Tesis

Uji Validitas Sistem Skoring Systemic Inflammatory Response Syndrome (SIRS), National Early Warning Score 2 (NEWS2), dan Rapid Emergency Medicine Score (REMS) dalam Memprediksi Luaran Buruk Pasien Sepsis di Unit Gawat Darurat RSUPN Dr. Cipto Mangunkusumo = Validity Assessment of Systemic Inflammatory Response Syndrome (SIRS), National Early Warning Score 2 (NEWS2), and Rapid Emergency Medicine Score (REMS) in Predicting Poor Outcomes in Sepsis Patients at the Emergency Department of Dr. Cipto Mangunkusumo National General Hospital.

Latar Belakang: Sepsis adalah masalah kesehatan global dengan mortalitas tinggi, terutama di negara berkembang seperti Indonesia. Deteksi dini pasien dengan risiko luaran buruk menjadi kunci penting dalam pengelolaan sepsis. Berbagai sistem skoring seperti Systemic Inflammatory Response Syndrome (SIRS), National Early Warning Score 2 (NEWS2), dan Rapid Emergency Medicine Score (REMS) dikembangkan untuk memprediksi luaran buruk, namun belum diketahui sistem mana yang paling akurat di populasi Indonesia. Metode: Penelitian ini merupakan studi kohort retrospektif menggunakan data rekam medis pasien sepsis dewasa di Unit Gawat Darurat RSUPN Dr. Cipto Mangunkusumo periode JanuariDesember 2023. Validasi performa skor SIRS, NEWS2, dan REMS dilakukan menggunakan analisis Area Under the Curve (AUC), sensitivitas, dan spesifisitas terhadap luaran buruk (mortalitas dalam 7 hari, menggunakan ventilasi mekanik, atau perawatan intensif) Hasil: Dari 347 pasien yang diikutsertakan dalam analisis, mayoritas pasien berusia < 60 tahun (68,88%), dengan jenis kelamin laki laki 46,97%. Infeksi saluran pernapasan menjadi sumber infeksi tersering (42,1%) dan proporsi komorbid keganasan 36,89%. Skor NEWS2 menunjukkan performa prediksi luaran buruk terbaik AUC = 0,951 (95% IK 0,92-0,97), diikuti oleh REMS AUC = 0,90 (95 % IK 0,87-0,93) dan SIRS AUC 0,673 (95% IK 0,61-0,72). Skor REMS menunjukkan performa kalibrasi terbaik dengan p= 0,99. Kesimpulan: Skor REMS memiliki performa diskriminasi dan kalibrasi terbaik dibandingkan dengan SIRS dan NEWS2 dalam memprediksi luaran buruk pasien sepsis di unit gawat darurat, sehingga dapat direkomendasikan sebagai alat skrining utama di unit gawat darurat rumah sakit Indonesia.
Kata Kunci: Sepsis, SIRS, NEWS2, REMS


Background: Sepsis is a significant global health issue associated with high mortality, particularly in developing countries like Indonesia. Early identification of patients at risk for poor outcomes is crucial in sepsis management. Scoring systems such as Systemic Inflammatory Response Syndrome (SIRS), National Early Warning Score 2 (NEWS2), and Rapid Emergency Medicine Score (REMS) have been developed to predict adverse outcomes, yet their comparative accuracy in the Indonesian population remains unclear. Methods: This retrospective cohort study analyzed medical records of adult sepsis patients from the emergency department at RSUPN Dr. Cipto Mangunkusumo from January to December 2023. Validation of SIRS, NEWS2, and REMS scoring systems was conducted using Area Under the Curve (AUC) analysis, sensitivity, and specificity in predicting poor outcomes (in-hospital 7 days mortality or intensive care support, or usage of mechanical ventilation during hospital stay) Results: Of the 347 patients included in the analysis, the majority were aged < 60 years (68.88%), with males accounting for 46.97%. Respiratory tract infection was the most common source of infection (42.1%), and the proportion of malignancy as a comorbidity was 36.89%. NEWS2 score demonstrated the best predictive performance for adverse outcomes AUC = 0.95 (95% CI 0.92- 0.97), followed by REMS AUC = 0.90 (95% CI 0.87-0.93) and SIRS AUC 0.673 (95% CI 0.61 - 0.72). The REMS score showed better calibration performance than NEWS2 on the HosmerLemeshow test, with p = 0.99. Conclusion: The REMS score demonstrated the best discrimination and calibration for predicting adverse outcomes in sepsis patients in the emergency department, outperforming NEWS2 and SIRS. Therefore, it can be recommended as the primary screening tool in Indonesian hospital emergency departments.
Keywords: Sepsis, SIRS, NEWS2, REMS