Sequential APACHE II Scores for Prediction of Mortality in Patients with Severe Malaria and Admitted to Critical Care Facilities

Abstract

Background/Aims: Both generic and disease specific prognostic scoring systems have been employed in areas of critical care. The score of Acute Physiology and Chronic Health Evaluation (APACHE) II is a generic point score which provides general assessment of illness and severity. In order to determine the mortality of critically ill malaria patients, this study aims to find the ideal time point for the APACHE II score. Materials and Methods: This longitudinal study was carried out after getting necessary ethics committee approval. Score of APACHE II on days 0, 2nd and 7th were evaluated and appropriate statistical tests were applied. Results: Out of 120 patients, 54 patients were of P. vivax, P. falciparum - 60 and six mixed infection patients. Mean APACHE II score was maximum on day 0 followed by day 2 and 7 in decreasing order. The APACHE II score was statistically greater on all days in patients who didn’t survive. The Receiver Operating Characteristic (ROC) curve when plotted showed score of APACHE II on day 2 - cut-off ≥14 to be the most valid in mortality prediction with a sensitivity of 64.28% and specificity of 87.80% as most of the results were on the left from the diagonal line and had greatest area under the curve. Conclusion: The prognostic score APACHE II was discovered to be helpful in patients with severe falciparum/vivax malaria who required intensive care treatment since the second day's sequential score significantly differentiates between survivors and non-survivors. The optimum cut-off for identifying individuals at high risk of mortality is 14, and the Day-2 APACHE II score is an ideal biomarker for predicting the outcomes of ICU patients. Recommendation: Malaria is a major cause of tropical sepsis in India leading to significant amount of mortality. Sequential APACHE II scoring instead of single time APACHE II score calculated on admission could have a role in evaluation of effectiveness of treatment, trend in recovery or dysfunction and prediction of mortality.