Abstract
Introduction. Elderly patients are at high risk of mortality after elective surgery. The aim of this study was to obtain and evaluate the performance of a mortality prediction model for elderly patients undergoing elective surgery.
Methods. The design of this study was a retrospective cohort using medical record data from 747 elderly subjects who underwent elective surgery in the period January 2015–December 2017 at Cipto Mangunkusumo Hospital (RSCM), Jakarta. This study used multivariate analysis with logistic regression to determine significant predictors that were included in the prediction model. The performance of the prediction model was assessed using the Hosmer-Lameshow test, and its discrimination ability was determined by calculating the Area Under the Curve (AUC).
Results. Of the total 747 elderly subjects, the elective surgery mortality rate was 14.5%. The predictor variables were functional status [ADL 9-11, OR 1.808 (95% CI 0.848-3.854); ADL 0-8, OR 3.382 (95% CI 1.724-6.634)], comorbidities [CCI 3-4, OR 12.206 (95% CI 5.317-28.018); CCI >5, OR 15.820 (95% CI 6.701-37.347)], albumin level <3 g/dL [OR 3.777 (95% CI 2.105-6.779)], type of surgery [grade II, OR 3.827 (95% CI 1.849- 7.923); grade III, OR 6.560 (95% CI 3.378-12.739)], and ASA status with an ASA score > 3 [OR 5.106 (95% CI 1.841-14.159)] were further included in the components of the surgical mortality predictor scoring system. The mortality prediction model was categorized into low risk (score < 7; probability of mortality 2.33%), medium risk (score 7-10; probability of mortality 25.22%), and high risk (score > 10; probability of mortality 74.67%). The prediction model showed good discrimination [AUC score 0.900 (95% CI: 0.873-0.927)] and good calibration (p=0.718 on Hosmer-Lameshow test).
Conclusion. The prediction model of mortality among elderly patients undergoing elective surgery, incorporating factors like functional capacity, comorbidities, preoperative serum albumin concentrations, surgical procedure type, and ASA classification, showed good performance.
Bahasa Abstract
Pendahuluan. Pasien usia lanjut berisiko tinggi mengalami mortalitas pasca operasi elektif. Tujuan penelitian ini adalah untuk mendapatkan dan mengevaluasi performa model prediksi mortalitas pasien usia lanjut yang menjalani pembedahan elektif.
Metode. Desain penelitian ini adalah kohort retrospektif dengan menggunakan data rekam medis dari 747 subjek usia lanjut yang menjalani pembedahan elektif periode Januari 2015–Desember 2017 di Rumah Sakit Cipto Mangunkusumo (RSCM), Jakarta. Penelitian ini menggunakan analisis multivariat dengan regresi logistik untuk menentukan prediktor signifikan yang masuk ke dalam model prediksi. Performa model prediksi dinilai menggunakan uji Hosmer-Lameshow dan kemampuan diskriminasinya ditentukan dengan menghitung Area Under Curved (AUC).
Hasil. Dari keseluruhan 747 subjek usia lanjut, didapatkan angka mortalitas pembedahan elektif sebesar 14,5%. Variabel faktor prediktor berupa status fungsional [ADL 9-11, OR 1,808 (IK 95% 0,848-3,854); ADL 0-8, OR 3,382 (IK 95% 1,724-6,634)], komorbiditas [CCI 3-4, OR 12,206 (IK 95% 5,317-28,018); CCI >5, OR 15,820 (IK 95% 6,701-37,347)], kadar albumin <3 g/dL [OR 3,777 (IK 95% 2,105-6,779)], jenis pembedahan [grade II, OR 3,827 (IK 95% 1,849-7,923); grade III, OR 6,560 (IK 95% 3,378-12,739)], dan status American Society of Anesthesiologists (ASA) dengan skor ASA > 3 [OR 5,106 (IK 95% 1,841-14,159)] selanjutnya disertakan dalam komponen sistem skor prediktor mortalitas pembedahan. Model prediksi mortalitas dikategorisasikan menjadi risiko rendah (skor < 7; probabilitas mortalitas 2,33%), risiko sedang (skor 7-10; probabilitas mortalitas 25,22%), dan risiko tinggi (skor > 10; probabilitas mortalitas 74,67%). Model prediksi menunjukkan diskriminasi yang baik [skor AUC 0,900 (IK 95%: 0,873-0,927)] dan kalibrasi yang baik (p=0,718 pada uji Hosmer-Lameshow).
Kesimpulan. Model prediksi mortalitas pasien usia lanjut yang menjalani pembedahan elektif dengan prediktor status fungsional, komorbiditas, kadar albumin serum pre operatif, jenis pembedahan, dan kategori ASA memiliki performa yang baik.
Kata Kunci: model prediksi, mortalitas, pasien usia lanjut, pembedahan elektif
Recommended Citation
Djafar, Fitria; Dwimartutie, Noto; Chandra, Susilo; and Harimurti, Kuntjoro
(2024)
"Model Prediksi Mortalitas Pembedahan Pasien Usia Lanjut yang Menjalani Pembedahan Elektif di Rumah Sakit Cipto Mangunkusumo,"
Jurnal Penyakit Dalam Indonesia: Vol. 11:
Iss.
2, Article 4.
DOI: 10.7454/jpdi.v11i2.1567
Available at:
https://scholarhub.ui.ac.id/jpdi/vol11/iss2/4
