Abstract:

Background: The global burden of critical illness is difficult to be quantified but it is estimated to be much higher in lower middle-income countries. Intensive care unit is a rapidly growing specialty in sub-Sahara Africa but there is a scarcity of epidemiological data. Severity Scoring System are commonly used in developed countries adult intensive care units for assessment of disease severity and predictions of mortality however there scoring systems are not routinely used in lower middle income countries. This makes difficulty in assessing severity of disease and predicting mortality risk. Sub-Saharan Africa developed severity scores have not been validated. This study was done to assess the utility of the severity scoring systems developed in Sub-Saharan Africa and APACHEII is assessment of disease severity and mortality risk.

Objectives: To assess the utility of Sub-Saharan Africa severity systems and APACHEII score in assessment of disease severity and prediction of mortality in AICU at Bugando Medical Centre.

Method: This prospective observational study was conducted in Adult intensive care unit at Bugando Medical Centre. A total of 173 adult patients were enrolled in this study during the study period. Patient from all disciplines were involved in the study. Demographic data and blood samples for arterial gas analysis, full blood counts and HIV tests were obtained from all the patients. Arterial blood gas analysis was done by using Abbot held hand ASTAT machine, full blood count was processed at Bugando Medical Centre main laboratory using system XS-500i machine and HIV test by bioline/unigold tests. Data were analysed using STATA version 13.

Results: A total of 173 study participants were enrolled. Their median [IQR] age was 40 [30-56] years majority being males about 60.1%. In hospital mortality of 39.9% was observed in which 34.9% death occurred during AICU stay while 5.0% occurred in the wards after being discharged from the AICU. Health insurance status was observed to be an independent predictor of mortality in which patients without health insurance had higher mortality (OR 3.1; 95%CI 1.2 – 9.9; p-value=0.024). HIV status was observed to be also an independent predictor of mortality with (OR 5.1; 95%CI 1.1 – 23.4; p-value=0.036), and patients who were admitted with lower GCS had an increased mortality rate (OR 9.0; 95% CI 2.4 – 34.4; p-value=0.001). Severity scoring from Sub Saharan Africa i.e. clinical sickness and Deranged physiological parameter scoring systems predicted well the mortality among our AICU population with an area under the receiver operating characteristic (AUROC) of 0.795 and 0.753 respectively. Generally all severity scoring system predicted well the outcome whoever APACHEII score was observed to be superior in predicting the mortality outcome among other Sub Saharan African scores with an AUROC of 0.849.

Conclusion and Recommendation: Sub Saharan Africa severity scores i.e. Sickness illness and Deranged physiologic parameter scores can be used to predict AICU mortality as compared to APACHE II at our AICU. Health insurance should be encouraged among citizen since lack of it is associated with increased AICU mortality risk. HIV patients should be considered to have high AICU mortality risk more care should be done.