Associated factors and dynamic coagulation changes of cefoperazone/sulbactam-associated coagulopathy: a retrospective case-control study from a single-center real-world ADR database

Objectives Cefoperazone/sulbactam is associated with severe coagulopathy, believed to be mediated by its N-methylthiotetrazole side chain through inhibition of the vitamin K cycle; however, comprehensive real-world data on risk factors remain limited. This study aimed to identify potential factors associated with cefoperazone/sulbactam-induced coagulopathy and to characterize the dynamic changes in coagulation parameters in affected patients. Methods This retrospective case-control study identified 228 patients with adverse drug reactions (ADRs) from a hospital database. Patients were divided into a coagulopathy group (n = 25) and an other-ADR control group (n = 203); baseline characteristics and coagulation parameters (PT, INR, APTT, PLT) were compared. Firth penalized logistic regression was used in an exploratory analysis to identify factors independently associated with coagulopathy. Results Among the 228 patients with ADRs, 25 (11.0%) developed coagulopathy. Compared with the control group, the case group was significantly older (70.08 vs. 61.92 years, P = 0.020) and had higher proportions of renal disease (52.0% vs. 11.3%), heart failure (24.0% vs. 5.9%), diabetes (36.0% vs. 14.8%), and diuretic (80.0%) or anticoagulant (56.0%) use (all P < 0.05). Treatment duration was significantly longer (9.68 vs. 2.98 days, P < 0.001), with 60.0% versus 13.3% receiving therapy for ≥7 days (P < 0.001), whereas daily dose did not differ (P = 0.900). In Firth penalized logistic regression, kidney disease (OR 8.72, 95% CI 3.11–26.31) and treatment duration ≥7 days (OR 10.83, 95% CI 4.08–31.61) were independently associated with coagulopathy, whereas age ≥65 years was not (OR 2.03, 95% CI 0.74–6.11; P = 0.171). At onset, PT, INR, and APTT were markedly prolonged but normalized after drug discontinuation (P < 0.001), whereas platelet counts remained stable (P > 0.05), a pattern consistent with vitamin K cycle inhibition. Conclusion In univariate analysis, advanced age, renal impairment, heart failure, diabetes, diuretic/anticoagulant use, and treatment duration ≥7 days were identified as potential factors. Kidney disease and treatment duration ≥7 days remained independently associated in penalized logistic regression. These findings underscore the need for risk-stratified monitoring protocols and suggest prophylactic vitamin K may be considered in high-risk patients. Future prospective studies with biomarker profiling are warranted.