Abstract
Background
The objective of this study was to ascertain co-occurring chronic conditions and expenditures associated with Parkinson’s disease among elderly individuals (age ≥ 65 years).
Methods
A retrospective, cross-sectional matched case–control design with data from Medical Expenditure Panel Survey (MEPS), a nationally representative survey of households in the United States was used. Elderly with Parkinson’s disease (N = 350) were compared to a matched control group (N = 1050) based on propensity scores. Ordinary Least Squares regressions on logged dollars were performed to understand the association between Parkinson’s disease and expenditures. All analyses accounted for the complex survey design of the MEPS and were conducted in SAS 9.3.
Results
Among elderly, the average total expenditures were $15,404 for those with Parkinson’s disease and $13,333 for those without Parkinson’s disease. Results from regressions revealed that elderly with Parkinson’s disease had 109% greater total expenditure compared to those without Parkinson’s disease, when only demographic and socioeconomic variables were entered in the model. When co-occurring chronic conditions were additionally included in the model, those with Parkinson’s disease had 84% greater expenditures compared to those without Parkinson’s disease.
Conclusions
Excess expenditures associated with Parkinson’s disease are partially driven by co-occurring conditions among individuals with Parkinson’s disease.
Keywords: Parkinson’s disease, Co-occurring conditions, Expenditure, Propensity score
1. Introduction
The prevalence of Parkinson’s disease (PD) has been found to be highly variable with worldwide standardized prevalence (across all ages) of 57–230 per 100,000 [1]. In the United States, an estimated 2% of elderly individuals over the age of 65 years have PD [2]. Existing literature suggests that compared to matched controls, individuals with PD experience higher comorbidity burden [3,4]. Using data from one county in Minnesota, it has been found that compared to gender and age matched controls without PD, individuals with PD had higher scores on Charlson Comorbidity Index (CCI) [3]. Using the same data, another study comparing the prevalence of mental health conditions found that the odds of anxiety and depression were approximately twice as high as among individuals with PD compared to their matched controls [4]. However, in a study of Medicare beneficiaries it was found that cardiovascular, renal, metabolic, and gastrointestinal conditions were not significantly different between beneficiaries with and without PD [5], suggesting that prevalence rates of some chronic conditions are comparable between the two groups.
Co-occurring conditions in individuals with PD can severely compromise their Health-Related Quality of Life (HRQoL) and can also lead to increased healthcare utilization and expenditures [6,7]. Although several studies have been conducted in European settings [8–10], only a handful of studies in the US have examined direct healthcare expenditures associated with PD [5,11–13]. Using commercial insurance claims, it has been estimated that average annual direct medical per-capita expenditures (in 2002 dollars) with PD ($23,101) was more than double compared to the individuals without PD ($11,247), with a projected annual direct healthcare expenditures of $23 billion for individuals with PD in the U.S. [11]. Among Medicare long-term care users, Medicare beneficiaries with PD had higher comorbidity cost ratios (ratio of average per person per year charges for PD alone vs. with comorbid conditions) compared to individuals without PD (adjusted to 2000 US dollars) [12]. Among nationally representative elderly Medicare beneficiaries, those with PD had approximately twice the total expenditures (expressed in 2002 dollars) compared to beneficiaries without PD [5].
Studies that examined the relationship between PD and direct healthcare expenditures have highlighted excess expenditures associated with PD. However, these studies have not examined the extent to which co-occurring conditions contribute to these excess expenditures. Therefore, the primary objective of this study is to examine incremental expenditures associated with PD and the role of co-occurring conditions in contributing to these excess expenditures. In addition, the present study builds on and improves existing research by including a nationally representative sample, recent data, use of a matched case–control design derived through rigorous statistical matching technique (i.e. propensity score matching).
2. Methods
2.1. Study design
A retrospective, cross-sectional matched case–control design was adopted for this study, matching individuals with PD to those without PD using a propensity score approach. A matched case–control design was adopted because individuals with PD may differ from individuals without PD in demographic characteristics, as well as physical and mental health status. To control for these systematic differences, we selected individuals with and without PD who were comparable on their demographic characteristics and health status.
