Printer Friendly

Assessing the social determinants of health care costs for Medicaid-enrolled adolescents in Washington State using administrative data.

1 | INTRODUCTION

Social determinants, or "the conditions in which people are born, grow, [and] live," have significant impacts on health and health care costs. (1) The family environment during childhood is a primary social determinant of health. Family adversity during childhood contributes to lifelong poor physical and mental health (2-6) and elevated medical costs. (7-10) Identification of the childhood experiences common in high-cost populations can help shape interventions and policies designed to reduce health spending. In 2015, U.S. national health expenditures equaled $3.2 trillion, almost 18% of the U.S. Gross Domestic Product (11) and Medicaid spending alone comprised $545 billion, or 17% of total expenditures. The current study examines medical costs of over 180 000 Medicaid-enrolled adolescents with the goal of identifying childhood adversities and social conditions most associated with increased health care utilization and costs. Using administrative data to link records for children with those of their biological parents confers a unique opportunity to assess social determinants and to evaluate impacts on health care costs net of other factors.

Historically, children age 0-18 account for approximately 25% of total Medicaid spending (11) and previous research links childhood adversity to increased Medicaid expenditures. A 2013 study demonstrated that maltreated Medicaid-enrolled children cost $2600 more per year than children who were not maltreated and estimated that excess cost due to maltreatment accounted for 9% of all child Medicaid expenditures. (7) Adverse Childhood Experiences (ACEs) studies investigate a broader set of adversities including maltreatment as well as family environment factors such as parent substance abuse, incarceration, and domestic violence. ACEs studies emphasize that individuals who experience a single ACE are likely to experience another ACE (12) and that the likelihood of negative health outcomes increases as the number of ACEs increases. (4,5,10,13) Child adversity, or ACEs, are important social determinants of health for young people, because they shape the conditions in which they are born, grow up, and live. Earlier identification of and intervention with children experiencing adversity could reduce health care costs. Since a small subset of the population accounts for a large percentage of total medical costs, (14,15) targeting a small number children, those experiencing childhood adversity, could have large economic impacts.

The current study used data from the Washington State Department of Social and Health Services Integrated Client Databases (16) to assess the association between health care costs and prior childhood adversity and to identify adversities that were the most impactful on health care costs. The utility of administrative data has been demonstrated and reviewed. (4,17-19) Using administrative data, child risk factors can be determined without reliance on retrospective self-report or caregiver report, and this method has been shown to replicate findings from studies that used self-report or national surveys. (4) Administrative datasets are large, providing high statistical power, and are longitudinal, with historical data from multiple decades. The ability to link service records for children to those of their parents and to integrate data from a variety of state systems delivers a comprehensive profile of childhood adversity. This information, combined with Medicaid claims, service encounters, and expenditures, provides a unique opportunity to identify childhood adversities associated with high-cost adolescents.

2 | METHODS

2.1 | Study cohort

The study cohort consisted of children insured by Washington State Medicaid for at least 1 month between 7/1/2014 and 6/30/2015 who were between the ages of 12 and 17 years old during their first month of coverage and who did not die during that period. Children eligible for inclusion had at least one biological parent identified through state administrative records. Of 233 054 individuals in the cohort, 181 176 (77.7%) were linked to at least one parent and were therefore included in the study cohort. Of the cohort, 30 483 (16.8%) children were linked to one biological parent, and 150 693 (83.2%) were linked to both biological parents.

Child and parent administrative data were linked to identify both child and parent risk factors that could impact health care costs. Parent risk factors were identified based on biological parents regardless of custody status. Data were compiled using the Washington State Department of Social and Health Services Integrated Client Databases, (16) which links data from various state systems on individual characteristics, services, outcomes, and costs. Data sources used in this study included the Department of Social and Health Services (child welfare, economic services, developmental disabilities, mental health, and substance abuse), Health Care Authority (Medicaid), State Patrol (arrests), Administrative Office of the Courts (filings, adjudications, or convictions), Department of Corrections (prison incarcerations), and the Department of Health (birth and death records). The study was granted an Institutional Review Board exempt status because it relied on deidentified retrospective records.

2.2 | Measures

2.2.1 | Independent variables

Independent variables were categorized as maltreatment history, parent risk factors, demographics, medical coverage, or child social and health risk factors (full descriptions in Table S1). Maltreatment history factors included sexual abuse, physical abuse, neglect, and out-of-home placements. Sexual abuse, physical abuse, and neglect were defined as any allegation with the indicated type of maltreatment that generated an intake, regardless of whether that intake led to an investigation or substantiation. During the study period, Washington State was rolling out a differential response system, where low-risk families referred to state Child Protective Services could be connected to services with no investigation or findings. Therefore, limiting to cases where abuse/neglect was investigated or where abuse or neglect was substantiated was not advisable. Prior research has found that unsubstantiated cases often have similar long-term health impacts as substantiated cases. (20) Recent abuse or neglect (in 2014) was distinguished from historical physical abuse, sexual abuse, and neglect (lifetime, excluding 2014) to discern the impact of temporality of maltreatment. Further, age at first maltreatment allegation and first out-of-home placement were included to measure possible age differences in impacts of maltreatment and out-of-home placement on costs. The out-of-home placement variable included foster care, kinship care, and congregate care placements. Extreme placement instability, defined as five or more lifetime placements, was also included as an adverse experience.

Parent risk factors included criminal behavior, domestic violence, disability, medical risk, (21) homelessness, poverty (food assistance (SNAP)--moderately low income; cash assistance (TANF)--very low income), mental health condition, substance use disorder, and death. Parent risk factors were assessed for the 5 years leading up to the study year (2010-2014), except where noted otherwise. The parent risk factors correspond to the original ACEs household dysfunction factors, with the addition of parental disability and medical risk. The maltreatment history variables and parent risk factors together measure the social conditions under which children grew up, or family-level social determinants of health. These variables and their relationship to costs were the central focus of this analysis.

