C-Reactive Protein Values after Colorectal Resection: Can We Discharge a Patient with a C-Reactive Protein Value >100? A Retrospective Cohort Study

BACKGROUND: C-reactive protein is a useful negative predictive test for the development of anastomotic leakage following colorectal surgery. Evolution of procedures (laparoscopy, enhanced recovery program, early discharge, complex redo surgery) may influence C-reactive protein values; however, this is poorly studied to date. OBJECTIVE: The aim of this study is to evaluate C-reactive protein as an indicator of postoperative complication and as a predictor for discharge. DESIGN: This is retrospective study of a consecutive monocentric cohort. SETTINGS: All patients undergoing a colorectal resection with anastomosis (2014-2015) were included. MAIN OUTCOMES MEASURES: C-reactive protein, leukocytosis, type of resection, and postoperative course were the primary outcomes measured. RESULTS: A total of 522 patients were included. The majority had either a colorectal (n = 159, 31%) or coloanal anastomosis (n = 150, 29%). Overall morbidity was 29.3%. C-reactive protein was significantly higher among patient having intra-abdominal complications at an early stage (day 1-2) (164.6 vs 136.2; p = 0.0028) and late stage (day 3-4) (209.4 vs 132.1; p < 0.0001). In multivariate analysis, early C-reactive protein was associated with BMI (coefficient, 4.9; 95% CI, 3.2-6.5; p < 0.0001) and open surgical procedures (coefficient, 43.1; 95% CI, 27-59.1; p < 0.0001), while late C-reactive protein value was influenced by BMI (coefficient, 4.8; 95% CI, 2.5-7.0; p = 0.0024) and associated extracolonic procedures (coefficient, 34.2; 95% CI, 2.7-65.6; p = 0.033). Sensitivity, specificity, negative predictive values, and positive predictive values for intra-abdominal complication were 85.9%, 33.6%, 89.3%, and 27.1% for an early C-reactive protein <100 mg/L and 72.7%, 75.4%, 89.4%, and 49.2% for a late C-reactive protein <100 mg/L. Four hundred seven patients with an uneventful postoperative course were discharged at day 8 ± 6.4 with a mean discharge C-reactive protein of 83.5 ± 67.4. Thirty-eight patients (9.3%) were readmitted and had a significantly higher discharge C-reactive protein (138.6 ± 94.1 vs 77.8 ± 61.2, p = 0.0004). Readmission rate was 16.5% for patients with a discharge C-reactive protein >100 mg/L vs 6% with C-reactive protein <100 mg/L (p = 0.0008). For patients included in an enhanced recovery program (discharge at day 4 ± 2.4), the threshold should be higher because discharge is around day 3 or 4. With a C-reactive protein <140, readmission rate was 2% vs 19%, (p = 0.056). LIMITATIONS: This study includes retrospective data. CONCLUSION: C-reactive protein <100 mg/L is associated with a lower risk of intra-abdominal complication and readmission rates.