Propensity score matching was performed using gender (male, female), race (white and other), perceived physical and mental health status (excellent/very good, good, fair/poor) and body mass index (BMI) [under weight or normal (BMI < 25 kg/m2) and overweight or obese (BMI ≥ 25 kg/m2)]. In addition, as our study pooled data from 10 years, two successive years were used to match, for e.g. 2000 and 2001, 2002 and 2003, 2004 and 2005, 2006 and 2007, 2008 and 2009 in order to avoid the matching of individuals with and without PD from distant years. Since the number of PD cases was low, for each individual with PD, 3 individuals without PD were selected in order to improve the power of the analysis [14].
2.2. Data source
We used data from annual releases of the Medical Expenditure Panel Survey (MEPS), a nationally representative survey of households and families. A wide array of information in terms of physical and mental health conditions, treatment, healthcare utilization and expenditures of non-institutionalized civilians in the United States are available in the MEPS. The household files from the MEPS provide information on the demographics, socio-economic status, employment, access to care, health status, and healthcare utilization and expenditures. Telephone interviews enquiring about the diagnosis (primary and other diagnoses) are conducted among a sample of physician offices and hospitals which rendered care to the MEPS sample population constituting the Medical Provider Component, which is often used to add and/or substitute the data on expenditures reported in the household component [15]. Medical conditions recorded in MEPS are self-reported. Studies conducted to demonstrate the accuracy of survey participants reported conditions have shown that household reports were consistent with provider reports for most of the salient conditions such as diabetes, mental health, and hypertension along with other conditions with a median sensitivity of approximately 70% [16].
2.3. Analytic sample
Our analytic sample consisted of elderly individuals aged 65 years or older and alive during the calendar years. We identified elderly individuals with PD using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes of 332.xx. In addition, we required that those without PD to have positive direct healthcare expenditures as all individuals with PD had positive direct healthcare expenditures. Matched controls were selected using the GREEDY 8 to 1 matching technique of propensity score [17].
The distribution of gender, race, health status, mental health status, and BMI categories, the matching variables, before and after propensity score matching are summarized in Table 1. Chi-square tests were used to determine the statistically significant differences between PD and no PD groups before and after propensity score matching. Before matching, all the characteristics except race were statistically significantly different between the two groups. Before matching, individuals with PD had a higher proportion of men (54.5%), whereas, the individuals without PD had higher proportion of women (57.8%). Overall, both the groups with and without PD comprised mostly whites (86.0% and 81.5% respectively). There were also statistically significant differences in physical health status between the two groups. A higher proportion of individuals with PD reported fair/poor health (53.9%) as compared to those without PD (23.7%). In terms of mental health status, 33% of the individuals with PD reported fair/poor mental health, while only 11.1% individuals without PD reported fair/poor mental health. With respect to BMI, a lower proportion of individuals with PD (54.5%) were in overweight/obese group compared to 62.7% of individuals without PD. From Table 1, it can be observed that, after propensity score matching, none of the characteristics were statistically significant, and the distributions of individuals with and without PD were balanced.
Table 1.
Description of sample before and after propensity score matching. Medical Expenditure Panel Survey (2000–2009).
| Before matching
|
After matching
|
|||||
|---|---|---|---|---|---|---|
| PD
|
Non-PD
|
Sig | PD
|
Non-PD
|
Sig | |
| Wt% | Wt% | Wt% | Wt% | |||
| 100.0 | 100.0 | 100.0 | 100.0 | |||
| Gender | ** | |||||
| Women | 45.5 | 57.8 | 45.5 | 48.7 | ||
| Men | 54.5 | 42.2 | 54.5 | 51.3 | ||
| Race | ||||||
| White | 86.0 | 81.5 | 86.0 | 86.6 | ||
| Others | 14.0 | 18.5 | 14.0 | 13.4 | ||
| Health status | *** | |||||
| Excellent/very good | 16.0 | 43.3 | 16.0 | 16.2 | ||
| Good | 30.1 | 33.0 | 30.1 | 27.1 | ||
| Fair/poor | 53.9 | 23.7 | 53.9 | 56.7 | ||
| Mental health status | *** | |||||
| Excellent/very good | 28.7 | 57.6 | 28.7 | 31.4 | ||
| Good | 38.0 | 31.4 | 38.1 | 34.6 | ||
| Fair/poor | 33.2 | 11.1 | 33.2 | 34.0 | ||
| Body mass indexa | * | |||||
| Under or normal weight | 45.5 | 37.3 | 45.5 | 43.7 | ||
| Overweight or obese | 54.5 | 62.7 | 54.5 | 56.3 | ||
Note: Based on 351 individuals with PD and 33,474 individuals without PD before matching; and 350 individuals with PD and 1050 individuals without PD after matching among elderly people aged 65 or older. Matching was conducted on gender, race/ethnicity, health status, mental health status, and body mass index. Asterisks represent statistical significance between women and men based on chi-square tests.
