Comparing efficacy of different scoring models to predict hepatic encephalopathy after TIPS in cirrhotic patients.

BACKGROUND

Assessing hepatic encephalopathy (HE) risk post-transjugular intrahepatic portosystemic shunt (TIPS) in cirrhotic patients is crucial. This study compares the predictive performance of Child-Pugh and Model for End-Stage Liver Disease (MELD), CLIFC-AD and Freiburg index of post-TIPS survival (FIPS) scores for overt and severe HE. To compare the predictive value of Child-Pugh, MELD, CLIFC-AD and FIPS scores for overt and severe HE post-TIPS in cirrhotic patients.

MATERIALS AND METHODS

We retrospectively analysed data from 406 cirrhotic TIPS patients (January 2017-January 2021). Scoring models were assessed for differentiation (C-index), calibration, clinical utility and overall performance at 1, 3, 6 and 12 months post-TIPS.

RESULTS

Predictive performance for overt HE post-TIPS was low across models. FIPS had superior predictive ability for severe HE at 1 and 12 months post-TIPS (C-index: 0.781, 0.705). FIPS and CLIFC-AD showed good predictive capacity for severe HE in sarcopenic patients at 1 and 12 months (FIPS: C-index 0.863, 0.757; CLIFC-AD: C-index 0.748, 0.732). FIPS had the highest hazard ratio for severe HE (HR = 3.520, 95% CI: 2.134-5.807) and CLIFC-AD for overt HE (HR = 2.132, 95% CI: 1.581-2.874).

CONCLUSION

FIPS and CLIFC-AD scores demonstrate significant predictive ability for severe HE post-TIPS, particularly in sarcopenic patients.