Child social and health risk factors included being a parent, medical risk, (21) medical complexity, (22) juvenile justice involvement, mental health need, and substance use. Child social and health risk factors were included to control for costly adolescent experiences in order to isolate impacts of child maltreatment and parent risk factors net of these variables. Demographic variables included age, sex, race, and limited English proficiency of the parent. Medical coverage controls included months of Medicaid coverage and months of third-party coverage. Medicaid coverage months were important to include to control for duration of access to Medicaid-paid services in the study period. Third-party coverage typically reflects enrollment in employer-based health insurance by families with incomes low enough to qualify for Medicaid. Under Medicaid rules, third-party coverage is the first payer for covered services and therefore is associated with lower Medicaid-paid health service costs.

2.2.2 | Dependent variable

Health care costs were defined as federal and state funds used to provide health services in the study year (2015) and were divided into medical system and behavioral health. In Washington State during this period, behavioral health services were provided through behavioral health organizations contracting with service providers while medical services were administered by a separate medical system. Therefore, this analysis separately attended to Medicaid costs in the medical system and those that accrued to the currently separate behavioral health service system.

In Washington State, medical assistance includes a mental health benefit for children that covers services received in a medical setting, including primary care. Because these services were delivered in a medical setting, they were included in the measure of medical system costs. More functionally impaired children were served through the behavioral health service system and these services were included in behavioral health costs. Medical costs also included prescriptions such as psychotropic medications. Due to psychotropic medication and some primary care mental health services falling under medical system costs, associations between risk factors and behavioral health costs, broadly defined, may be an underestimate. There are substantive reasons to examine medical and behavioral health costs separately, including prior research on child adversity's impact on both adolescent physical and mental health. (4,23)

Medical system costs were calculated from the provider amount paid for fee-for-service claims, the managed care amounts paid to reported provider for managed care encounters, and other fees derived from the Medicaid reporting system in Washington State. Medical costs included the following types of costs: outpatient hospital emergency department (ED), outpatient hospital non-ED, inpatient, prescription drug, ambulatory care, and other costs (eg, transportation, medical equipment). Behavioral health costs were calculated based on payments made to behavioral health service providers and include outpatient mental health services, substance use disorder treatment, behavioral rehabilitation services, and inpatient mental health treatment.

2.3 | Analysis

The study population was described including summary statistics on the independent and dependent variables. Bivariate comparisons identified the prevalence of select risk factors in the top decile relative to the bottom nine deciles of the cohort based on total medical system or behavioral health costs. Regression modeling was conducted to isolate the relationship between medical system and behavioral health costs and our measures of interest.

Because both behavioral health costs and medical system costs were zero for a nontrivial proportion of the population of interest, a two-step model was used to predict costs. In the first step, two separate logistic regression models predicting behavioral and medical system costs, respectively, were fit to assess which factors contributed to the likelihood of receiving care. In the second step, two ordinary least squares (OLS) models predicting behavioral and medical system costs for the subset of youth with nonzero costs were run using the same predictor variables as in the first step. The second step was also estimated using generalized linear models (GLM) with the gamma distribution and log link function. The substantive conclusions were not different in the GLM models, so the results from the OLS regression are presented here because the resultant coefficients are more easily interpretable as changes in dollars associated with each factor. We note that OLS regression models are most commonly used in health care risk-adjustment processes because they predict mean expenditures more accurately for all risk and demographic subgroups, which renders results more readily interpretable in health care policy contexts. (24) All costs are listed as 2015 dollars, rounded to the nearest tens or hundreds of dollars in the text and signify the predicted increase or decrease in costs per individual for a given risk factor net of other factors in the model.

We combined the results from the first and second step in order to estimate the overall impact on behavioral and medical system costs of each predictive factor of interest. In order to calculate the overall impact, the predicted probability of having any expenditures from the first step was multiplied by the predicted value of expenditures from the second step for a subject with and without the focal risk factor, and then the average difference for all subjects was calculated to show the overall impact. A confidence interval of the overall impact was estimated by bootstrapping with 1000 replicates. All statistical analyses were run in SAS Statistical Software version 9.4 (SAS Institute Inc., Cary, NC, USA).

3 | RESULTS

Descriptive statistics for the cohort are presented in Table 1. The 181 176 adolescents were 49.2% male and 50.8% female. Participant age ranged from 12 to 17, with an average age of 14.4. On average, participants were Medicaid insured for 11.0 months of the year-long outcome period with an average of 1.2 months of third-party liability coverage. Racial composition was 42.8% non-Hispanic white, 29.2% of Hispanic origin, 25.6% non-Hispanic minority, and 2.3% unknown. More than one-third (36.3%) of the cohort had a mental health condition, 40.9% were reported to state CPS for neglect or abuse during their lifetime, and 78.8% experienced at least one family risk factor.

On average, each cohort member's yearly medical costs were $1688, while median costs were $605 indicating a skewed distribution of medical costs with a small number of very costly children. A majority of youth had low or no medical system costs: 13.4% of individuals had no medical costs for the year. An average of $632 was spent on behavioral health costs per cohort member, while the median was $0 due to nearly 87% of the cohort receiving no behavioral health services in the year.

To characterize risk factors associated with high costs, the percentage of individuals who experienced parent and child risk factors in the top decile and bottom nine deciles for medical costs and for behavioral health costs were compared (Table 2). The adolescents with the highest health care costs were 1.4-2.0 times more likely to have experienced at least five parent risk factors than the lower deciles, and 1.4-1.9 times more likely to have experienced abuse or neglect or 5 + out-of-home placements. In particular, high-cost adolescents were more likely to have experienced recent maltreatment: Those in the top decile for medical costs were 2.0 times more likely to have experienced recent maltreatment and those in the top decile for behavioral health costs were 3.9 times more likely to have experienced recent maltreatment.