Wt%: Weighted percentage, HS: High School.
p < 0.001;
0.001 ≤ p < 0.01.
Numbers do not add up to total numbers due to missing data.
2.4. Measures
2.4.1. Dependent variables
Dependent variables for this study consisted of total, inpatient, outpatient, emergency room, prescription drugs, home health agency and other expenditures (which included dental care vision care and other expenditures). Direct expenditures are defined as payments for different types of services (example: inpatient, outpatient) and from different sources (out-of-pocket or direct payments from the individuals, payments from private insurance, Medicaid, Medicare, Workers’ Compensation, and others) [18]. As our data spanned across 10 years, we also expressed expenditures in real dollars. Medical care services part of the annual consumer price index (CPI) was utilized to transform/convert all expenditures to 2009 constant dollars. The CPI was obtained from the Bureau of Labor Statistics [19]. In addition, expenditure data were skewed to the right; logarithmic transformations of the expenditure data were conducted in order to fit a linear model.
2.5. Co-occurring conditions
Co-occurring conditions were identified using ICD-9-CM codes and Clinical Classification codes. We included the following co-occurring physical conditions: arthritis, asthma, cancer, diabetes, eye, gastro esophageal reflux disorder (GERD), heart disease, hypertension, osteoporosis, stroke, and thyroid disorders and mental health conditions: anxiety and depression. We also used total number of chronic physical conditions as an independent variable in multivariable models.
2.6. Other independent variables
Other independent variables consisted of marital status (married and other), area of residence (metro and non-metro), education (less than high school, high school, and more than high school education), and poverty status (poor, near poor, middle income, and high income).
2.7. Statistical analyses
T-tests were used to examine the differences in average total, inpatient, outpatient, emergency room, pharmacy, home healthcare, and other expenditures between PD and no PD groups. Ordinary Least Square (OLS) regressions were conducted on logged expenditures (2009 dollars) controlling for different co-occurring conditions. All analyses accounted for the complex survey design of the MEPS and were conducted in SAS 9.3. As the current study pooled multiple years (2000–2009), the HC-036 file was used which contains the proper variance structure to obtain appropriate estimates from MEPS data that have been pooled over multiple years and where one or more years are from 1996 to 2001 [20]. The percentage change in expenditure was calculated using the formula expβ–1 [21].
3. Results
Table 2 presents the average total and type of expenditures (inpatient, outpatient, emergency room, prescription drugs, home healthcare, and other) and standard errors among individuals with and without PD. The average total expenditures were $15,404 and $13,333 for individuals with and without PD respectively. The average prescription drug expenditures were higher among individuals with PD ($3787) compared to those without PD ($2515). The average home healthcare expenditures for individuals with PD ($2559) were more than double compared to individuals without PD ($1017). Bottom panel of Table 2 presents average expenditures by service type based on users of specific services. Among users of specific services, the average prescription drug and home health-care expenditures were higher for the individuals with PD ($3843 and $9229 respectively) compared to individuals without PD ($2628 and $6254 respectively). When comparing the ratio of means, it can be seen that the home healthcare expenditures (2559/1017) and prescription drug expenditures (3787/2515) were 2.5 times and 1.5 times higher among individuals with PD compared to those without PD respectively. We did not observe statistically significant differences in inpatient, emergency room, outpatient and other expenditures.
Table 2.
Average total and type of expenditures and standard errors. Elderly individuals with and without PD. Medical Expenditure Panel Survey (2000–2009).
