Abstract / Summary
Abstract Background. Carbapenem-resistant Enterobacteriaceae (CRE) are a critical-priority pathogen, and prior antimicrobial exposure is the principal modifiable driver of their emergence through selective pressure. Local antibiogram data are essential for rational empirical prescribing but remain scarce for the Chattogram region of Bangladesh. We examined the prevalence of CRE infection, the relationship between prior antibiotic exposure and resistance phenotype, and the antibiogram profile of isolates. Methods. In this retrospective observational study, 15,714 patients observed at a tertiary care centre between January 2021 and December 2024 were screened; 113 with microbiologically confirmed CRE infection formed the analytic cohort. Demographic, comorbidity, prior-exposure and outcome data were abstracted with a standardised questionnaire; prior exposure was defined as documented receipt of an antibiotic class within 30 days. Susceptibility was determined for 19 agents and multidrug resistance (MDR) defined per Magiorakos criteria. Prevalence was estimated with Wilson 95% confidence intervals (CI); associations were tested with the Fisher exact test and the Cochran–Armitage test for trend. Results. Period prevalence of CRE infection was 0.72% (95% CI 0.60–0.86%), or 7.19 per 1000 patients observed. Prior antibiotic exposure was near-universal (98/113, 86.7%), most often carbapenem (80.5%) and third-/fourth-generation cephalosporin (77.9%). Every exposure class was significantly associated with an MDR phenotype (all P < 0.001), and a clear dose–response gradient was observed: MDR rose from 0.0% among patients exposed to no prior class and 0.0% among those exposed to 1–2 classes, to 22.2% at 3–4 classes and 53.9% among those who had received all five (Cochran–Armitage z = 4.96, P < 0.0001). Duration of definitive therapy for the index episode was not associated with resistance (P = 0.143) or mortality (P = 0.677). Antibiogram profiling showed extensive resistance to cephalosporins and fluoroquinolones (> 84% for most agents) and to meropenem (77.1% of 48 tested), whereas colistin (81.6% susceptible) and tigecycline (88.2% susceptible) retained the greatest in-vitro activity. Overall, 31 of 106 patients with a determinate outcome died (29.2%). Conclusions. CRE infection affected approximately 7 per 1000 patients observed. Cumulative prior antibiotic exposure showed a strong, graded association with multidrug resistance, consistent with the selective-pressure hypothesis and providing direct local evidence for carbapenem-sparing empirical strategies. The antibiogram indicates that colistin and tigecycline remain the principal reserve agents in this setting, and that empirical reliance on cephalosporins or fluoroquinolones is no longer defensible.