Latar belakang: Model prediksi risiko operasi memiliki peranan penting pada tindakan operasi katup jantung. Perubahan karakter pasien dan fasilitas pembedahan dalam waktu tertentu dapat mempengaruhi nilai prediksi skor risiko operasi.Tujuan: Mengetahui perbandingan validasi EuroSCORE II, skor Ambler dan skor Harapan Kita dalam memprediksi mortalitas di rumah sakit pasca operasi katup jantung.Metode: Dilakukan studi kohort retrospektif terhadap 416 pasien yang menjalani operasi katup jantung pada periode November 2018 hingga Desember 2019. Data berasal dari rekam medis dengan metode sampling konsekutif. Didapatkan nilai kalibrasi dan diskriminasi EuroSCORE II, skor Ambler dan skor Harapan Kita.Hasil: Angka kematian yang diobservasi sebesar 6,7%. EuroSCORE II, skor Ambler and skor Harapan Kita memiliki kalibrasi yang baik (uji Hosmer-Lemeshow p=0,065, p=0,233 and p=0,314). Kemampuan diskriminasi skor dalam memprediksi kematian di rumah sakit EuroSCORE II (AUC 0,763; 95% IK;0.660-0.867), diikuti skor Ambler (AUC 0.748; 95% IK; 0.655-0.841) dan skor Harapan Kita (AUC 0,694; 95% IK; 0.584-0.804)Kesimpulan: EuroSCORE II, skor Ambler dan skor Harapan Kita memiliki validasi yang cukup baik. Kalibrasi ketiga skor baik dengan kalibrasi skor Harapan Kita relatif lebih baik dari dua skor lainnya, sedangkan nilai diskriminasi skor Harapan Kita di bawah EuroSCORE II dan skor Ambler.

---

Background: Preoperative risk prediction models have important role in cardiac valve surgical management. Changing in patient characteristics and surgical facilities over time, might affect the predicting value of those scoring system.

Objective: This study aimed to compare the validation of EuroSCORE II, Ambler score and Harapan Kita score in predicting in-hospital mortality at patients underwent heart valve surgery

Methods: Cohort restrospective study was performed at 416 patients who underwent heart valve surgery from November 2018 to December 2019. Data was taken from the medical records by consecutive sampling method. The calibration and discrimination value of EuroSCORE II, Ambler score and Harapan Kita score were obtained.

Results: Observed in-hospital mortality was 6,7%. EuroSCORE II, Ambler score and Harapan Kita score have good calibration (Hosmer-Lemeshow test p=0,065, p=0,233 and p=0,314). The discriminative value of these three scores in predicting in-hospital mortality for EuroScore II AUC 0,763 (95% CI; 0.660-0.867), Ambler score AUC 0.748 (95% CI; 0.655-0.841) and Harapan Kita score AUC 0,694 (95% CI; 0.584-0.804)

Conclusion: EuroSCORE II, Ambler score and Harapan Kita score have fairly good validation. Those scoring system have good calibration with Harapan Kita score calibration relatively better than EuroSCORE and Ambler score, meanwhile Harapan Kita score has less discrimination value than EuroScore II and Ambler score.