| Parkinson’s disease (N = 350)
|
No Parkinson’s disease (N = 1050)
|
|||
|---|---|---|---|---|
| Average | S.E. | Average | S.E. | |
| Among all users | ||||
| Total* | 15,404 | 999 | 13,333 | 631 |
| Inpatient | 5354 | 731 | 5528 | 441 |
| Emergency room | 303 | 34 | 275 | 44 |
| Outpatient | 2748 | 198 | 3412 | 416 |
| Prescription*** | 3787 | 197 | 2515 | 78 |
| Home healthcare*** | 2559 | 385 | 1017 | 104 |
| Other | 654 | 66 | 586 | 34 |
| Parkinson’s disease
|
No Parkinson’s disease
|
|||||
|---|---|---|---|---|---|---|
| N | Average | S.E. | N | Average | S.E. | |
| Among users of specific servicesa | ||||||
| Inpatient | 90 | 21,438 | 2449 | 271 | 21,312 | 1200 |
| Emergency room | 106 | 1030 | 89 | 262 | 1069 | 150 |
| Outpatient | 339 | 2871 | 193 | 992 | 3582 | 437 |
| Prescription*** | 345 | 3843 | 200 | 1002 | 2628 | 78 |
| Home healthcare*** | 92 | 9229 | 927 | 172 | 6254 | 352 |
| Other | 206 | 1035 | 90 | 565 | 1018 | 57 |
Note: Based on 350 individuals with PD and 1050 matched controls without PD, aged 65 years or older and were alive during particular calendar year. Asterisks represent significant group differences by the presence of PD using t-tests.
SE: Standard Error.
p < 0.001;
0.001 ≤ p < 0.01;
0.01 ≤ p < 0.05.
Type of expenditures among users will not add to average total expenditure because of the denominators.
We also examined the rates of co-occurring chronic physical and mental health conditions among individuals with PD and without PD. Except for depression and diabetes, rates of other co-occurring conditions were not statistically significantly between the two groups after propensity score matching. The proportion of individuals with depression were higher (26.1%) among individuals with PD as compared to individuals without PD (13.1%) whereas, the proportion of individuals with diabetes was lower among individuals with PD (19.1% vs. 24.9%) compared to individuals without PD (data not shown in tabular form). Overall, the average number of chronic physical conditions was slightly lower (2.4) among individuals with PD as compared to individuals without PD (2.7).
As the differences in home healthcare and prescription drug expenditures between those with and without PD were very high as measured by the ratio of means, we further examined the difference in home healthcare and prescription drug expenditures for specific chronic conditions among individuals with and without PD. Table 3, presents average home healthcare and prescription drug expenditures for each co-occurring chronic condition for PD and no PD groups. The ratios of the means (average expenditures for PD group divided by average expenditures of no PD group) are also presented to represent the relative differences in expenditures. For all conditions, except cancer, the average home healthcare expenditures were higher for individuals with PD as compared to those without PD. Specifically, among individuals with anxiety, diabetes, and thyroid conditions, home healthcare expenditures among those with PD were about 5 times those of individuals without PD. Further exploration of individual co-occurring conditions revealed that diabetes, anxiety and depression were most influential and positively associated with home healthcare expenditures among individuals with PD compared to those without PD (data not presented). Similarly, for each specific condition, prescription drug expenditures among individuals with PD were higher compared to the matched controls without PD.
Table 3.
Average total and type of expenditures by co-occurring condition. Elderly individuals with and without PD. Medical Expenditure Panel Survey (2000–2009).
| Home healthcare
|
Ratio of means | Prescription drugs
|
Ratio of means | |||||
|---|---|---|---|---|---|---|---|---|
| PD | Non-PD | Sig | PD | Non-PD | Sig | |||
| Anxiety | 5731 | 1108 | *** | 5.17 | 4932 | 3430 | ** | 1.44 |
| Arthritis | 2514 | 1517 | ** | 1.66 | 4643 | 2955 | *** | 1.57 |
| Cancer | 1402 | 1130 | 1.24 | 3965 | 3053 | ** | 1.30 | |
| Emphysema | 3916 | 1821 | * | 2.15 | 6029 | 3828 | *** | 1.58 |
| Diabetes | 4756 | 939 | *** | 5.06 | 5602 | 3631 | *** | 1.54 |
| Heart disease | 3935 | 1657 | *** | 2.37 | 4752 | 3197 | *** | 1.49 |
| Hypertension | 2271 | 1069 | ** | 2.13 | 4579 | 3140 | *** | 1.46 |
| Depression | 3595 | 1587 | *** | 2.26 | 5022 | 3845 | ** | 1.31 |
| Stroke | 4686 | 1556 | *** | 3.01 | 5472 | 3671 | *** | 1.49 |
| Thyroid | 5072 | 999 | *** | 5.08 | 4584 | 2815 | *** | 1.63 |
Note: Based on 350 individuals with PD and 1050 matched controls without PD, aged 65 years or older and were alive during particular calendar year. Asterisks represent significant group differences by the presence of PD using t-tests. Ratio of means is computed as the ratio of expenditures for PD group divided by average expenditures of no PD group.
