Pengaruh Luas Luka Bakar, Usia, Trauma Inhalasi dan Skor Qsofa terhadap Mortalitas Pasien

Dzaky Haidar Afif, Dr. dr. M Rosadi Seswandhana, Sp.B., Sp.B.P.R.E., Subsp. L.B.L. (K)

2025 | Tesis-Spesialis | S2 Ilmu Bedah

Latar Belakang: 

Cedera luka bakar masih menjadi beban kesehatan global yang signifikan, terutama di negara berpenghasilan rendah dan menengah yang memiliki keterbatasan akses terhadap unit perawatan intensif luka bakar khusus. Identifikasi dini pasien berisiko tinggi sangat penting untuk triage dan tata laksana yang tepat. Model prediksi mortalitas luka bakar konvensional umumnya mengandalkan parameter demografis dan karakteristik cedera, sehingga mungkin tidak mencerminkan perburukan fisiologis awal secara memadai. Penelitian ini bertujuan untuk mengidentifikasi prediktor independen mortalitas pada pasien luka bakar dan mengembangkan sistem skoring klinis sederhana yang mengintegrasikan keparahan luka bakar dan disfungsi organ dini.

Metode:

Studi retrospektif ini menganalisis data dari registri luka bakar prospektif di RSUP Dr. Sardjito, pusat rujukan tersier luka bakar di Indonesia, antara Januari 2020 hingga Desember 2024. Variabel klinis yang dicatat saat pasien tiba di unit gawat darurat meliputi usia, total body surface area (TBSA), cedera inhalasi, dan skor quick Sequential Organ Failure Assessment (qSOFA). Analisis regresi logistik multivariat dilakukan untuk mengidentifikasi prediktor independen mortalitas selama perawatan dan menyusun sistem skoring prognostik berbobot. Performa model dievaluasi menggunakan analisis kurva receiver operating characteristic (ROC).

Hasil:

Sebanyak 276 pasien luka bakar diikutsertakan dalam penelitian ini. Prediktor independen mortalitas yang teridentifikasi adalah usia ?60 tahun, TBSA 20–50%, TBSA >50%, cedera inhalasi, dan skor qSOFA ?1. Sistem skoring sederhana dengan rentang 0 hingga 7 poin kemudian dikembangkan. Model menunjukkan diskriminasi yang sangat baik dengan area under the ROC curve (AUC) sebesar 0,920 (95% CI 0,88–0,95). Pada nilai cut-off skor ?3, model mencapai sensitivitas 74,7?n spesifisitas 90,2%.

Kesimpulan:

Penelitian ini mengusulkan model prediksi mortalitas luka bakar yang sederhana dan mengintegrasikan keparahan cedera serta perburukan fisiologis dini menggunakan qSOFA. Sistem skoring ini menunjukkan performa prediktif yang sangat baik dan berpotensi menjadi alat praktis untuk stratifikasi risiko mortalitas dini pada pasien luka bakar di instalasi gawat darurat maupun di fasilitas dengan sumber daya terbatas.

Background:

Burn injuries continue to represent a significant global health burden, particularly in low- and middle-income countries where access to specialized burn intensive care units is limited. Early identification of high-risk patients is crucial for appropriate triage and management. Conventional burn mortality prediction models mainly rely on demographic and injury-related parameters and may not adequately reflect early physiological deterioration. This study aimed to identify independent predictors of mortality in burn patients and to develop a simplified clinical scoring system incorporating burn severity and early organ dysfunction.

Methods:

This retrospective study analyzed data from a prospective burn registry at Dr. Sardjito General Hospital, a tertiary referral burn center in Indonesia between January 2020 and December 2024. Clinical variables recorded at emergency department presentation included age, total body surface area (TBSA), inhalation injury, and quick Sequential Organ Failure Assessment (qSOFA) score. Multivariate logistic regression analysis was performed to identify independent predictors of in-hospital mortality and to construct a weighted prognostic scoring system. Model performance was evaluated using receiver operating characteristic (ROC) curve analysis.

Results:

A total of 276 burn patients were included. Independent predictors of mortality were age ?60 years, TBSA 20–50%, TBSA >50%, inhalation injury, and qSOFA score ?1. A simplified scoring system ranging from 0 to 7 points was derived. The model demonstrated excellent discrimination with an area under the ROC curve (AUC) of 0.920 (95% CI 0.88–0.95). At a cut-off score ?3, the model achieved a sensitivity of 74.7% and specificity of 90.2%.

Conclusion:

This study proposes a simplified burn mortality prediction model that integrates injury severity and early physiological deterioration using qSOFA. The scoring system demonstrated excellent predictive performance and may provide a practical tool for early mortality risk stratification of burn patients in emergency and resource-limited settings.

Kata Kunci : burn injuries, mortality prediction, TBSA, inhalation injury, qSOFA, bedside scoring system, BAIqS, logistic regression, ROC curve, resource-limited settings