Given that a variety of child and parent risk factors were more frequently experienced by the adolescents with the highest health care costs, two-step regression modeling was used to isolate risk factors associated with increased health care usage and costs while controlling for other variables. First-stage logistic regression modeling (Table 3) showed that youth with histories of abuse, neglect, or out-of-home placement were more likely to have costs in the behavioral health system relative to similar children without maltreatment or foster care history. Recent (prior year) abuse or neglect more than doubled a young person's odds of behavioral health system costs (OR = 2.39), while any lifetime physical abuse and neglect were associated with increased odds of behavioral health costs of 12% and 18%, respectively. Recent out-of-home placements (at age 12 or older) were associated with increased likelihood of behavioral health costs by about 34%. Youth who experienced extreme placement instability (5+ placements) were also more likely to experience behavioral health system costs (OR = 1.44).

History of neglect, abuse, or out-of-home placement generally had no association with the likelihood of incurring medical costs or was associated with lower likelihood. Any lifetime neglect was associated with lower likelihood of medical costs (OR = 0.87), as were out-of-home placements starting before the age of 12 (age 0-3: OR = 0.72; age 4-11: OR = 0.76), and abuse or neglect in the prior year (OR = 0.91). However, youth with a history of sexual abuse had a slightly higher likelihood of incurring medical system costs (OR = 1.09) relative to similar children without these risk factors. No cost differences were found by age at first abuse or neglect allegation.

Most parent risk factors showed no discernable pattern of associations with the likelihood of incurring behavioral health and medical system costs, with the exception of parent mental health need, which was associated with increased likelihood of costs in both systems by 9%-12%, and food assistance and cash assistance use (proxy for low and very low income), which was associated with decreased likelihood of costs in both systems by 8%-20%.

Second-step regression modeling (Table 4) for children with nonzero systems costs showed that child maltreatment was associated with greater costs, especially to the behavioral health system. Abuse and neglect within the year prior to this study and historical physical and sexual abuse were associated with increased behavioral health system costs. Recent abuse was associated with annual per person behavioral health cost increases of $3600, lifetime physical abuse was associated with cost increases of $840, and lifetime sexual abuse was associated with cost increases of $1600 net of other factors. A history of neglect alone did not significantly predict behavioral health costs, net of other factors.

For medical system costs, recent abuse and neglect were associated with a cost increase of $470, history of sexual abuse was associated with a cost increase of $340, and history of neglect was associated with a cost increase of $260. A history of physical abuse was not associated with increased medical costs, net of other controls.

History of out-of-home placement at any age was associated with higher behavioral health system costs, though the biggest monetary impact was found for extreme placement instability ($9800). Recent placement at age 12 or later was associated with a $3700 increase in behavioral health costs; in contrast, first placement at age 0-3 increased behavioral health costs by $1100 and first placement at age 4-11 increased behavioral health costs by $1400. For medical system costs, a first placement in early life (before age three) was associated with modest increases in costs ($440), as was placement instability ($740). No differences in behavioral health or medical costs were found by age at first abuse or neglect allegation.

The associations between parent risk factors and costs showed few consistent patterns. However, low-income status was associated with lower costs, net of other factors in the model. While the bivariate comparisons indicated that costlier children had experienced more parent risk factors, when regression modeling controlled for maltreatment experiences and other child characteristics, largely mixed or nonsignificant associations were found for parent risk factors.

Overall estimates of the impacts of maltreatment and parent risk measures on behavioral health costs and medical costs were calculated based on the results of the two-step models (Table 5). Placement instability had the largest overall association with increased behavioral health costs ($1700) and medical costs ($620). Prior year abuse and neglect had the next largest estimated association with increased behavioral health costs ($1100) and medical costs ($400). Experiencing a first out-of-home placement at age 12 or older was also associated with a large increase in behavioral health costs ($700), but not with medical costs. First out-of-home placement experiences before the age of 12 were associated with smaller increases in behavioral health costs ($150-$170), while in the case of medical costs, increases of about $340 were only found for children who had a first placement prior to age 4. Maltreatment experiences also had associations with overall costs. Physical and sexual abuse allegation history was associated with increased behavioral health costs of $160 and $240, respectively, with neglect showing no association. In the case of medical costs, sexual abuse and neglect were associated with overall cost increases of $300 and $210, respectively, while historical physical abuse showed no association.

Parent risk factors did not exhibit the same strong association with increased behavioral health and medical costs. Some risk factors, like homelessness and parent death, were associated with modest increases in behavioral health costs, but had estimates near zero for medical costs. Other factors were associated with moderate decreases in costs or were not associated with costs at all. The only parental factor that showed a strong association with costs was poverty. Adolescents whose parents had used food or cash assistance had lower behavioral health and medical costs. Of all the adverse experiences examined in our models, maltreatment, out-ofhome placement, and placement instability were the factors consistently associated with increased costs to the behavioral health and medical systems.

4 | DISCUSSION

By using integrated administrative records, the association between parent and child risk factors and annual health care costs was assessed for a population of over 180 000 Medicaid-insured adolescents without reliance on retrospective self-report. Among the social determinants examined here, child maltreatment, out-of-home placement, and placement instability had the largest cost impact to the Medicaid system (Tables 3, 4 and 5). These experiences were associated with higher use of behavioral health services, even when controlling for pre-existing behavioral health conditions. In contrast, parent risk factors, even though they influence the social conditions under which children grow up, had more modest, mixed effects on costs.

