A new method for predicting SIRS after percutaneous transhepatic gallbladder drainage

Area of Science:

• Hepatobiliary Surgery
• Clinical Prediction Modeling
• Inflammatory Response Syndromes

Background:

• Systemic inflammatory response syndrome (SIRS) is a significant risk following percutaneous transhepatic gallbladder drainage (PTGBD).
• Early prediction of SIRS is crucial for timely clinical intervention and prevention of severe complications.

Purpose of the Study:

• To develop and validate a predictive model for SIRS occurrence post-PTGBD.
• To identify key clinical features associated with SIRS development.

Main Methods:

• Logistic regression was employed to construct a predictive model using easily obtainable clinical features.
• Model performance was assessed using receiver operating characteristic (ROC) curves and decision curves on training and validation sets.
• Nine clinical features were selected: C-reactive protein (CRP), Fever, Direct Bilirubin (DBIL), Obstruction, Bile properties, Procalcitonin (PCT), Length, Width, and Volume factor.

Main Results:

• The developed prediction model demonstrated strong performance with an Area Under the Curve (AUC) of 0.83 in the training set and 0.81 in the validation set.
• Decision curve analysis indicated that the nine-feature model outperformed models based on single clinical features.
• The model accurately predicted SIRS occurrence, highlighting its clinical utility.

Conclusions:

• A robust prediction model incorporating nine clinical features has been developed for SIRS post-PTGBD.
• This model offers a valuable tool for clinicians to anticipate SIRS, enabling proactive management and potentially improving patient outcomes.
• Early identification and intervention guided by this model can mitigate the progression to severe complications.