Background: Patients with high-risk acute cholecystitis are challenging to manage because many cannot tolerate early surgery. Percutaneous transhepatic gallbladder drainage (PTGBD) is therefore commonly used; however, the clinical response after drainage is not always satisfactory. Reliable indicators for early identification of patients at risk of poor treatment response remain limited. Although inflammatory and nutritional biomarkers have shown potential value in risk stratification, their role in predicting response to PTGBD has not been fully clarified. Accordingly, this study aims to evaluate the predictive value of the preoperative C-reactive protein/albumin ratio (CAR) for poor treatment response in high-risk patients with acute cholecystitis undergoing PTGBD.

Methods: Clinical data from 309 patients with high-risk acute cholecystitis who underwent PTGBD at our institution between January 2023 and January 2025 were retrospectively reviewed. Based on post-procedural treatment response, patients were categorized into a good response group (n = 231) and a poor response group (n = 78). Receiver operating characteristic (ROC) curve analysis was applied to assess the predictive performance of CAR. Multivariable logistic regression analysis was used to determine independent predictors of poor treatment response, and a predictive model was subsequently developed.

Results: Seventy-eight patients (25.2%) developed poor treatment response. Preoperative CAR was higher in the poor response group than in the good response group [5.82 (3.65–8.96) vs 2.68 (1.35–4.52), p < 0.001]. ROC curve analysis showed that the area under the curve (AUC) of CAR for predicting poor treatment response was 0.812 (95% confidence interval (CI): 0.758–0.866), which was higher than that of C-reactive protein (CRP) (AUC = 0.756) and albumin (AUC = 0.718) (p < 0.05); the optimal cutoff value was 3.85, with a sensitivity of 74.4% and specificity of 76.2%. Multivariable logistic regression analysis showed that CAR ≥3.85 (odds ratio (OR) = 3.42, 95% CI: 1.89–6.18), TG18 Grade III (OR = 2.28, 95% CI: 1.26–4.12), time from onset to PTGBD ≥72 h (OR = 2.15, 95% CI: 1.18–3.92), diabetes mellitus (OR = 1.86, 95% CI: 1.04–3.33), and CCI ≥4 (OR = 1.92, 95% CI: 1.08–3.41) were independent risk factors for poor treatment response (all p < 0.05). The high CAR group (CAR ≥3.85) had longer time to temperature normalization, abdominal pain relief, and WBC normalization compared to the low CAR group (all p < 0.001).

Conclusion: Preoperative CAR shows good performance in identifying high-risk acute cholecystitis patients who are more likely to experience an unfavorable response to PTGBD. A CAR value ≥3.85 is associated with an increased likelihood of poor treatment response and may assist clinicians in early risk stratification and treatment planning.