PERFORMA SIMPLIFIED ACUTE PHYSIOLOGY SCORE 3 SEBAGAI PREDIKTOR MORTALITAS PADA UNIT RAWAT INTENSIF KARDIOVASKULAR

Abstract

Background: Severity of illness scoring systems has gained increasing popularity in Intensive Care Units (ICUs) since 1980s. Physicians used them for predicting mortality and assessing illness severity in clinical trials. The Simplified Acute Physiology Score 3 (SAPS 3) is the only score that can predict hospital mortality within an hour of admission to ICU. Although this scoring systems has been widely used in ICUs, they have not been commonly applied in Intensive Cardiovascular Care Units (ICVCUs) since the population is quite different especially in disease subset. Therefore, the objective of this study was to evaluate the parameters in the SAPS 3 scoring system performance for predicting mortality in ICVCU population. Methods: This was an observational study with cross-sectional approach using secondary data from RAICOM (Registry of Acute and Intensive Cardiovascular Care on Outcome) taken from September 2013 September 2014 in the ICVCU National Cardiovascular Centre Harapan Kita, Jakarta, Indonesia. The secondary data were collected, analysed, and matched with SAPS 3 variables. All missing and invalid data were excluded. All data was processed and the SAPS 3 score was calculated in each patient. Multivariate analysis with logistic regression was conducted to evaluate the significance of the parameters in predicting mortality. Discrimination was assessed by area under the receiver operator characteristic curve (AUROC). Calibration was assessed by Hosmer-Lemeshow goodness-of-fit test through calculating the ratio of observed?to?expected numbers of deaths. Results: A total of 233 patients were included in this study and the observed hospital mortality was 16.7% (39/233). The patients enrolled were divided into survivors and nonsurvivors. Bivariate analyses of SAPS 3 variables showed intra-hospital location before ICVCU admission, use of vasoactive agents, reasons for ICVCU admission, infection, Glasgow Coma Score (GCS), creatinine level, and platelet count were significantly different between nonsurvivors than survivors (P<0.05). The SAPS 3 score was significantly higher in nonsurvivors than survivors. The AUC (95% confidence intervals [CIs]) for SAPS 3 score was 0.752 (0.6690.835). The Hosmer?Lemeshow goodness?of?fit test for SAPS 3 demonstrated a Chi?square test score of 1.729, P = 0.943. Multivariate logistic regression was conducted for all variables that were probably correlated to prognosis. Eventually, intermediate ward as intra-hospital location before ICVCU admission was selected as an independent risk factors for predicting mortality (OR 4.165; 95% CI 1.462-11.864; P=0.008), whereas surprisingly the presence of community-acquired pneumonia (CAP) before ICVCU admission was a protective factor from hospital mortality (OR 0.224; 95% CI 0.068-0.730; P=0.013). Conclusion: Parameters in the SAPS 3 score system exhibited satisfactory performance in discrimination. In predicting hospital mortality, these parameters also showed good calibration for estimating hospital mortality. Intermediate ward as intra-hospital location before ICVCU admission appeared to be independently associated with mortality whereas patients with CAP comorbid as a protective factor against mortality. Despite the good result of this study, there are still plenty room of improvement for developing similar score in the future specifically for ICVCU population.

Keywords

• Hosmer?Lemeshow goodness?of?fit Test
• Intensive Cardiovascular Care Unit
• Independent Risk Factor
• Probability of Hospital Mortality
• Prognosis
• Receiver Operating Characteristic