This study found that the adolescents with the highest health care costs tended to experience more parental ACEs compared to adolescents with lower expenses (Table 2), which mirrors prior research demonstrating that more adverse experiences lead to a greater likelihood of negative health outcomes. (4, 5, 10, 13) However, two-stage regression modeling (Tables 3 and 4) demonstrated that the experiences most impactful on costs were those related to child maltreatment (abuse, neglect, out-of-home placement, placement instability) as opposed to parent risk factors (parent domestic violence, criminal justice involvement, mental health need, and disability). The one pattern in parent risk factors across all models was lower utilization and costs among adolescents whose parents relied on cash or food assistance (proxies for low income). This finding could indicate less need for services because of connections to cash and food benefits or could indicate barriers to health care access for low-income individuals. Although adolescents with high health care costs experienced a variety of childhood adversities, the factors most associated with increased health care spending were related to maltreatment. Given that approximately 40% of the studied youth had been reported for some form of abuse or neglect during their lifetime, programs targeting maltreatment, both through prevention and intervention, could have a large impact on costs.

A sizeable body of research evaluates short-term (7,25,26) and long-term (6,9,27) health care costs associated with childhood maltreatment or estimated total costs of child maltreatment. (28-31) Child maltreatment increases costs in multiple domains (health care, criminal justice, education, lifetime lost productivity, and social services), all of which contribute to its large economic burden. In Washington State, the 2012 cost of child welfare services alone was over $500 million. (28) The estimated lifetime costs of maltreatment for all new abuse and neglect cases in 2008 were $124 billion, (31) which is comparable to the economic burden associated with health conditions such as diabetes. (33)

Prior research demonstrates that childhood maltreatment negatively affects childhood mental health, (4,34) with associated costs estimated at over $1 billion annually. (30) In this study, physical or sexual abuse increased annual behavioral health costs by approximately $150-$250, and sexual abuse or neglect increased annual medical system costs by $200-$300. Neglect or abuse within the prior year increased behavioral health costs by $1100 and medical costs by about $400. In this study, recent maltreatment was associated with larger cost increases than lifetime measures, which concurs with prior research demonstrating that recent adversity has a larger impact on adolescent health than earlier life adversity. (10)

Prior research suggests that as the number of recent foster care placements increases for children, the likelihood of high mental health service usage increases. (35) The current study found that high placement instability had the largest association with increased costs of all social determinants examined and that it was associated with cost increases to both the behavioral health and medical system. Five or more out-of-home placements were associated with a $1700 increase in behavioral health costs and $620 increase in medical costs. Age at first placement had differing impacts on behavioral health and mental health system costs. A first out-of-home placement after age 11 was associated with a $700 increase in behavioral health costs, while the associations were smaller (around $150) for first placements prior to age 12. In contrast, first out-of-home placement before the age of 4 was associated with around $340 in medical costs, whereas later first out-of-home placement experiences were not associated with medical costs.

These findings suggest that preadolescent interventions targeting child maltreatment could have a large impact on health care costs associated with maltreatment and multiple foster care placements. Meta-analyses assessing the efficacy of early interventions on child outcomes find that overall, early interventions benefit children and the economy. (36-41) In particular, programs designed to decrease child maltreatment have been shown to not only be successful, (39,42) but to provide net positive benefit to cost ratios. (40) Prevention of abuse and neglect and interventions in response to maltreatment require a cross-systems approach that connects multiple key policy sectors, including health care. (43) For example, primary health care visits offer a point of contact for universal prevention efforts and screening of children for maltreatment risk factors, as well as an opportunity to make referrals for evidence-based interventions in response to abuse and neglect. (44)

This study had a number of limitations. Adolescents were included only if they could be linked with at least one biological parent through administrative data, resulting in the exclusion of 22% of the potential cohort. Parent risk factors were based solely on biological parents, which did not account for risk factors potentially associated with other caregivers. Further, the frequency and/or severity of risk factors was not coded--for example, parent domestic violence signified only that one or more occurrence was documented and also would not register undocumented events. The current study only examined health care costs associated with child risk factors in adolescents. Future studies should expand to younger age groups and expand beyond health care costs (eg, criminal justice costs).

This study was also limited because it was observational in nature and established associations but not causality. There are few alternatives to observational studies when examining social determinants of health, given that maltreatment and parental risk factors cannot be randomly assigned to youth. The analyses included controls for a number of factors that drive costs, such as medical complexity and behavioral health conditions, but there remains a possibility that unmeasured factors affecting both maltreatment and costs could account at least partially for the identified associations.

This study used administrative data linking Medicaid-insured adolescents to their biological parents to assess the impact of childhood social conditions (social determinants) on health care costs. A small subset of the studied population accounted for the majority of costs. Although these adolescents with high health care expenses tended to experience multiple risk factors, multivariate analyses indicated that the childhood adversities that most increased costs were abuse, neglect, out-of-home placement, and out-of-home placement instability. These results suggest that prevention and intervention efforts focused on children and families who are at risk for child welfare system involvement and for those children who have already experienced maltreatment could substantially reduce the health care costs in Medicaid-insured adolescent populations.

ACKNOWLEDGMENTS

Joint Acknowledgment/Disclosure Statement: This study is an extension of a project funded in part by the Washington State Department of Social and Health Services, Division of Behavioral Health and Recovery. The authors would like to thank David Mancuso, Andy Glenn and Ken Lee for their consultation on the use of medical cost data and concepts; Barbara Felver for design and formatting of tables; Irina Sharkova and Lijian He for their earlier development of family risk factor measures; and the RDA data teams and program partners who make this work with integrated data possible.

ORCID

Deleena A. Pattern [iD] http://orcid.org/0000-0002-8223-4753

REFERENCES

(1.) Marmot M, Friel S, Bell R, Houweling TA, Taylor S. Closing the gap in a generation. Lancet. 2008;372:1661-1669.

(2.) Springer KW, Sheridan J, Kuo D, Carnes M. Long-term physical and mental health consequences of childhood physical abuse: results from a large population-based sample of men and women. Child Abuse Negl. 2007;31(5):517-530.

