Efficacy of APACHE II score at ICU Discharge in predicting post-ICU mortality and determine ICU readmission at the Colonial War Memorial Hospital, Fiji.

Researcher: Atalifo M1
Supervisors: Christopher J2, Bennett E3, Nasedra L2
Researcher Affiliations:
1Registrar, Department of Anaesthesia, Intensive Care & Hyperbaric Service, Fiji
2Consultant, Department of Anesthesia, Intensive Care & Hyperbaric Service, Fiji
3Senior Lecturer, Department of Anesthesia, College of Medicine, Nursing and Health Sciences, Fiji National University, Suva, Fiji.
Keywords; Intensive Care Unit, APACHE II Score, Post-ICU Mortality, Readmission
Introduction: Despite successful discharge to the wards from ICU, some patients deteriorate and either require ICU readmission or result in death whilst in the ward (post-ICU mortality). Therefore, in this study, we assessed the efficacy of APACHE II score at discharge in predicting post-ICU mortality and determine ICU readmission for the same period of hospital admission.
Methodology: This is a prospective, non-interventional, observational study for a period of one year (May 2018 to April 2019) in an adult mixed ICU setting.
Results: Of the 244 patients that were successfully discharged from ICU during the observed period, 21 (8.6%) died post-ICU discharge (post-ICU mortality) and 19 (7.8%) patients deteriorated requiring ICU readmission. The ‘area under the receiver operating characteristic’ (AUROC) curve analyses for the APACHE IIA (admission) score and APACHE IID (discharge) score in predicting post-ICU mortality were 0.654 (95% CI 0.538 to 0.769) and 0.773 (95% CI 0.670 to 0.875) respectively (p =0.013). A cut-off APACHE IID was calculated to be 12.5 (sensitivity 0.905; specificity 0.546) for post-ICU mortality. Similarly, the AUROC curve analyses for the APACHE IIA and APACHE IID scores in predicting ICU readmission were 0.647 (95% CI 0.512 to 0.782) and 0.775 (95% CI 0.689 to 0.861) respectively (p =0.050). A cut-off APACHE IID score was calculated to be 12.5 (sensitivity 0.947; specificity 0.546) for ICU readmission. Multivariate analysis illustrated the ‘number of comorbidities’ (odds ratio (OR) 1.703, 95% CI 1.046 to 2.773, p=0.032) to be a significantly associated with post-ICU mortality. The APACHE IID score (OR 1.214, 95% CI 1.081 to 1.364, p=0.001) and the ‘number of comorbidities’ (OR 0.487, 95% CI 0.241 to 0.985, p=0.045) were significantly associated predictors for ICU readmission.
Conclusion: This study concludes that the APACHE II score at discharge is effective and superior to APACHE II score at admission in predicting post-ICU mortality and ICU readmission with a cut-off score of 12.5.