PREVALENCE OF EXTENDED-SPECTRUM BETA LACTAMASE AND CARBAPENEMASE-ENCODING GENES IN SURGICAL SITE INFECTION ISOLATES: IMPLICATIONS FOR SURGICAL ANTIMICROBIAL PROPHYLAXIS

Abstract

Background: Cefazolin remains the guideline-recommended agent for surgical antimicrobial prophylaxis in most procedures, but Indian practice has drifted towards third-generation cephalosporins, and the prevalence of extended-spectrum β-lactamase (ESBL) and carbapenemase determinants among the organisms that cause surgical site infection (SSI) is largely unquantified at the level of individual institutions. Prophylaxis policy cannot be rationally set without those data. Objectives: To determine the prevalence of ESBL- and carbapenemase-encoding genes among Gram-negative isolates from post-surgical wound infections at a tertiary care institute, to describe their distribution by surgical specialty, wound class and study quarter, and to quantify the in vitro coverage that candidate prophylaxis regimens would provide against these organisms. Methods: A prospective observational study was conducted jointly by the Departments of Microbiology and Surgery at NAMO Medical Education and Research Institute, Silvassa, over twelve months from July 2025 to July 2026. SSI was defined by CDC/NHSN criteria. Gram-negative isolates underwent identification and susceptibility testing interpreted per CLSI M100, and multiplex PCR for blaCTX-M, blaTEM, blaSHV, blaNDM, blaOXA-48, blaKPC, blaVIM, blaIMP, blaOXA-23 and blaOXA-51. The prophylactic regimen administered was recorded prospectively. Prevalence is reported with Wilson 95% confidence intervals; trends across quarters were tested by the Cochran–Armitage test. Prophylaxis–pathogen discordance was defined as recovery of an isolate non-susceptible in vitro to the agent the patient had actually received. Coverage of candidate regimens was modelled at isolate level, a regimen counting as covering an isolate if that isolate was susceptible to any component. Results: Of 2,184 procedures monitored, 178 met criteria for SSI (8.15%; 95% CI 7.10–9.40); 156 were culture positive, yielding 187 isolates, of which 135 (72.2%) were Gram-negative. Among 88 Enterobacterales, 56 (63.6%; 95% CI 53.2–72.9) carried an ESBL determinant and 28 (31.8%; 95% CI 23.0–42.1) a carbapenemase determinant. blaCTX-M predominated (47; 53.4%), followed by blaTEM (39; 44.3%) and blaSHV (22; 25.0%); blaNDM (17; 19.3%) and blaOXA-48 (13; 14.8%) were the commonest carbapenemases, co-carried by 6 isolates. Twenty-five isolates carried both an ESBL and a carbapenemase determinant, and thirteen distinct determinant combinations were observed. blaOXA-23 was present in 16 of 21 Acinetobacter baumannii (76.2%). ESBL determinants were more frequent in contaminated or dirty wounds (38/52 versus 18/36; OR 2.71, 95% CI 1.11 6.65; p = 0.027). Ceftriaxone with or without metronidazole was the regimen administered to 96 of 156 patients (61.5%), and cefazolin to 22 (14.1%); overall, 117 of 151 evaluable patients (77.5%; 95% CI 70.2–83.4) yielded an isolate non-susceptible to the agent they had received. Modelled coverage of all 187 isolates was 19.8% for cefazolin, 21.4% for cefuroxime and 24.6% for ceftriaxone; adding gentamicin to cefazolin raised coverage to 49.2%, and no regimen short of meropenem with vancomycin exceeded 60%. Conclusions: Two-thirds of Enterobacterales causing SSI at this centre carried an ESBL determinant and almost one-third a carbapenemase determinant, a burden that leaves cephalosporin prophylaxis — whether first- or third generation — without meaningful in vitro activity against the organisms that actually caused breakthrough infection. These data do not support broadening prophylaxis to carbapenems, which would accelerate the very resistance they seek to circumvent, and they cannot by themselves establish what prophylaxis should be, because isolates from patients who developed SSI despite prophylaxis are a selected sample. They do establish that the local determinant burden is high enough to warrant a formal institutional prophylaxis policy, routine reporting of a surgical antibiogram, and restriction of third-generation cephalosporin prophylaxis, for which no rationale survives these data.