(3.) Chapman DP, Whitfield CL, Felitti VJ, Dube SR, Edwards VJ, Anda RF. Adverse childhood experiences and the risk of depressive disorders in adulthood. J Affect Disord. 2004;82(2):217-225.

(4.) Lucenko BA, Sharkova IV, Huber A, Jemelka R, Mancuso D. Childhood adversity and behavioral health outcomes for youth: an investigation using state administrative data. Child Abuse Negl. 2015;47:48-58.

(5.) Felitti VJ, Anda RF, Nordenberg D, et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. Am J Prev Med. 1998;14(4):245-258.

(6.) Arnow BA. Relationships between childhood maltreatment, adult health and psychiatric outcomes, and medical utilization. J Clin Psychiatry. 2004;65(Suppl 1):10-15.

(7.) Florence CS, Brown DS, Fang X, Thompson HF. Health care costs associated with child maltreatment: impact on medicaid. Pediatrics. 2013;132(2):312-318.

(8.) Chartier MJ, Walker JR, Naimark B. Separate and cumulative effects of adverse childhood experiences in predicting adult health and health care utilization. Child Abuse Negl. 2010;34(6):454-464.

(9.) Bonomi AE, Anderson ML, Rivara FP, et al. Health care utilization and costs associated with childhood abuse. J Gen Intern Med. 2008;23(3):294-299.

(10.) Flaherty EG, Thompson R, Dubowitz H, et al. Adverse childhood experiences and child health in early adolescence. JAMA Pediatr. 2013;167(7):622-629.

(11.) Centers for Medicare & Medicaid Services. NHE Fact Sheet. https://www.cms.gov/research-statistics-data-and-systems/statistics-trends-and-reports/nationalhealthexpenddata/nhe-fact-sheet.html. Published 2015. Accessed August 8, 2017.

(12.) Dong M, Anda RF, Felitti VJ, et al. The interrelatedness of multiple forms of childhood abuse, neglect, and household dysfunction. Child Abuse Negl. 2004;28(7):771-784.

(13.) Anda RF, Brown DW. Adverse Childhood Experiences & Population Health in Washington: The Face of a Chronic Public Health Disaster: Results from the 2009 Behavioral Risk Factor Surveillance System. 2010. http://www.wvlegislature.gov/senatel/majority/poverty/ACEsinWashington2009BRFSSFinalReportCrittenton.pdf. Accessed June 18, 2017.

(14.) Pires S, Grimes K, Gilmer T, Allen K, Mahadevan R, Henricks T. Identifying Opportunities to Improve Children's Behavioral Health Care: An Analysis of Medicaid Utilization and Expenditures. 2013. http://www.chcs.org/media/ldentifying-Opportunities-to-Improve-Childrens-Behavioral-Health-Care2.pdf. Accessed April 27, 2017.

(15.) Caspi A, Houts RM, Belsky DW, et al. Childhood forecasting of a small segment of the population with large economic burden. Nat Hum Behav. 2016;1(1):5.

(16.) Mancuso D. DSHS Integrated Client Database (No. 11.205). Olympia, WA; 2014. https://www.dshs.wa.gov/sites/default/files/SESA/rda/documents/research-11-205.pdf. Accessed August 9, 2017.

(17.) Riley GF. Administrative and claims records as sources of health care cost data. Med Care. 2009;47(Supplement):S51-S55.

(18.) Arthur KC, Lucenko BA, Sharkova IV, Xing J, Mangione-Smith R. Using state administrative data to identify social complexity risk factors for children. Ann Fam Med. 2018;16(1):62-69.

(19.) VirnigBA, McBean M. Administrative data for public health surveillance and planning. Annu Rev Public Health. 2001;22:213-230.

(20.) Hussey JM, Marshall JM, English DJ, et al. Defining maltreatment according to substantiation: distinction without a difference? Child Abuse Negl. 2005;29(5):479-492.

(21.) Kronick R, Gilmer T, Dreyfus T, Lee L. Improving health-based payment for Medicaid beneficiaries: CDPS. Health Care Financ Rev. 2000;21(3):29-64.

(22.) Simon TD, Cawthon ML, Stanford S, et al. Pediatric medical complexity algorithm: a new method to stratify children by medical complexity. Pediatrics. 2014;133(6):el647-el654.

(23.) Hussey JM, Chang JJ, Kotch JB. Child maltreatment in the United States: prevalence, risk factors, and adolescent health consequences. Pediatrics. 2006;118(3):933-942.

(24.) Centers for Medicare and Medicaid Services. March 31, 2016, HHS-Operated Risk Adjustment Methodology Meeting. 2016. https://www.cms.gov/CCIIO/Resources/Forms-Reports-and-Other-Resources/Downloads/RA-March-31-White-Paper-032416.pdf. Accessed March 15, 2018.

(25.) Peterson C, Xu L, Florence C, Parks SE. Annual cost of U.S. hospital visits for pediatric abusive head trauma. Child Maltreat. 2015;20(3):162-169.

(26.) Wright MS, Litaker D. Childhood victims of violence. Hospital utilization by children with intentional injuries. Arch Pediatr Adolesc Med. 1996;150(4):415-420.

(27.) Walker EA, Unutzer J, Rutter C, et al. Costs of health care use by women HMO members with a history of childhood abuse and neglect. Arch Gen Psychiatry. 1999;56(7):609-613.

(28.) Child Welfare League of America. Washington's Children 2016.2016:1-8.

(29.) Washington State Institute for Public Policy. Benefit-Cost Technical Documentation. Olympia; 2017. http://www.wsipp.wa.gov/TechnicalDocumentation/WsippBenefitCostTechnicalDocumentation.pdf. Accessed June 28, 2017.

(30.) Gelles R. Perlman S. Estimated annual cost of child abuse and neglect; 2012. http://preventchildabuse.org/wp-content/uploads/2016/02/PCA_COM2012-1.pdf. Accessed June 27, 2017.

