Background: Recent data have shown that the proportion of older adult patients admitted to intensive care units is increased and the severity of illness is an independent risk factor associated with mortality. The aim of the current study was to compare the prognostic value of the Glasgow Coma Scale (GCS) and GCS-Age Prognosis (GAP) scores in older adult patients (aged ≥65 years) admitted to Medical Intensive Care Unit (MICU).
Methods: This was a prospective study of 168 consecutive older adult patients admitted to medical ICU during a 14-month period. For each patient, the GCS and GAP score in the first 24hours of admission and demographic characteristics were calculated and recorded. For statistical analysis, the logistic regression, Receiver operator characteristic (ROC) curve, and Hosmer-Lemeshow test were used (95% confidence interval).
Results: Survivors had a significantly higher GCS and GAP scores in the first 24h of MICU admission compared with nonsurvivors (p<0.001, p<0.001, respectively). The discrimination power of both models was good ((area under curve [AUC]:83.8% (standard error [SE]:3%), AUC: 85.4% (SE: 2.9%), respectively). Based on the Hosmer-Lemeshow goodness of fit test, just GCS had an acceptable calibration (x2=13.18, p=0.068).
Conclusions: For older adult patients admitted to the MICU, GCS and GAP scores reliably predict outcomes. Based on AUCs the discrimination power of models was good, but the calibration was acceptable just for GCS, thus the GCS is the better predictive model than GAP and useful in determining the prognosis of older adult patients in MICU.