SE: Standard Error,
p < 0.001;
0.001 ≤ p < 0.01;
0.01 ≤ p < 0.05.
Table 4 displays the intercept and parameter estimates for individuals with PD from separate OLS regressions on logged total and types of healthcare expenditures. Although comparisons of average expenditures between PD and no PD group did not reveal statistically significant differences in inpatient, emergency room and outpatient expenditures, unadjusted OLS regressions on logged expenditures revealed that the total, inpatient, emergency room, prescription drugs, and home healthcare expenditures were significantly higher among individuals with PD compared to matched-controls without PD (Model 1). The parameter estimates indicated that for total expenditures, those with PD had 111% greater expenditures (β = 0.745) compared to those without PD; for home healthcare expenditures those with PD had 438% greater expenditures (β = 1.683) compared to those without PD. After controlling for gender, race, marital status, metro status, education, poverty status, and MEPS year, we found that individuals with PD had 109% greater total expenditures compared to those without PD (Model 2). Similar results were observed for inpatient, home healthcare, prescription drug, and other expenditures. When the number of chronic conditions and presence of depression and anxiety were included in the model (Model 3), it was observed that individuals with PD had higher total (84%), inpatient (70%), emergency room (92%), prescription drugs (139%), and home healthcare (379%) compared to matched controls without PD.
Table 4.
Intercept and parameter estimates for PD from separate Ordinary Least Square regressions on logged healthcare expenditures (2009 Dollars). Medical Expenditure Panel Survey, 2000–2009.
| Model 1
|
Model 2
|
Model 3
|
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Beta | SE | Sig | Beta | SE | Sig | Beta | SE | Sig | |
| Total | |||||||||
| Intercept | 8.343 | 0.011 | *** | 8.680 | 0.058 | *** | 7.699 | 0.048 | *** |
| PD | 0.745 | 0.070 | *** | 0.741 | 0.071 | *** | 0.611 | 0.054 | *** |
| No PD | |||||||||
| Inpatient | |||||||||
| Intercept | 1.644 | 0.021 | *** | 2.781 | 0.124 | *** | 1.611 | 0.117 | *** |
| PD | 0.747 | 0.244 | ** | 0.700 | 0.248 | ** | 0.533 | 0.223 | * |
| No PD | |||||||||
| Emergency room | |||||||||
| Intercept | 1.107 | 0.016 | *** | 1.599 | 0.085 | *** | 0.961 | 0.084 | *** |
| PD | 0.770 | 0.159 | *** | 0.763 | 0.156 | *** | 0.652 | 0.150 | *** |
| No PD | |||||||||
| Outpatient | |||||||||
| Intercept | 6.736 | 0.016 | *** | 6.877 | 0.085 | *** | 5.806 | 0.079 | *** |
| PD | 0.254 | 0.135 | 0.248 | 0.131 | 0.146 | 0.120 | |||
| No PD | |||||||||
| Prescription drugs | |||||||||
| Intercept | 6.659 | 0.018 | *** | 6.961 | 0.088 | *** | 5.389 | 0.081 | *** |
| PD | 1.061 | 0.071 | *** | 1.079 | 0.073 | *** | 0.870 | 0.080 | *** |
| No PD | |||||||||
| Home healthcare | |||||||||
| Intercept | 0.688 | 0.018 | *** | 1.231 | 0.103 | *** | 0.685 | 0.097 | *** |
| PD | 1.683 | 0.231 | *** | 1.708 | 0.235 | *** | 1.567 | 0.227 | *** |
| No PD | |||||||||
| Others | |||||||||
| Intercept | 3.701 | 0.032 | *** | 3.360 | 0.113 | *** | 2.937 | 0.116 | *** |
| PD | 0.228 | 0.239 | 0.254 | 0.236 | 0.153 | 0.231 | |||
| No PD | |||||||||
Note: Based on 350 individuals with PD and 1050 individuals without PD, aged 65 years or older and were alive during particular calendar year. Asterisks represent significant group differences by the presence of PD using t-tests.
Model 1 included Parkinson’s disease as the independent variable.
Model 2 included Parkinson’s disease, gender, race, marital status, metro status, education, region, and MEPS year as independent variables.
