Understanding health care utilization of older adults with Parkinson’s disease in Taiwan - Role of psychological variables

Abstract

Aims: As a common neurodegenerative chronic illness in old age, Parkinson’s disease (PD) is a leading reason of long-term health care utilization. While the multifaceted economic concerns in this disorder are internationally well documented, the psychological aspects have not received the same attention. The present study, based on a modified Andersen model, aimed at improving the understanding of predictors addressing health care utilization among a sample of older Taiwanese adults with PD, specifically exploring the role of the psychological component and investigating the relationships between depressive symptoms, health care utilization and health-related quality of life (HRQOL). The specific goals of the present research project were to (1) get an insight into the characteristics of health care utilization in PD; (2) test a modified Andersen model with additional psychological variables such as knowledge, attitudes, the social norms and perceived control; (3) establish path models predicting differentiated types of health care utilization examined in this study; and to (4) explore the relationships between health care utilization, depressive symptoms and HRQOL among a sample of older Taiwanese adults with PD. Methods: This study was conducted in Taipei, Taiwan, in 2009, using a cross-sectional design and a questionnaire orally administered to 284 persons with PD aged 65 and over, who had been recruited via location sampling in four outpatient clinic settings. A final useable response rate of 70.4 per cent was achieved, which included a sample of 200 participants, consisting of 104 males (52%) and 96 females (48%). Statistical description and logistic regression analysis in SPSS Version 16 and path analysis in LISREL Version 8.72 were conducted for the data analysis. Results: Visits to hospital out-patient departments were found to be the most utilized health care service (approximately 70%), followed by prescription refill slips (52.5%), supportive devices (55.5%) and family care (48%). According to logistic regression analysis, the modified Andersen model showed a good fit with the data and accounted for between 23.6% and 53.4% of the variance (Negelkerke R2). Need was indicated to be the most significant risk factor. It is to be noted that risk factors concerning the utilization of single service were suggested to vary from service to service, depending on the service being measured. Moreover, based on an ‘aggregate’ approach (single services had been aggregated into four types of services), path analysis revealed significant effects of need, followed by the psychological and predisposing components. Addressing the utilization of medical and care services, the role of psychological variables turned out to be significant. In detail, ADL, attitudes towards health care and age were indicated to have total effects on the utilization of medical services; ADL, age, education level, selective and compensatory control were significantly linked to the utilization of care services. In regards with HRQOL among older adults with PD, the findings indicated that selective control and ADL had positive total effects on HRQOL, whereas co-morbidity, age and education level were negatively linked to HRQOL. Additionally, the utilization of care services was proved to have negative total effects on HRQOL. Moreover, once depressive symptoms were taken into consideration, depressive symptoms, age, co-morbidity, education level, ADL and selective control were suggested to be significantly linked to HRQOL. Particularly, high levels of depressive symptoms were more likely to be associated with an increasing utilization of medical and care services. Additionally, the utilization of care services was proved to have negative total effects on HRQOL. Indirect effects on HRQOL and a complex interplay inherent in the modified behavioral model were also identified. Conclusion: Next to need, the psychological variables in the modified Andersen model were proved to have total effects on health care utilization. Health care services were recommended to be categorized into different types with the aim of extending the understanding of multifaceted health care utilization and HRQOL in PD. Moreover, the impacts of selective control, depressive symptoms and the utilization of care services on HRQOL were identified. Accordingly, screening and treatment for depression and implementing behavioral intervention programs with the concept of perceived control were likely to improve HRQOL among the older adults with PD.