An inflammation–nutrition composite index and its dynamic change for predicting postoperative wound complications

Postoperative wound complications are an important cause of morbidity, and routinely available inflammatory markers may assist with early risk assessment. This study evaluated a newly proposed exploratory composite index, the Wound Inflammation–Nutrition Index (WINA), and its change between postoperative day 1 and postoperative day 3 (ΔWINA) in relation to postoperative wound complications. This retrospective observational study included 86 patients who underwent general surgical procedures between January 2020 and December 2025. WINA was calculated by multiplying the C-reactive protein-to-albumin ratio by the neutrophil-to-lymphocyte ratio at postoperative days 1 and 3. ΔWINA was calculated as the POD3 value minus the POD1 value. Logistic regression, receiver operating characteristic analysis, model comparison, bootstrap internal validation, and sensitivity analyses were performed. Postoperative wound complications occurred in 24 patients (27.9%). WINA values at POD1 and POD3 and ΔWINA were higher in patients with wound complications (all p < 0.001). In the parsimonious multivariable model, ΔWINA per 100-unit increase was associated with wound complications (adjusted OR: 2.12, 95% CI: 1.44–3.13, p = 0.001), together with diabetes mellitus (adjusted OR: 2.89, 95% CI: 1.01–8.29, p = 0.048) and operative time per 30-minute increase (adjusted OR: 1.67, 95% CI: 1.08–2.58, p = 0.021). ΔWINA had an AUC of 0.872 (95% CI: 0.789–0.941), with 83.3% sensitivity, 80.6% specificity, 62.5% positive predictive value, and 92.6% negative predictive value at a cut-off of 245.0. Its discrimination did not differ significantly from WINA at POD3. Adding ΔWINA to the clinical model increased the AUC from 0.772 to 0.903 (ΔAUC: 0.131, p = 0.003). The association remained after excluding complications diagnosed on or before POD3. ΔWINA was associated with postoperative wound complications and improved the performance of a parsimonious clinical model. However, it did not outperform WINA measured at POD3, and the derived cut-off requires external validation before clinical use.