Impact of Multimorbidity Patterns on Elderly Hospitalization and Mortality: A Case Study of Kwadaso S.D.A Hospital

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
In this thesis, we aimed to identify multimorbidity patterns, study their impact on the length of hospitalization (LOH) until death and on mortality among elderly patients. The study utilized a sample of 984 elderly general clinic patients aged 50 years or older from data gathered from Kwadaso S.D.A hospital. The multimorbidity patterns were identified by exploratory tetrachoric factor analysis. Patients were assigned to any pattern if they had at least two diseases with factor loadings of 0.50 or more in absolute value on the corresponding pattern. However, to study the impact of the multimorbidity patterns, accelerated failure time (AFT) models with proportional hazards (PH) were used. The hazards were that of the exponential and Weibull models. Three multimorbidity patterns were identified: cardio-metabolic and pain disorders (CMPD), cardio-pulmonary disorders (CPD) and gastrointestinal, low back pain and anxiety disorders (GLAD). These patterns affected 52.2% of the entire sample. The Weibull model (AIC=239.7) provided a better fit when compared to the exponential model (AIC=259.3). Results from the Weibull model revealed that the median LOH until death was decreased by a factor of 0.46 and a factor of 0.45 for patients with CMPD and patients with CPD respectively when compared to patients with GLAD. Therefore, the estimated median LOH until death were 16 days for patients with CMPD and patients with CPD and 35 days for those with GLAD. Hospitalized elderly patients with these multimorbidity patterns, especially those with CPD and CMPD were vulnerable to increasing likelihood of mortality.
A thesis submitted to the Department of Mathematics, Kwame Nkrumah University of Science and Technology in partial fulfilment for the degree of Master of Philosophy, May-2013