Abstract
The purpose of this exploratory study was to examine whether demographic and disability variables predict total health care expenditure of Wisconsin PROMISE. The findings are intended to assist in promoting cost-effectiveness for future similar initiates. This study data were extracted from Wisconsin PROMISE data set. This study had a total of 1,443 youth with disabilities (Mage = 14.89). The majority of participants were male (69%). Our results indicated that some demographic and disability–related characteristics are associated with total health care expenditure in control with VR case during PROMISE, control without VR case during PROMISE, and treatment group. Overall, findings of the current study suggest demographic and disability variables do assist in predicting total health care expenditure of Wisconsin PROMISE.
Employment provides many financial and psychosocial benefits to youth with disabilities, their families, societies, and government (Anderson et al., 2019; Hartman et al., 2019). Since Wisconsin PROMISE started, the employment rate among youth has increased from 1% (2013) to 67% (Hartman et al., 2019), resulting in millions of dollars annual cost savings (Anderson et al., 2019). For example, “overall savings for SSI, state supplement, and Medicaid are $1,309,876/year for the 5% PROMISE youth currently earning SGA, and a projection of savings of $3,325,065 for the 5% of the approximately 10,000 Wisconsin SSI youth ages 14 to 18” (Anderson et al., 2019, p. 259). In addition, Anderson et al. (2019) reported that youth who received targeted case management and family navigation generated approximately twice as much in tax revenue compared with youth who were in PROMISE control without a Vocational Rehabilitation (VR) case. The total expenditure to implement PROMISE for youth with disabilities was US$229 million (Enayati & Shaw, 2019), and service costs during 6-year period were about US$8.2 million for Wisconsin PROMISE participants (Anderson et al., 2019). To promote cost-effectiveness of future similar initiates, it is important to examine correlates of total annual expenditures. Therefore, we aim to examine whether demographic and disability variables predict total health care expenditure of Wisconsin PROMISE. This exploratory study aims to inform clinicians, researchers, policy makers, and stakeholders in their work with youth with disabilities.
Method
This study data were extracted from Wisconsin PROMISE data set. For detailed study procedures, please see Hartman et al. (2019). Participants responded to questions related to contact information, household-related information, youth information, health status information, and parent information. The total health care expenditure data come from the PROMISE Management Information System (MIS), and we calculated total health care expenditure by summing all expenditure for participants. Based on data homogeneity requirement, we conducted independent-samples t tests, Kruskal–Wallis H tests, Mann–Whitney U tests, and analyses of variance (ANOVAs) to compare group differences. Finally, we conducted a series of regression analysis to examine whether demographic and disability–related variables predict the total health care expenditure in three groups.
Results
This study had a total of 1,443 youth with disabilities (Mage = 14.89). The majority of participants were male (69%) and Black (47.9%). We categorized disabilities as developmental and intellectual disabilities (40.1%), including autism and intellectual disability; emotional and behavioral disability (18.7%); specific learning disability (11.8%); sensory disabilities (2.1%), including visual impairment, hearing impairment, and speech or language impairment; physical disabilities (1.5%), including traumatic brain injury and orthopedic impairment; and other health impairments (25.9%). Some participants received treatment (TG) (49.8%), and others were in control group (17.2%) with VR case during PROMISE [CwVR] and 33.1% without VR cases during PROMISE [Cw/oVR]. Please see Table 1.
Demographic Characteristics and Total Health Care Expenditure.
Group Differences
CwVR
We found that total health care expenditure for control with VR case during PROMISE did not change based on gender differences, t(246) = .48, p = .63. ANOVA results revealed that total health care expenditure did not change in race groups, F(4, 243) = .84, p = .50. Kruskal–Wallis H tests results revealed that total health care expenditure did not change in disability types, H = 10.95, p = .052.
Cw/oVR
Mann–Whitney U test results revealed that total health care expenditure significantly changed in gender groups, with females having significantly higher total health care expenditures than males, U = 17,919.00, p = .00. ANOVA results revealed that total health care expenditure did not change in race groups, F(4, 472) = .96, p = .43. Kruskal–Wallis H tests results revealed that total health care expenditure did not change in disability types, H = 9.66, p = .09.
TG
We found that total expenditure did not change based on gender differences, t(716) = −.73, p = .47. ANOVA results revealed that total health care expenditure did not change in race groups, F(4, 713) = 1.81, p = .13. Kruskal–Wallis H tests results revealed that total health care expenditure significantly changed in disability types, with physical disabilities having the highest total health care expenditure followed by developmental and intellectual disabilities, H = 13.51, p = .02.
Predicting Total Expenditure
Regression analysis for CwVR result revealed that demographic and disability variables accounted for a significant proportion of variance in total health care expenditure, R = .21, R2 = .04, F(4, 243) = 2.73, p < .05. Specifically, age at intake (β = .17) positively associated with total health care expenditure. Regression analysis for Cw/oVR result revealed that demographic and disability variables accounted for a significant proportion of variance in total health care expenditure, R = .24, R2 = .06, F(4, 472) = 7.44, p < .001. Specifically, age at intake (β = .17), gender (β = .15), and disability type (β = .09) positively associated with total health care expenditure. Regression analysis for TG result revealed that demographic and disability variables accounted for a significant proportion of variance in total health care expenditure, R = .22, R2 = .05, F(4, 713) = 9.26, p < .001. Specifically, age at intake (β = .20) positively associated with total health care expenditure.
Discussion
We found that total health care expenditure did not change based on gender differences for CwVR and treatment subsamples although females had significantly higher total health care expenditure compared with males in Cw/oVR subsample. Finally, we found that total health care expenditure significantly changes in disability types for Cw/oVR and treatment subsamples, but not CwVR subsample. For both Cw/oVR and treatment subsamples, physical disabilities were found to have the highest amount of total health care expenditure compared with other disability types. Although there is limited study on this topic, a previous research studied the costliest conditions in the United States (Druss et al., 2002). Authors reported that the two most expensive conditions were respiratory malignancies and ischemic heart disease (Druss et al., 2002), which are physical conditions.
After we found some group differences in total health care expenditure, we examined whether any demographic and disability characteristic predicts total health care expenditure in youth with disabilities in three subgroups. We found that age at the time of intake significantly predicts total health care expenditure in all subgroups, indicating that increase in age positively predicts total health care expenditure. Regarding gender, we found that being female is significantly associated with total health care expenditure in Cw/oVR subgroup. Regarding disability type, we found that having developmental and intellectual disabilities positively associated with total health care expenditure in only Cw/oVR subgroup.
Overall, findings of the current study suggest demographic and disability variables do assist in predicting total health care expenditure of Wisconsin PROMISE. A combination of factors such as the age of intake, gender, race, and disability can contribute to the total health care expenditure.
Footnotes
Acknowledgements
The views expressed herein do not necessarily represent the positions or policies of the Department of Education, the Wisconsin Department of Workforce Development, or their federal or state partners. No official endorsement by the U.S. Department of Education or the Wisconsin Department of Workforce Development of any product, commodity, service, or enterprise mentioned in this publication is intended or should be inferred.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The contents of this paper were developed under a cooperative agreement with the U.S. Department of Education, Office of Special Education Programs, associated with PROMISE Award #H418P140002. Selete Avoke served as the project officer.