Tesis

Model Prediksi Mortalitas Pasien Usia Lanjut yang Menjalani Pembedahan Elektif di Rumah Sakit Cipto Mangunkusumo = Predictive Model for Mortality in Elderly Patients Undergoing Elective Surgery in Cipto Mangunkusumo Hospital.

Latar Belakang: Secara global, jumlah penduduk usia lanjut terus meningkat yang diiringi dengan jumlah pasien usia lanjut yang menjalani pembedahan juga meningkat. Pasien usia lanjut memerlukan perhatian khusus dalam persiapan, saat pembedahan dan pasca pembedahan karena kemunduran sistem fisiologis dan farmakologi sehingga lebih berisiko mengalami komplikasi. Tujuan: Mendapatkan angka mortalitas, model prediksi, serta performa model prediksi pasien usia lanjut yang menjalani pembedahan elektif di RSCM. Metode: Penelitian menggunakan desain kohort retrospektif dengan metode sampling konsekutif . Data sekunder rekam medis pasien usia lanjut yang menjalani pembedahan elektif di RSCM periode Januari 2015 – Desember 2017 dianalisis dengan program statistik SPSS Statistics 20.0 untuk analisis univariat, bivariat, multivariat, Receiving Characteristics Operator (ROC), dan analisis bootstrapping pada uji kalibrasi HosmerLemeshow. Hasil: Terdapat 747 subjek penelitian yang dianalisis untuk mendapatkan angka mortalitas dan prediktor yang bermakna untuk disertakan sebagai komponen sistem skor. Sebanyak 108 (14,5%) pasien meninggal pascabedah. Variabel status fungsional, komorbiditas, kadar albumin serum preoperatif, jenis pembedahan, dan status fisik ASA merupakan variabel yang secara statistik independen berhubungan dengan mortalitas. Sistem skor yang dibuat memiliki nilai AUC = 0,900 (KI 95% 0,873 – 0,927). Kalibrasi sistem skor baik dengan nilai p>0,05. Hasil ini konsisten setelah dilakukan bootstrapping. Kesimpulan : Angka mortalitas pasien geriatri yang menjalani pembedahan elektif adalah 14,5%. Prediktor dan komponen skor prediksi mortalitas pembedahan elektif pada pasien usia lanjut yaitu status fungsional, komorbiditas, kadar albumin serum preoperatif, jenis pembedahan, dan kategori ASA. Model prediksi memiliki kualitas kalibrasi dan diskriminasi yang baik dan kuat.
Kata Kunci: mortalitas, pasien usia lanjut, pembedahan elektif, model prediksi


Background: Globally, the number of elderly population continues to grow. It is accompanied by the increasing number of older people undergoing surgery. Elderly patients need certain care in preoperative, intraoperative,and postoperative phase since they are more likely to develop postoperative complication due to physiological and pharmacological deterioration. Aim: To get mortality rate, predictive model, and the performance of predictive model in elderly patients undergoing elective surgery in RSCM. Methods: This study is a retrospective cohort study with consecutive sampling method . Secondary data from patients’ medical record who underwent elective surgery from January 2015-December 2017 is analysed using SPSS Statistics 20.0 for univariate, bivariate, multivariate, and Receiving Operator Characteristics (ROC) and SPSS Statistics 20.0 for bootstrapping analysis in Hosmer- Lemeshow calibration test. Results: All 747 subjects are analysed to get mortality rate and predictor variables that are statiscally significant included as scoring system components. A hundred eight patients (14.5%) died within thirty days after surgery. Functional status, comorbidities, preoperative serum albumin level, type of surgery, and ASA physical status are independently associated with mortality. A scoring system composed of above predictors has an AUC value at 0.900 (95% CI 0.873 – 0.927). This scoring system shows good calibration with p>0,05 and this result is consistent even after bootstrapping analysis. Conclusion: The mortality rate of elderly patients undergoing elective surgery in RSCM is 14.5%. Scoring system for predicting mortality in elderly patients undergoing elective surgery consist of functional status, comorbidities, preoperative serum albumin levels, type of surgery and ASA physical status. The predictive model shows good calibration and strong discrimination.
Keywords: mortality, elderly patient, elective surgery, prediction model