Multi-marker approach using C-reactive protein, procalcitonin, neutrophil CD64 index for the prognosis of sepsis in intensive care unit: a retrospective cohort study

Background We aimed to explore the prognostic utilities of C-reactive protein (CRP), procalcitonin (PCT), neutrophil CD64 (nCD64) index, in combination or alone, in septic patients. Methods We retrospectively included 349 septic patients (based on Sepsis 3.0 definition). The primary outcome was 28-day all-cause mortality. Cox regression model, receiver-operating characteristic (ROC) curve, reclassification analysis, Kaplan–Meier survival curves were performed to evaluate the predictive efficacy of the above parameters. Results CRP, nCD64 index were independent predictors of 28-day mortality for sepsis in the Cox regression model [CRP, HR 1.004 (95% CI 1.002–1.006), P < 0.001; nCD64 index, HR 1.263 (95% CI 1.187–1.345, P < 0.001]. Area under the ROC curve (AUC) of CRP, PCT, nCD64 index, nCD64 index plus PCT, nCD64 index plus CRP, were 0.798 (95% CI 0.752–0.839), 0.833 (95% CI 0.790–0.871), 0.906 (95% CI 0.870–0.935), 0.910 (95% CI 0.875–0.938), 0.916 (95% CI 0.881–0.943), respectively. nCD64 plus CRP performed best in prediction, discrimination, and reclassification of the 28-day mortality risk in sepsis. The risk of 28-day mortality increased stepwise as the number of data exceeding optimal cut-off values increased. Conclusions nCD64 index combined with CRP was superior to CRP, PCT, nCD64 index and nCD64 index plus PCT in predicting 28-day mortality in sepsis. Multi-marker approach could improve the predictive accuracy and be beneficial for septic patients.