(31.) Fang X, Brown DS, Florence C, Mercy JA. The economic burden of child maltreatment in the United States and implications for prevention. Child Abuse Negl. 2012;36(2):156-165.

(32.) Brown DS, Fang X, Florence CS. Medical costs attributable to child maltreatment: a systematic review of short- and long-term effects. Am J Prev Med. 2011;41(6):627-635.

(33.) American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. Diabetes Care. 2013;36(4):1033-1046.

(34.) Turner HA, Finkelhor D, Ormrod R. The effect of lifetime victimization on the mental health of children and adolescents. Soc Sci Med. 2006;62(1):13-27.

(35.) Rubin DM, Alessandrini EA, Feudtner C, Mandell DS, Localio AR, Hadley T. Placement stability and mental health costs for children in foster care. Pediatrics. 2004;113(5):1336-1341.

(36.) Anderson LM, Shinn C, Fullilove MT, et al. The effectiveness of early childhood development programs: a systematic review. Am J Prev Med. 2003;24(3 Suppl.):32-46.

(37.) Gorey KM. Early childhood education: a meta-analytic affirmation of the short- and long-term benefits of educational opportunity. Sch Psychol Q. 2001;16(1):9-30.

(38.) Karoly L, Kilburn M, Cannon J. Early Childhood Interventions: Proven Results, Future Promise. RAND Corporation. 2005. https://doi.org/10.7249/MG341

(39.) MacLeod J, Nelson G. Programs for the promotion of family wellness and the prevention of child maltreatment: a meta-analytic review. Child Abuse Negl. 2000;24(9):1127-1149.

(40.) Lee S, Aos S, Pennucci A. What works and what does not? Benefits-cost findings from WSIPP. Washingt State Inst Public Policy. 2015. https://doi.org/10.1002/capr.12016

(41.) Campbell FA, Conti G, Heckman JJ, et al. Early Childhood investments substantially boost adult health. Science (80-). 2014;343(6178):1478-1485.

(42.) Eckenrode J, Ganzel B, Henderson CR, et al. Preventing child abuse and neglect with a program of nurse home visitation: the limiting effects of domestic violence. JAMA. 2000;284(11): 1385-1391.

(43.) Fortson BL, Klevens J, Merrick MT, Gilbert LK, Alexander SP. Preventing Child Abuse and Neglect: A Technical Package for Policy, Norm, and Programmatic Activities, Vol. 66. Atlanta GA; 2016. https://doi.org/10.15620/cdc.38864

(44.) Hornor G. Child maltreatment: screening and anticipatory guidance. J Pediatr Heal Care. 2013;27(4):242-250.

SUPPORTING INFORMATION

Additional supporting information may be found online in the Supporting Information section at the end of the article.

How to cite this article: Patton DA, Liu Q, Adelson JD, Lucenko BA. Assessing the social determinants of health care costs for Medicaid-enrolled adolescents in Washington State using administrative data. Health Serv Res. 2019;54:52- 63. https://doi.org/10.1111/ 1475-6773.13052

Deleena A. Patton PhD [iD] | Qinghua Liu PhD | Jaimie D. Adelson PhD | Barbara A. Lucenko PhD

Division of Research and Data Analysis (RDA), Washington State Department of Social and Health Services, Olympia, Washington

Correspondence

Deleena A. Patton, PhD, Washington State Department of Social and Health Services, Division of Research and Data Analysis (RDA), Olympia, WA 98504-5204.

Email: deleena.patton@dshs.wa.gov

DOI: 10.1111/1475-6773.13052
TABLE 1 Demographic characteristics and prevalence of risk factors
within the population

Variable                                              No.

Total N                                           181 176
Gender
Female                                             89 165
Male                                               92 011
Age
Average (SD)                                           14.4 (1.7)
Average months of coverage in SFY2015
Medicaid (SD)                                          11.0 (2.5)
Third-party liability (SD)                              1.2 (3.4)
Biological parents identified
One                                                30 483
Two                                               150 693
Maltreatment
Maltreatment type
Any abuse or neglect                               74 072
Physical abuse                                     31 343
Sexual abuse                                       12 536
Neglect                                            66 896
Any out-of-home placement                          15 862
No. out-of-home placements (for >1 placement)
1-2 placements                                       9800
Child social and health risk factors
Mental health condition                            65 702
Substance use disorder                             10 388
Arrest or conviction                                 6387
Has a biological child                               1872
Parent risk factors
Domestic violence                                  30 224
Low income
Cash or food assistance (proxy                    128 842
for low income)
Food assistance only (proxy                        80 990
for moderate low income)
Cash assistance (proxy for                         47 852
extremely low income)
Medical
High medical risk score (CDPS)                     42 049
Disability                                         24 693
Died                                                 5647

Variable                                              %

Total N                                            100%
Gender
Female                                            49.2%
Male                                              50.8%
Age
Average (SD)
Average months of coverage in SFY2015
Medicaid (SD)
Third-party liability (SD)
Biological parents identified
One                                               16.8%
Two                                               83.2%
Maltreatment
Maltreatment type
Any abuse or neglect                              40.9%
Physical abuse                                    17.3%
Sexual abuse                                       6.9%
Neglect                                           36.9%
Any out-of-home placement                          8.8%
No. out-of-home placements (for >1 placement)
1-2 placements                                    61.8%
Child social and health risk factors
Mental health condition                           36.3%
Substance use disorder                             5.7%
Arrest or conviction                               3.5%
Has a biological child                             1.0%
Parent risk factors
Domestic violence                                 16.7%
Low income
Cash or food assistance (proxy                    71.1%
for low income)
Food assistance only (proxy                       44.7%
for moderate low income)
Cash assistance (proxy for                        26.4%
extremely low income)
Medical
High medical risk score (CDPS)                    23.2%
Disability                                        13.6%
Died                                               3.1%

