Clinical outcomes of patients with chronic liver disease admitted to selected tertiary care hospitals, Ethiopia: Prospective cohort study

Abstract

Background: Chronic liver disease (CLD) is a series of liver derangement persisting for at least 6 months. The cause of CLD can be viral and/or non-viral. In Ethiopia, despite the high burden report; data on different aspects of CLD is limited. Objective: To assess clinical outcomes and associated factors among chronic liver disease patients admitted to medical wards of selected hospitals. Methods: A prospective cohort study was conducted on 109 adult CLD patients recruited using purposive sampling technique from April 1, 2018 –October 5, 2018. Data was collected using tool comprised of pertinent parameters, entered into Epi-Data 4.2.0.0 for cleaning and analyzed using STATA 13.0. Result of descriptive analysis was presented using text, tables and figures. Kaplan-Maier and cox-regression analysis was used to compare the survival experience and to determine explanatory variables, respectively. Hazard ratios with a p-value <0.05 was considered to declare statistical significance. Result: A total of 109 CLD patients (77.98% male) were included. Mean (±SD) age of the participants was 39.03 ± 13.80. From the total of 109 CLD patients, 52 (47.71%) were with viral etiology. The overall median length of hospital stay was 7 (4 -11) days. Seventeen (15.60%) patients developed major acute in-hospital complications of CLD; 9(8.26%) were from viral group. The incidence rate (IR) of in-hospital acute complications of CLD in the viral group was insignificantly lower (crude IRR=0.911 [95% CI, 0.311-2.714, p= 0.424]). Duration since diagnosis (AHR=1.029 [95%CI, 1.004-1.054, p=0.025]) and aspartate amino transaminase (AST) level (AHR=1.007 [95%CI, 1.003 -1.010, p<0.001]) were the identified predictors for in-hospital acute complication of CLD. The cumulative mortality from admission to 30 days of hospital discharge was 38 (34.86%); 18 (16.51%) were from patients with viral etiology. Of these, 31 (28.4%) deaths were in-hospital; 13 (11.93%) were from viral group. IR of in-hospital mortality was insignificantly lower in the viral group (crude IRR= 0.635 [95% CI, 0.286 - 1.372, p=0.108]). Furthermore, a higher median survival time [29 days (13-29 days)] was identified for the viral group (log rank, p=0.04). Mean corpuscular volume (MCV) was the independent predictor of in-hospital mortality (AHR=1.004 [95%CI, 1.001 - 1.007, p=0.013]). Conclusion: HBV was the commonest etiology identified in this study. Incidence of acute inhospital complications of CLD and death were insignificantly lower among patients with viral etiology. Approximately, one death was observed for four admissions with CLD and the median survival time was significantly higher for the viral group. Prolonged duration since diagnosis and the increment in AST level were found to increase the rate for incidence of acute in-hospital complications of CLD. Since these factors are associated with progression of CLD, availing immunoprophylaxis and targeted treatments might benefit CLD patients. Furthermore, increment in MCV level at admission was identified to increase the risk of inhospital mortality. As a result, attention is required in the early detection and correction of hematologic abnormalities. Finally, because of the small and unequal sample size used, the difference in clinical outcomes among patients with or without viral etiology cannot be concluded confidently; a further study with adequate sample size is recommended.