Model Pelayanan Kesehatan BPJS Berbasis Random Forest Pada RSI Siti Khadijah

ABSTRACT Healthcare services provided to BPJS beneficiaries are expected to deliver high-quality, efficient, and patient-centered care. However, several challenges remain, including prolonged waiting times, complicated administrative procedures, and variations in patients' perceptions of the quality of services received. This study aimed to develop a predictive model of BPJS user satisfaction using the Random Forest algorithm and to identify service factors that should be prioritized for improvement at RSI Siti Khadijah. A quantitative approach was employed involving 204 respondents. Data were collected using a questionnaire consisting of 25 Likert-scale indicators covering learnability, efficiency, system reliability, ease of use, and user satisfaction. The findings demonstrated that the Random Forest model achieved excellent performance, with accuracy, balanced accuracy, precision, recall, F1-score, and ROC-AUC values of 1.000. The efficiency dimension obtained the lowest mean score (4.535), indicating the need for greater attention in service improvement. In addition, SKOR_EFF, EFF3, RL5, SKOR_RL, and RL4 were identified as the most influential factors affecting users' satisfaction with healthcare services. Keywords : BPJS, Healthcare Services, Random Forest, User Satisfaction. ABSTRAK Pelayanan kesehatan bagi peserta BPJS Kesehatan dituntut mampu memberikan layanan yang berkualitas, efisien, dan berorientasi pada kepuasan pasien. Namun, masih dijumpai berbagai permasalahan, seperti lamanya waktu tunggu, prosedur administrasi yang kurang sederhana, serta perbedaan persepsi pasien terhadap mutu pelayanan yang diterima. Penelitian ini bertujuan mengembangkan model prediksi kepuasan pengguna layanan BPJS menggunakan algoritma Random Forest serta mengidentifikasi faktor pelayanan yang menjadi prioritas perbaikan di RSI Siti Khadijah. Penelitian menggunakan pendekatan kuantitatif terhadap 204 responden dengan instrumen kuesioner yang terdiri atas 25 indikator skala Likert meliputi aspek kemudahan dipelajari, efisiensi, keandalan sistem, kemudahan penggunaan, dan kepuasan pengguna. Hasil penelitian menunjukkan bahwa model Random Forest memiliki kinerja sangat baik dengan nilai akurasi, balanced accuracy, precision, recall, F1-score, dan ROC-AUC sebesar 1,000. Aspek efisiensi memperoleh skor rata-rata terendah (4,535), sedangkan SKOR_EFF, EFF3, RL5, SKOR_RL, dan RL4 merupakan faktor yang paling berpengaruh terhadap kepuasan pengguna terhadap pelayanan Kesehatan. Kata Kunci: BPJS, Pelayanan Kesehatan, Random Forest, Kepuasan Pengguna.