Variable                                       No.      %

Race or ethnicity
Hispanic                                    52 820  29.2%
Non-Hispanic White                          77 587  42.8%
Non-Hispanic Black                            9279   5.1%
Non-Hispanic Indian                           3119   1.7%
Non-Hispanic Asian                            4834   2.7%
Non-Hispanic Pacific Islander                 2031   1.1%
Multiple races (non-Hispanic)               27 239  15.0%
Unknown race                                  4258   2.3%

Limited English (parent)                    30 101  16.6%


No. out-of-home placements, continued
3-5 placements                                4458  28.1%
6-8 placements                                1070   6.7%
9 + placements                                 534   3.4%
Age at first placement (for >1 placement)
0-3                                           7876  49.7%
4-11                                          6270  39.5%
12+                                           1716  10.8%

PMCA
Nonchronic                                 126 604  69.9%
Chronic noncomplex                          38 167  21.1%
Chronic complex                             16 405   9.1%

Mental health condition                     71 761  39.6%
Substance use disorder                      44 142  24.4%
Arrest or conviction                        57 737  31.9%

Homeless                                     41417  22.9%

Number of parent risk factors (ACEs)

0                                           38 472  21.2%
1-3                                         88 610  48.9%
4-6                                         43 442  24.0%
7+                                          10 652   5.9%

ACEs, Adverse Childhood Experiences; No., number; PMCA, Pediatric
Medical Complexity Algorithm.

TABLE 2 Prevalence of child adversities in the top decile relative to
the rest of the population for medical and behavioral health costs

                          Medical cost
                          Bottom 90 percent  Top 10 percent    Relative
                                                               risk

Total N                        163 059             18 117       -
Total cost                $121 746 543       $184 091 206       -
Cost per adolescent               $747            $10 161       -
Risk factors
5+ parental ACEs risk               19.1%              29.0%    1.5
factors
Any abuse or neglect in             39.3%              55.2%    1.4
lifetime
Any abuse or neglect in              7.5%              15.0%    2.0
prior year
5 + placements (among               13.8%              21.4%    1.5
children ever placed out
of home)

                          Behavioral health cost
                          Bottom 90 percent  Top 10 percent    Relative
                                                               risk

Total N                      162 416               18 760       -
Total cost                $1 271 699         $113 183 929       -
Cost per adolescent               $8                $6033       -
Risk factors
5+ parental ACEs risk             18.3%                35.4%    1.9
factors
Any abuse or neglect in           37.8%                68.0%    1.8
lifetime
Any abuse or neglect in            6.3%                24.8%    3.9
prior year
5 + placements (among             12.6%                23.5%    1.9
children ever placed out
of home)

TABLE 3 First step models predicting any costs (relative to no costs)
to the behavioral health system and medical system using logistic
regression

                                                    Behavioral health
                                                    system
                                                    Coefficient

Intercept                                           -3.49
Maltreatment history
Any lifetime physical abuse (excluding prior year)   0.12
Any lifetime sexual abuse (excluding prior year)     0.05
Any lifetime neglect (excluding prior year)          0.17
Any abuse or neglect in prior year                   0.87
First out-of-home placement at age 0-3               0.01
First out-of-home placement at age 4-11             -0.04
First out-of-home placement at age 12 or older       0.29
5 or more out-of-home placements                     0.36
First abuse or neglect at age 0-3                    0.11
First abuse or neglect at age 4-11                   0.15
First abuse or neglect at age 12-17                  0.21
Parent risk factors
Parent has mental health condition                   0.11
Parent has a substance use disorder                  0.09
Parent was homeless                                  0.07
Parent was involved in criminal justice system       0.01
Parent was disabled                                  0.11
Parent was involved in domestic violence             0.02
Parent used food assistance, not cash               -0.09
Parent used cash assistance                         -0.14
Parent has disabling condition                      -0.03
Parent died                                          0.15
Child and family demographics
Age in years                                        -0.05
Female                                               0.26
Hispanic                                             0.13
Non-Hispanic Black                                   0.06
Non-Hispanic Indian                                 -0.37
Non-Hispanic Asian                                  -0.14
Non-Hispanic Pacific Islander                       -0.64
Multiple races (non-Hispanic)                        0.35
Reference is non-Hispanic White
Parent was limited English proficient               -0.18
Medical coverage
Count of Medicaid eligibility months                 0.07
Count of third-party liability coverage months       0.00
Child social and health risk factors
Chronic noncomplex health conditions                 0.59
Chronic complex health conditions                    0.65
Reference is nonchronic noncomplex conditions
Child has mental health condition                    1.23
Child has substance use disorder                     0.92
Child is involved in criminal justice system         0.73
Child is a parent                                   -0.68
Child medical risk score (CDPS)                     -0.01

                                                    Behavioral health
                                                    system
                                                    Odds-ratio

