Background
In intensive care units, scoring systems allow the assessment of disease severity and provide an estimate of in-hospital mortality. The use of scoring systems in critical care units has reduced many problems, especially for elderly patients.
Objective
The objective of this study was to evaluate APACHE II, APACHE IV and SOFA as predictors of outcomes in critically ill elderly patients in the geriatric critical care unit (CCU).
Patients and methods
A prospective observational study was carried out in the geriatrics and gerontology department's CCU at Ain Shams University Hospitals. The study included 106 elderly patients from both sexes aged 60 years old and older who were admitted to the Geriatrics CCU between March 2023 and August 2023. APACHE II and APACHE IV scores were calculated on admission. SOFA score was recorded on admission and every 48 hours until discharge. All enrolled patients were followed up, and outcomes were recorded as survivors and nonsurvivors. Observed mortality rates were compared with predicted mortality rates for the APACHE II, APACHE IV, and SOFA.
Results: At the end of the study period, 54 (50.95%) patients were survivors and 52 (49.05%) were non-survivors. There was a highly significant increase in all scores in non-survivor patients. Also, they showed a good ability to predict mortality rates, except for SOFA Initial.
Conclusion: Discrimination and calibration were better for all studied score models. However, SOFA Highest had the best calibration and discrimination. SOFA delta showed the highest specificity, and APACHE II showed the highest sensitivity.