Development and validation of an explainable rule-based clinical decision support system for childhood immunization using rule traceability

Childhood immunization is essential for preventing vaccine-preventable diseases, yet immunization decisions in primary healthcare are often affected by false contraindications, inconsistent guideline interpretation, and dynamic clinical conditions. This study developed and evaluated an Explainable Clinical Decision Support System (X-CDSS) that integrates rule-based reasoning, dynamic clinical variables, and explainable rule traceability to provide transparent, guideline-compliant immunization recommendations. A Research and Development (R&D) approach with iterative prototyping was employed. Knowledge from the Indonesian national immunization guideline, scientific literature, field observations, and healthcare professionals was formalized into 35 IF–THEN production rules implemented using a forward-chaining inference mechanism in a web-based application. The system was evaluated through functional testing, clinical validation using 40 representative scenarios, performance evaluation, and User Acceptance Testing involving seven healthcare professionals. The proposed system achieved an accuracy of 92.5%, precision of 96.6%, recall of 93.3%, an F1-score of 94.9%, and a Cohen's Kappa coefficient of 0.81, with an overall usability score of 4.3/5.0. These findings demonstrate that the proposed X-CDSS provides reliable, transparent, and clinically accountable decision support for routine childhood immunization in primary healthcare.