Targeting indocyanine green where it matters most: a CRP-stratified cohort study in acute cholecystitis.

Abstract

Background: Intraoperative indocyanine green (ICG) fluorescence cholangiography has been linked to fewer conversions to open surgery in emergency cholecystectomy, but it adds cost, and evidence for its effect on conversion and operative time is graded moderate to low even by the consensus statement recommending routine use. We asked whether preoperative C-reactive protein (CRP), a marker of surgical difficulty and conversion risk, identifies patients in whom ICG's benefit is greatest.

Methods: Prospective cohort of 221 patients undergoing emergency laparoscopic cholecystectomy for acute cholecystitis, stratified by ICG use (85 ICG vs 136 conventional). We tested the interaction between CRP and ICG use on conversion using Firth penalized logistic regression, and on operative time, length of stay, and complications using linear/logistic regression with a log-CRP sensitivity analysis. We compared conversion rates across CRP tertiles and calculated a CRP-stratified estimated absolute risk reduction (ARR) and number needed to treat (NNT).

Results: Conversion occurred only in the conventional arm (12/136, 8.8%; 0/85 ICG), rising from 0 to 15.4% across CRP tertiles. The formal CRP×ICG interaction for conversion was not significant (Firth model, p = 0.878), reflecting the absence of ICG-arm events; the ARR associated with ICG use rose from not estimable to 9.8 points (95% CI 3.9-22.5; NNT≈10) in the middle tertile and 15.4 points (95% CI 8.0-27.5; NNT 6.5) in the highest. The interaction was significant for operative time (p = 0.045), strengthening under a log-CRP sensitivity model (p = 0.001), with CRP-related lengthening of surgery markedly smaller in the ICG arm.

Conclusions: In this cohort, ICG's clinical benefit was not uniform across the CRP range. The CRP×ICG interaction reached significance for operative time; for conversion, the pattern was consistent and clinically plausible but not statistically significant, and should be regarded as hypothesis-generating. A CRP-informed strategy for prioritising ICG may be worth testing prospectively.