Reasons for admission in intensive care unit (ICU) and factors associated with poor outcome in referral public hospitals in Rwanda

Abstract

Background: The quality of care in ICU of hospitals from low-income countries is continuously improving, but mortality remains very high thus, current study aimed at investigating reasons of admission, outcome, predictors of ICU mortality and performance for mortality prediction of MEWS/MEO WS and that of qSOFA. Methods: This study was retrospective cross ±sectional with both descriptive and analytical aspect. It was conducted at CHUK and CHUB and 443 adult and pediatric patients managed in intensive care unit from January 2019 to December 2020 were recruited in this study .Multivariate analysis was conducted to determine predictors of ICU mortality and adjusted odd ratio with their 95% CI were calculated and ROC was used to assess diagnostic accuracy of MEWS/MEOWS and of qSOFA for mortality prediction. MEWS/MEOWS and qSOFA were recorded at admission time. Results: The commonest reason for admission was shock at 49% followed by decreased level of consciousness and Respiratory distress at 20.8% and 14% respectively. The majority of patients (33%) were admitted post- surgery .The mortality rate was 47% and majority of patients in shock died (58%). Age (AOR: 1.01, 95% CI: 1.00-1.02), lack of formal-education (AOR: 2.22, 95% CI: 1.05-4.65), invasive mechanical ventilation (AOR: 5.48, 95% CI: 2.40-12.46) and 0(:6 0(2:6 VFRUHV $25 &, -32.80) were factors linked with ICU poor outcome. The adjusted AUROC of the model of prediction of mortality by qSOFA and MEWS/MEOWS shows that both MEWS/MEOWS and qSOFA have fair accuracy in predicting the mortality in ICU patients with adjusted AUROC=0.6808. Conclusion: Surgical conditions are increasing and patients are being admitted in shock and with altered level of consciousness and it explains why mortality rate in ICU is persistently high. Age, education level, invasive mechanical ventilation and MEWS/MEOWS are predictors of mortality in ICU and when it came to mortality prediction tool in ICU, both MEWS/MEOWS and qSOFA tools have fair accuracy in predicting ICU mortality