Clinical Scorecard: Development and Validation of a Machine Learning Model and Simplified Risk Score Based on SIRI for Predicting Postherpetic Neuralgia: A Multicenter Prospective Cohort Study

At a Glance

Key Highlights

• SIRI is an independent predictive factor for PHN development.
• The eXtreme Gradient Boosting (XGBoost) model showed optimal predictive performance.
• A simplified risk scoring table based on SHAP analysis achieved high clinical practicability.
• 511 out of 1361 patients (37.55%) developed PHN during follow-up.
• Early intervention is recommended for patients with a risk score ≥18.

Guideline-Based Recommendations

Diagnosis

• Use SIRI as a predictive biomarker for PHN.

Management

• Implement early active intervention for high-risk patients.

Monitoring & Follow-up

• Regularly assess SIRI and other predictive factors in patients with HZ.

Risks

• Patients with a risk score ≥18 are at increased risk for developing PHN.

Patient & Prescribing Data

Patients diagnosed with herpes zoster.

Focus on early identification and intervention for patients at high risk of PHN.

Clinical Best Practices

• Incorporate SIRI into routine assessments for patients with HZ.
• Utilize machine learning models for improved predictive accuracy.
• Adopt the simplified risk scoring tool for clinical decision-making.

Related Resources & Content

• Study on SIRI and PHN