Diagnostic performance of procalcitonin/lactate ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) in predicting bacterial infection across different severities of diabetic ketoacidosis

Abstract

Abstract Background Diabetic ketoacidosis is frequently accompanied by systemic inflammatory responses that complicate the early identification of bacterial infection. Methods We retrospectively analyzed 158 adult patients admitted with diabetic ketoacidosis and evaluated the diagnostic performance of the PLR(procalcitonin-to-lactate ratio) and the NLR (neutrophil-to-lymphocyte ratio) using ROC analysis. Results Patients with infection were significantly older than those without infection (53.97 vs. 44.42 years, p < 0.001). Procalcitonin, lactate, PLR, and NLR levels were significantly higher in infected patients (p < 0.05 for all). Overall ROC analysis demonstrated significant diagnostic value for both PLR (AUC 0.791, 95% CI 0.722–0.861) and NLR (AUC 0.730, 95% CI 0.651–0.808). When stratified by diabetic ketoacidosis severity, PLR demonstrated higher diagnostic accuracy in patients with severe DKA (AUC 0.875, p = 0.034), whereas NLR did not remain statistically significant in this subgroup (p = 0.572). In multivariable analyses adjusted for age and renal function, both PLR and NLR remained independently associated with infection, with a stronger association observed for PLR. Conclusions The procalcitonin-to-lactate ratio appears to be a reliable biomarker for identifying bacterial infection in diabetic ketoacidosis and may support more rational antibiotic decision-making.