Laboratory-based surveillance of antimicrobial non-susceptibility in Enterobacter species over 12 years at a tertiary-care hospital in Saudi Arabia

Background Enterobacter species are increasingly associated with antimicrobial non-susceptibility, but long-term surveillance of temporal trends across antibiotic classes is limited in southern Saudi Arabia. Methods We retrospectively studied non-duplicate Enterobacter isolates from adults at a tertiary-care hospital in southern Saudi Arabia (August 2012–August 2024), interpreted by contemporaneous CLSI criteria. Temporal trends were modelled by logistic regression as the annual percent change (APC) in the odds of non-susceptibility, adjusted for care setting, specimen source, age, and sex. Carbapenem-resistant Enterobacterales (CRE) and difficult-to-treat resistance (DTR; Kadri criteria) were analysed over time; sensitivity analyses excluded partial years and used breakpoint-comparable periods. Results Of 1,319 Enterobacter isolates, 91.8% belonged to the Enterobacter cloacae complex. Adjusted trends declined for amikacin (APC −6.1%), gentamicin (−9.4%) and cefepime (−9.1%) and rose for carbapenems (+13.1% to +15.8%), levofloxacin (+6.3%) and piperacillin–tazobactam (+7.0%), with ciprofloxacin stable. The levofloxacin and piperacillin–tazobactam increases did not persist in breakpoint-comparable analyses and are not unequivocal biological increases. Inpatient isolates were more non-susceptible than outpatient isolates by approximately 16–39 percentage points (gentamicin 31.7% vs 14.8%; piperacillin–tazobactam 38.7% vs 23.1%; cefotaxime 63.9% vs 41.5%; cefepime 45.7% vs 28.8%; ceftazidime 61.4% vs 35.3%; ceftriaxone 57.4% vs 18.8%; all p ≤ 0.003). CRE rose early and remained common (APC + 22.8%; 28.7% overall), whereas DTR rose later and stayed infrequent (7.5% of eligible isolates; APC + 16.4%; highest annual value 20.2% in 2023); this difference in timing and overall frequency, not the partially sampled year 2024, indicates dissociation between carbapenem non-susceptibility and loss of all first-line agents. Conclusion Non-susceptibility among Enterobacter species was heterogeneous and antibiotic-specific. The dissociation between CRE and DTR in timing and overall frequency supports using clinically meaningful phenotypes alongside antibiotic-specific trend analysis for stewardship and reporting; the levofloxacin and piperacillin–tazobactam increases warrant caution, as they did not persist in breakpoint-comparable analyses.