Model 3 included Parkinson’s disease, gender, race, marital status, metro status, education, region, chronic disease number, depression-anxiety, and MEPS year as independent variables.
SE: Standard Error.
p < 0.001;
0.001 ≤ p < 0.01;
0.01 ≤ p < 0.05.
4. Discussion
This study examined excess healthcare expenditures associated with PD and the extent to which various co-occurring conditions contributed to the increased expenditures among individuals with PD. We found that in the model without co-occurring conditions, the average total expenditures among individuals with PD was 109% higher compared to matched controls without PD, when only demographic and socioeconomic variables were entered in the model. After adjusting for co-occurring conditions, depression and anxiety, the average total expenditures among individuals with PD was 84% higher as compared to those without PD. Taken together these two findings suggest that the increased expenditures among individuals with PD compared to those without PD can be partially explained by burden from the co-occurring conditions. These findings emphasize the importance of reducing the comorbidity burden through co-management of PD and other chronic conditions among individuals with PD. Our findings have implications for healthcare delivery. Given that evidence-based approaches to the management of co-occurring disorders are scarce, providers caring for PD patients need to recognize the complexities in treating co-occurring conditions, particularly the risk of polypharmacy and may need to adopt a “holistic” approach.
Results from this study are consistent with a previous study based on Medicare beneficiaries in which the average total expenditures for individuals with PD was 70% higher compared to individuals without PD [5]. Our study extended this prior study, by using a matched case–control design and more recent data. Furthermore, our study makes unique contribution by providing information on the extent to which co-occurring conditions contributed to the excess expenditures among individuals with PD compared to matched controls.
We also observed that inpatient, home healthcare, and prescription drug expenditures were higher for individuals with PD compared to those without PD. Noteworthy finding was that home healthcare expenditures among individuals with PD and diabetes was five times higher compared to those without PD and with diabetes. Individuals with diabetes have greater home healthcare expenditures compared to those without diabetes [22], perhaps due to morbidity burden and/or greater assistance with activities of daily living and medication management compared to those without diabetes [23]. Additionally, the presence of PD may exacerbate the morbidity burden [5]. Therefore, we speculate that the combination of these two conditions may lead to greater need for assistance [24], which in turn can lead to higher home healthcare utilization and expenditures. In terms of prescription drug expenditures, higher expenditures among the individuals with PD compared to those without PD may be attributed to higher rates of polypharmacy among individuals with PD [25]. With the advent of new drugs and progress of the disease, polypharmacy use is widespread among individuals with PD [25].
Co-occurring diabetes, anxiety, and depression among individuals with PD compared to those without PD was found to be the major drivers of excess home healthcare expenditures. Depression [26] and anxiety [27] has been shown to be associated with elevated psychosocial disabilities such as mobility issues and ADL impairment among individuals with PD. Therefore, co-occurring diabetes, anxiety, and depression may amplify the negative health outcomes among individuals with PD. These findings underscore the need to adopt a multidisciplinary and collaborative care approach with access to specialists such as endocrinologists and mental health specialists in addition to neurologists.
Finding from this study also indicated that inpatient expenditures were the major component of the total expenditures comprising of 35% of average total expenditures. While these findings are consistent with the existing literature [11,28], future studies need to examine the reasons behind greater inpatient use. Some reasons for higher inpatient expenditures may include higher severity of PD, associated co-occurring conditions like trauma due falls, vascular disease, and increased motor symptoms [29].
The strengths of this study included the use of nationally representative data on community-dwelling individuals with PD, use of comprehensive set of variables and utilization of propensity score matching which furnishes more accurate and equitable appraisal by eliminating the non-comparable subjects in both groups [30].
Limitations of this study included that all variables were self-reported and subject to recall bias. However, MEPS conducts regular interviews at an interval of 4–5 months (total of five rounds) to minimize recall bias. As with any other cross-sectional study, this study also can only point to associations but cannot establish the cause-effect relationship. Information on severity and duration of PD, severity and duration of other co-occurring conditions was not available; such information could have further explained the differences in expenditures between the two groups.
5. Conclusion
Despite the limitations, our study highlighted the excess expenditures associated with PD compared to matched controls without PD. Although prevalence of chronic conditions between the PD and no PD group were comparable, some of the excess expenditures could be explained by the burden of co-occurring conditions. Collaborative care models might be helpful to reduce the complications of PD and thereby reduce the economic burden of Parkinson’s disease.
