bloodstream infections; CRE, carbapenem-resistant Enterobacterales; PD, patient-days; Q, quarter.