Multidrug-Resistant Bacterial Infections in Intensive Care Units: A Retrospective Analysis of Pathogen Epidemiology, Antimicrobial Resistance, and Clinical Risk Factors

Background Adult ICU patients are highly vulnerable to bacterial infections because of critical illness, invasive devices, prolonged hospitalization and antimicrobial exposure. Increasing multidrug resistance, particularly among Gram-negative pathogens, complicates empirical treatment and requires reliable local epidemiological surveillance and effective stewardship planning. Aims To characterize ICU-associated bacterial pathogens, antimicrobial-resistance patterns, MDR burden, carbapenem resistance and clinical factors associated with MDR infection in adults. Methods A retrospective study analyzed records from ICU patients. Specimens were cultured and isolates identified using routine procedures. Susceptibility was interpreted using recognized breakpoint standards. MDR was defined as non-susceptibility to at least one agent in three or more antimicrobial categories, while carbapenem resistance was assessed among tested Gram-negative isolates. Descriptive statistics and multivariable logistic regression were performed using Python; associations were reported as adjusted odds ratios with 95% confidence intervals. Results Among 1,135 specimens, isolates demonstrated a substantial resistance burden. Acinetobacter baumannii showed highest MDR prevalence and carbapenem resistance, followed by Klebsiella pneumoniae and Pseudomonas aeruginosa. Carbapenem resistance occurred in 87.3% of A. baumannii, 51.0% of P. aeruginosa, 42.8% of K. pneumoniae and 9.7% of Escherichia coli. Recent antibiotic exposure was the strongest factor associated with MDR infection (adjusted OR, 3.18; 95% CI, 2.32–4.36; p < 0.001), followed by mechanical ventilation and ICU stay exceeding seven days. Diabetes and chronic kidney disease were not independently associated after adjustment. Conclusion The findings support continuous ICU surveillance, careful antibiotic use and strengthened infection-prevention practices, while recognizing that this single-cohort study cannot establish causality or national prevalence.