Acknowledgments
Mr. Bhattacharjee received research assistance and Dr. Sambamoorthi was partially supported for infrastructure from the West Virginia Collaborative Health Outcomes Research of Therapies and Services (WV CoHORTS) Center.
References
- 1.Muangpaisan W, Mathews A, Hori H, Seidel D. A systematic review of the worldwide prevalence and incidence of Parkinson’s disease. J Med Assoc Thai. 2011;94:749–55. [PubMed] [Google Scholar]
- 2.Alves G, Forsaa EB, Pedersen KF, Dreetz Gjerstad M, Larsen JP. Epidemiology of Parkinson’s disease. J Neurol. 2008;255(Suppl 5):18–32. doi: 10.1007/s00415-008-5004-3. [DOI] [PubMed] [Google Scholar]
- 3.Leibson CL, Maraganore DM, Bower JH, Ransom JE, O’Brien PC, Rocca WA. Comorbid conditions associated with Parkinson’s disease: a population-based study. Mov Disord. 2006;21:446–55. doi: 10.1002/mds.20685. [DOI] [PubMed] [Google Scholar]
- 4.Shiba M, Bower JH, Maraganore DM, McDonnell SK, Peterson BJ, Ahlskog JE, et al. Anxiety disorders and depressive disorders preceding Parkinson’s disease: a case-control study. Mov Disord. 2000;15:669–77. doi: 10.1002/1531-8257(200007)15:4<669::aid-mds1011>3.0.co;2-5. [DOI] [PubMed] [Google Scholar]
- 5.Noyes K, Liu H, Li Y, Holloway R, Dick AW. Economic burden associated with Parkinson’s disease on elderly medicare beneficiaries. Mov Disord. 2006;21:362–72. doi: 10.1002/mds.20727. [DOI] [PubMed] [Google Scholar]
- 6.Karlsen KH, Larsen JP, Tandberg E, Maeland JG. Influence of clinical and demographic variables on quality of life in patients with Parkinson’s disease. J Neurol Neurosurg Psychiatr. 1999;66:431–5. doi: 10.1136/jnnp.66.4.431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hobson DE, Lix LM, Azimaee M, Leslie WD, Burchill C, Hobson S. Healthcare utilization in patients with Parkinson’s disease: a population-based analysis. Parkinsonism Relat Disord. 2012;18:930–5. doi: 10.1016/j.parkreldis.2012.04.026. [DOI] [PubMed] [Google Scholar]
- 8.von Campenhausen S, Winter Y, Rodrigues e Silva A, Sampaio C, Ruzicka E, Barone P, et al. Costs of illness and care in Parkinson’s disease: an evaluation in six countries. Eur Neuropsychopharmacol. 2011;21:180–91. doi: 10.1016/j.euroneuro.2010.08.002. [DOI] [PubMed] [Google Scholar]
- 9.Winter Y, von Campenhausen S, Brozova H, Skoupa J, Reese JP, Botzel K, et al. Costs of Parkinson’s disease in eastern Europe: a Czech cohort study. Parkinsonism Relat Disord. 2010;16:51–6. doi: 10.1016/j.parkreldis.2009.07.005. [DOI] [PubMed] [Google Scholar]
- 10.Winter Y, Balzer-Geldsetzer M, Spottke A, Reese JP, Baum E, Klotsche J, et al. Longitudinal study of the socioeconomic burden of Parkinson’s disease in Germany. Eur J Neurol. 2010;17:1156–63. doi: 10.1111/j.1468-1331.2010.02984.x. [DOI] [PubMed] [Google Scholar]
- 11.Huse DM, Schulman K, Orsini L, Castelli-Haley J, Kennedy S, Lenhart G. Burden of illness in Parkinson’s disease. Mov Disord. 2005;20:1449–54. doi: 10.1002/mds.20609. [DOI] [PubMed] [Google Scholar]
- 12.Pressley JC, Louis ED, Tang MX, Cote L, Cohen PD, Glied S, et al. The impact of comorbid disease and injuries on resource use and expenditures in parkinsonism. Neurology. 2003;60:87–93. doi: 10.1212/wnl.60.1.87. [DOI] [PubMed] [Google Scholar]
- 13.Rubenstein LM, Chrischilles EA, Voelker MD. The impact of Parkinson’s disease on health status, health expenditures, and producitivity: estimates from the National Medical Expenditure Survey. Pharmacoeconomics. 1997;12:486–98. doi: 10.2165/00019053-199712040-00006. [DOI] [PubMed] [Google Scholar]
- 14.Wacholder S, Silverman DT, McLaughlin JK, Mandel JS. Selection of controls in case-control studies. III. Design options. Am J Epidemiol. 1992;135:1042–50. doi: 10.1093/oxfordjournals.aje.a116398. [DOI] [PubMed] [Google Scholar]
- 15.Medical Expenditure Panel Survey (MEPS) [accessed September 2012]; Available from: http://meps.ahrq.gov/mepsweb/survey_comp/mpc_data_collection.jsp.
