Latar Belakang: Sepsis merupakan salah satu masalah kesehatan di rumah sakittermasuk di ruang Intensive Care Unit (ICU) dan angka kematiannya masih tetaptinggi meskipun dengan tatalaksana yang maksimal dan biaya yang besar. Kematianmerupakan hal yang sulit untuk diprediksi. Pasien yang telah diresusitasi denganbaik masih berpeluang untuk mengalami kematian karena proses disfungsi organyang terus berlanjut akibat tingginya tingkat inflamasi. Inflamasi yang tidakterkontrol memicu stress oksidasi dan necroptosis. Penelitian terakhir menunjukkankadar protein carbonyl (PCO) dan receptor-interacting protein kinase 3 (RIPK3)tinggi pada pasien sepsis dan dapat digunakan untuk memprediksi kematian.Penelitian ini bertujuan untuk menilai seberapa besar kegagalan resusitasi, kadarPCO, dan kadar RIPK3 dapat memprediksi kematian pada pasien sepsis.Metode: Rancangan penelitian ini adalah kohort prospektif di ruang resusitasi danICU RSUP. Dr. Moh. Hoesin (RSMH) Palembang. Penelitian dimulai setelahsertifikat etik dan izin lokasi diterbitkan sejak bulan Februari sampai Agustus 2019.Kriteria penerimaan meliputi pasien berusia 18 tahun atau lebih yang didiagnosissepsis. Kriteria penolakan meliputi keluarga menolak diikutsertakan dalampenelitian, pasien tidak dirawat di ICU, terlambat didiagnosis (lebih dari 24 jam),hamil dan didiagnosis mati batang otak. Kriteria pengeluaran meliputi pasienmeninggal kurang dari 4 jam setelah diagnosis ditegakkan dan pasien tidak dapatdilakukan follow up dalam waktu 28 hari. Tim peneliti yang telah dilatihsebelumnya mengidentifikasi semua pasien yang memenuhi kriteria penelitian.Semua subjek penelitian mendapatkan resusitasi standar dan diambil sampel darahuntuk diperiksakan ke laboratorium. Pasien diamati selama 28 hari: apakahmengalami kematian atau tidak. Kegagalan resusitasi didefinisikan sebagai kadar laktat ≥2 mmol/l atau reduksi laktat <20%. Data yang didapatkan dianalisis denganuji statistik yang sesuai menggunakan piranti lunak program STATA.Hasil: Didapatkan total 72 subjek penelitian, 13 dikeluarkan karena meninggalkurang dari 4 jam setelah diagnosis ditegakkan. Dari hasil analisis bivariatdidapatkan hubungan antara kegagalan resusitasi (RR 1,36; IK95% 0,965-1,916; p0,085), kadar PCO (RR 2,37; IK95% 1,348-4,194; p 0,0001), dan kadar RIPK3 (RR5,86; IK95% 2,07-16,61; p <0,0001). Dari hasil multivariat hanya didapatkan satuvariabel yang bermakna yaitu kadar RIPK3 (RR 5,39; IK95% 1,490-19,478; p0,010). Setelah dikontrol dengan variabel perancu usia, komorbiditas dan skorAPACHE II didapatkan variabel RIPK3 memiliki RR 4,64 dengan IK95% 1,233-17,479; p 0,023).Simpulan: Kegagalan resusitasi, kadar PCO dan kadar RIPK3 dapat memprediksikematian pada pasien sepsis.

---

Background: Sepsis remains one of the health problems at the hospital including

intensive care unit (ICU) since its mortality is still high despite maximal efforts on

therapy. Mortality is an unpredictable event. Patients who were properly

resuscitated still have a probability of mortality because of severe inflammatory

state which may lead to ongoing organ dysfunctions. Uncontrolled inflammation

will trigger oxidative stress and necroptosis. Recent study showed that high level

of protein carbonyl (PCO) and receptor-interacting protein kinase 3 (RIPK3) in

septic patients could be used to predict mortality. This study wished to analyze the

ability resuscitation failure, PCO level and RIPK3 level to predict mortality in

septic patients.

Methods: This prospective cohort study was conducted at resuscitation room and

ICU of RSUP. dr. Moh. Hoesin (RSMH), a single tertiary teaching hospital in

Palembang, South Sumatera. This study was started after ethical and location

authorization were unleashed in February to August 2019. Inclusion criteria were

18 years old or above patients that were diagnosed with sepsis. Exclusion criteria

were patients whose family did not give any consent to participate the study,

patients that were not treated at the ICU, had a late diagnosis (>24 h), pregnant, and

diagnosed with brain dead. Drop out criteria including died <4 h after diagnosed

and patients that could not be followed in 28 days. Investigators were trained to

identified all eligible patients. Subjects had a standard resuscitation and their blood

was taken to be examined at the laboratory. Patients were observed in 28 days

whether there were any mortality or not. Failed resuscitation defined by examined

lactate level ≥ 2 mmol/l or lactate reduction<20%. Data was statistically analyzed

with STATA™.

Results: Seventy two subjects were included to the study but 13 of them were

dropped out because died within 4 h after diagnosed. From bivariate analysis, there

was an association between failed resuscitation (RR 1.36; CI95% 0.965-1.916; p

0.085), PCO level (RR 2.37; CI95% 1.348-4.194; p 0.0001), and RIPK3 level (RR

5,86; CI95% 2.07-16.61; p <0.0001). From multivariate analysis using cox

regression time constant, the only variable statistically significant was RIPK3 (RR

5.39; CI95% 1.490-19.478; p 0.010). After adjusted by confounding variables,

including age, comorbidities, and APACHE II score, RIPK3 had RR 4.64 with CI

95% 1.233-17.479; p 0.023.

Conclusions: Failed resuscitation, PCO level, and RIPK3 level can predict

mortality in sepsis patients