A Predictive Regression Model of Health Care Expenditures in Privately Insured Individuals with Alzheimer's Disease
Open Access
- Author:
- Peterson, Emily Nancy
- Graduate Program:
- Public Health Sciences
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- March 21, 2013
- Committee Members:
- Douglas L Leslie, Thesis Advisor/Co-Advisor
- Keywords:
- healthcare costs
Alzheimer's disease
cost of Alzheimer's
non-normal cost data regression
public health cost data - Abstract:
- Abstract Background: Alzheimer’s disease (AD) is a neurodegenerative disorder incurring significant social and economic costs due to increasing prevalence rates and associated higher costs of care. Average annual costs of caring for patients with AD have been estimated at $80 billion to $100 billion in the United States. Previous literature has identified several important factors that influence costs of care including symptoms, severity, age, and the presence of comorbid conditions. However, much of the AD costs literature measures total costs of AD in cross-sectional designs, and therefore cannot examine changes in costs over time. This study evaluates total costs of AD over time, as well as specific components of total costs (inpatient costs, outpatient costs and prescription drug costs). This study also evaluates differences in costs of care for patients with AD, with and without coexisting conditions. In this study, the aim is to extend upon previous research of US healthcare expenditures for AD, and estimate the effects of coexisting conditions on the direct healthcare costs of AD (total costs, inpatient costs, outpatient costs, and prescription drug costs). Objectives: This study aims to develop a predictive model for healthcare expenditures in the United States for patients diagnosed with AD. It investigates the magnitude of health care expenditures among privately insured individuals with AD. It will distinguish patterns of high expenditures within each type of care (outpatient, inpatient and prescription drug) and identify characteristics of patients associated with higher expenditures. Lastly, it evaluates the differences in costs of care between patients diagnosed with AD alone, and patients that have comorbid conditions. Methods: This retrospective study used 2008-2010 Truven Health MarketScan® Commercial Claims and Encounters data to examine total and component healthcare costs per patient by year. Random effects models estimate the effects of patients’ characteristics (age, gender, region, and presence of significantly associated comorbidities) as well as year (2008-2010) on direct costs of care. Direct costs included outpatient costs, inpatient costs, prescription drug costs, and total costs of care. Results: Total direct costs increased 19.4% from approximately $1,871 per patient in 2008 to $2,233 in 2010. This substantial increase is attributed to increases in patient age and increases in costs across time. Annual outpatient costs increased 12.6% from $662 per patient in 2008 to $746 in 2010 among AD patients who used outpatient services. Average annual inpatient costs increased 2.9% from $7,790 in 2008 to $8,016 in 2010 among AD patients who used inpatient services. Prescription drug costs increased 28.8% from $1,735per patient per year in 2008 to $2,236 in 2010 for AD patients who used AD medications. The number of patients using inpatient services, outpatient services and prescription drugs for AD also increased from 2008 to 2010. Regression analyses showed the following indicators to have positive significant effects on total direct costs for individuals with AD: year, age, and region of residence (p-value<0.05). The regression analyses also showed that two out of the three comorbid conditions included in the study, chronic obstructive pulmonary disorder and chronic liver disorder, had significant negative effects on total direct costs, p-value<0.001. Total direct costs, outpatient, inpatient, and prescription drug costs were consistently lower for patients with AD who had comorbidities compared to patients diagnosed with AD alone. Conclusion: Total direct costs of caring for individuals with AD increased substantially over time. Increases in costs were attributed to increases over time, age, number of years from diagnosis, and region of residence. Gender was not significantly associated with total direct costs. Presence of chronic obstructive pulmonary disorder or chronic liver disease resulted in significant decreases in total direct costs, which could be attributed to increases in costs directed towards comorbid conditions instead of attributing costs to initial diagnosis of AD. Findings from this study will help healthcare providers and policy makers to better manage future costs of patients with AD. The implications from this study will show health policy makers that within the population of AD patients, there are specific characteristics that identify higher costs expenditures. It will also demonstrate the differences in costs of care specific to the type of care, and how to better prepare financially for patients with AD according to the types of services used. Lastly, this study shows a pattern of recording healthcare costs that may in fact deemphasize total direct costs of Alzheimer’s disease when comorbidities are present.