- 16.Machlin S, Cohen J, Elixhauser A, Beauregard K, Steiner C. Sensitivity of household reported medical conditions in the Medical Expenditure Panel Survey. Med Care. 2009;47:618–25. doi: 10.1097/MLR.0b013e318195fa79. [DOI] [PubMed] [Google Scholar]
- 17.Coca-Perraillon M. [accessed September 2012];Matching with propensity scores to reduce bias in observational studies. 2006 :1–10. Available from: http://www.nesug.org/proceedings/nesug06/an/da13.pdff.
- 18.Agency for Healthcare Research and Quality (AHRQ) [accessed September 2012]; Available from: http://meps.ahrq.gov/mepsweb/data_files/publications/st387/stat387.pdf.
- 19.Bureau of Labor Statistics (BLS) [accessed September 2012]; Available from: http://data.bls.gov/cgi-bin/surveymost.
- 20. [accessed September 2012];Medical Expenditure Panel Survey (MEPS) HC-036. Available from: http://meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h36/h36u09doc.pdf.
- 21.Halvorsen Robert PR. The interpretation of dummy variables in semi-logarithmic equations. Am Econ Rev. 1980;70:474–5. [Google Scholar]
- 22.American Diabetes Association (ADA) Statement. Economic costs of diabetes in the U.S in 2007. Diabetes Care. 2008;31:596–615. doi: 10.2337/dc08-9017. [DOI] [PubMed] [Google Scholar]
- 23.Helaine RE. Diabetes among recipients of home health services in the United States. Home Heal Care Manage Prac. 2012;24:234–41. [Google Scholar]
- 24.Vickers LF, O’Neill CM. An interdisciplinary home healthcare program for patients with Parkinson’s disease. Rehabil Nurs. 1998;23:286–9. 299. doi: 10.1002/j.2048-7940.1998.tb01806.x. [DOI] [PubMed] [Google Scholar]
- 25.Tuite P, Riss J. Recent developments in the pharmacological treatment of Parkinson’s disease. Expert Opin Investig Drugs. 2003;12:1335–52. doi: 10.1517/13543784.12.8.1335. [DOI] [PubMed] [Google Scholar]
- 26.Oluwadamilola OO, Njideka OU, Frank OI, Mustapha DA. Frequency of cognitive impairment and depression in Parkinson’s disease: a preliminary case-control study. Niger Med J. 2012;53:65–70. doi: 10.4103/0300-1652.103544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Marinus J, Visser M, Martinez-Martin P, van Hilten JJ, Stiggelbout AM. A short psychosocial questionnaire for patients with Parkinson’s disease: the SCOPAPS. J Clin Epidemiol. 2003;56:61–7. doi: 10.1016/s0895-4356(02)00569-3. [DOI] [PubMed] [Google Scholar]
- 28.Vossius C, Nilsen OB, Larsen JP. Parkinson’s disease and nursing home placement: the economic impact of the need for care. Eur J Neurol. 2009;16:194–200. doi: 10.1111/j.1468-1331.2008.02380.x. [DOI] [PubMed] [Google Scholar]
- 29.Martignoni E, Godi L, Citterio A, Zangaglia R, Riboldazzi G, Calandrella D, et al. Comorbid disorders and hospitalisation in Parkinson’s disease: a prospective study. Neurol Sci. 2004;25:66–71. doi: 10.1007/s10072-004-0232-5. [DOI] [PubMed] [Google Scholar]
- 30.D’Agostino RB., Jr Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med. 1998;17:2265–81. doi: 10.1002/(sici)1097-0258(19981015)17:19<2265::aid-sim918>3.0.co;2-b. [DOI] [PubMed] [Google Scholar]