EP727 - The Impact Of C-Reactive Protein On CT Scan Utilization After Colorectal Surgery: A Retrospective Review

Purpose/Background: Postoperative anastomotic leak is a severe complication often investigated with computed tomography (CT) based on clinical suspicion, or more recently based on routine postoperative serial C-Reactive Protein (CRP) measurements. This practice may contribute to increased CT utilization with possible increased healthcare costs and patient exposure to radiation. This study evaluates the real-world impact of routine CRP monitoring on the utilization and diagnostic yield of CT scans following elective colorectal surgery. We aimed to quantify how this biomarker-driven approach influences clinical decision-making and patient outcomes.

Methods/Interventions: A retrospective review of 611 consecutive patients from a single tertiary care center undergoing colorectal surgery with anastomosis between January 2021 and December 2023 was performed. Routine postoperative CRP monitoring was standard practice for this cohort. Full, detailed data extraction was conducted for 189 patients who met at least one of the following inclusion criteria: an elevated CRP above predefined cutoffs (CRP>110 on POD1, 148 on POD 3, 115 on POD5 or 50 point increase between any 2 measurements), a postoperative CT scan, or a re-intervention. The performance of elevated CRP as a screening test for a clinically significant anastomotic leak-defined as a complication requiring surgical or radiological re-intervention-was evaluated on the subset of 166 patients from this cohort included on the basis of their CRP.

Results/Outcomes: The rate of clinically significant anastomotic leak requiring re-intervention was 2.9% in the total screened population and 9.5% amongst the fully reviewed cohort. A postoperative CRP elevation was a powerful predictor of imaging, associated with a 7-fold increase in the odds of receiving a CT scan (OR 6.95, p< 0.0001). The subsequent diagnostic yield for a leak or abscess was significantly higher in the CRP-positive group compared to the CRP-negative group (33% vs. 9%, p=0.04). In CRP-negative patients, the yield remained low unless prompted by clear clinical signs of sepsis, where it rose to 40%. As a screening test, an elevated CRP demonstrated a high sensitivity of 89% and a negative predictive value (NPV) of 94% for predicting the need for re-intervention, with a positive predictive value (PPV) of 12.2%.

Conclusion/Discussion: Postoperative CRP elevation effectively guides and rationalizes CT utilization, leading to a respectable diagnostic yield for significant complications. This study confirms that serial CRP is a highly sensitive screening test. Its high NPV provides powerful clinical reassurance, supporting its routine use in safely de-escalating investigations for low-risk patients. The integration of serial CRP into standard postoperative pathways is a valuable, evidence-based strategy for reducing unnecessary imaging and improving healthcare resource allocation.