Intercept
Maltreatment history
Any lifetime physical abuse (excluding prior year)  1.12 (***)
Any lifetime sexual abuse (excluding prior year)    1.06
Any lifetime neglect (excluding prior year)         1.18 (***)
Any abuse or neglect in prior year                  2.39 (***)
First out-of-home placement at age 0-3              1.01
First out-of-home placement at age 4-11             0.96
First out-of-home placement at age 12 or older      1.34 (***)
5 or more out-of-home placements                    1.44 (***)
First abuse or neglect at age 0-3                   1.12 (*)
First abuse or neglect at age 4-11                  1.17 (***)
First abuse or neglect at age 12-17                 1.24 (***)
Parent risk factors
Parent has mental health condition                  1.12 (***)
Parent has a substance use disorder                 1.09 (***)
Parent was homeless                                 1.07 (***)
Parent was involved in criminal justice system      1.01
Parent was disabled                                 1.11 (***)
Parent was involved in domestic violence            1.02
Parent used food assistance, not cash               0.92 (***)
Parent used cash assistance                         0.87 (***)
Parent has disabling condition                      0.97
Parent died                                         1.16 (***)
Child and family demographics
Age in years                                        0.95 (***)
Female                                              1.30 (***)
Hispanic                                            1.14 (***)
Non-Hispanic Black                                  1.06
Non-Hispanic Indian                                 0.69 (***)
Non-Hispanic Asian                                  0.87 (***)
Non-Hispanic Pacific Islander                       0.53 (***)
Multiple races (non-Hispanic)                       1.42 (***)
Reference is non-Hispanic White
Parent was limited English proficient               0.83 (***)
Medical coverage
Count of Medicaid eligibility months                1.07 (***)
Count of third-party liability coverage months      1.00
Child social and health risk factors
Chronic noncomplex health conditions                1.81 (***)
Chronic complex health conditions                   1.92 (***)
Reference is nonchronic noncomplex conditions
Child has mental health condition                   3.42 (***)
Child has substance use disorder                    2.50 (***)
Child is involved in criminal justice system        2.08 (***)
Child is a parent                                   0.50 (***)
Child medical risk score (CDPS)                     0.99

                                                    Medical system
                                                    Coefficient

Intercept                                           -2.15
Maltreatment history
Any lifetime physical abuse (excluding prior year)  -0.01
Any lifetime sexual abuse (excluding prior year)     0.09
Any lifetime neglect (excluding prior year)         -0.14
Any abuse or neglect in prior year                  -0.10
First out-of-home placement at age 0-3              -0.33
First out-of-home placement at age 4-11             -0.28
First out-of-home placement at age 12 or older      -0.12
5 or more out-of-home placements                    -0.08
First abuse or neglect at age 0-3                    0.09
First abuse or neglect at age 4-11                   0.03
First abuse or neglect at age 12-17                 -0.04
Parent risk factors
Parent has mental health condition                   0.08
Parent has a substance use disorder                  0.00
Parent was homeless                                 -0.04
Parent was involved in criminal justice system       0.03
Parent was disabled                                  0.02
Parent was involved in domestic violence            -0.02
Parent used food assistance, not cash               -0.09
Parent used cash assistance                         -0.23
Parent has disabling condition                       0.12
Parent died                                         -0.07
Child and family demographics
Age in years                                         0.02
Female                                               0.41
Hispanic                                             0.27
Non-Hispanic Black                                  -0.11
Non-Hispanic Indian                                  0.33
Non-Hispanic Asian                                  -0.05
Non-Hispanic Pacific Islander                       -0.23
Multiple races (non-Hispanic)                        0.05
Reference is non-Hispanic White
Parent was limited English proficient                0.34
Medical coverage
Count of Medicaid eligibility months                 0.30
Count of third-party liability coverage months      -0.08
Child social and health risk factors
Chronic noncomplex health conditions                 0.61
Chronic complex health conditions                    0.90
Reference is nonchronic noncomplex conditions
Child has mental health condition                    0.42
Child has substance use disorder                    -0.19
Child is involved in criminal justice system        -0.27
Child is a parent                                    0.48
Child medical risk score (CDPS)                      1.51

                                                    Medical system
                                                    Odds-ratio

Intercept
Maltreatment history
Any lifetime physical abuse (excluding prior year)  0.99
Any lifetime sexual abuse (excluding prior year)    1.09 (*)
Any lifetime neglect (excluding prior year)         0.87 (**)
Any abuse or neglect in prior year                  0.91 (*)
First out-of-home placement at age 0-3              0.72 (***)
First out-of-home placement at age 4-11             0.76 (***)
First out-of-home placement at age 12 or older      0.89
5 or more out-of-home placements                    0.92
First abuse or neglect at age 0-3                   1.09
First abuse or neglect at age 4-11                  1.03
First abuse or neglect at age 12-17                 0.96
Parent risk factors
Parent has mental health condition                  1.09 (***)
Parent has a substance use disorder                 1.00
Parent was homeless                                 0.97
Parent was involved in criminal justice system      1.03
Parent was disabled                                 1.02
Parent was involved in domestic violence            0.99
Parent used food assistance, not cash               0.91 (***)
Parent used cash assistance                         0.80 (***)
Parent has disabling condition                      1.12 (***)
Parent died                                         0.93
Child and family demographics
Age in years                                        1.02 (***)
Female                                              1.50 (***)
Hispanic                                            1.32 (***)
Non-Hispanic Black                                  0.90 (**)
Non-Hispanic Indian                                 1.40 (***)
Non-Hispanic Asian                                  0.95
Non-Hispanic Pacific Islander                       0.80 (***)
Multiple races (non-Hispanic)                       1.05 (*)
Reference is non-Hispanic White
Parent was limited English proficient               1.41 (***)
Medical coverage
Count of Medicaid eligibility months                1.35 (***)
Count of third-party liability coverage months      0.93 (***)
Child social and health risk factors
Chronic noncomplex health conditions                1.84 (***)
Chronic complex health conditions                   2.45 (***)
Reference is nonchronic noncomplex conditions
Child has mental health condition                   1.52 (***)
Child has substance use disorder                    0.83 (***)
Child is involved in criminal justice system        0.77 (***)
Child is a parent                                   1.62 (***)
Child medical risk score (CDPS)                     4.54 (***)

(***) P < 0.001; (**) P < 0.01; (*) P < 0.05.

TABLE 4 Second stage models predicting costs (in dollars) to the
behavioral health system and medical system using OLS regression

                                                    Behavioral health
                                                    system
                                                    Coef. (SE)

Intercept                                           2804 (1030)
Maltreatment history
Any lifetime physical abuse (excluding prior year)     844 (218) (***)
Any lifetime sexual abuse (excluding prior year)      1617 (267) (***)
Any lifetime neglect (excluding prior year)            274 (345)
Any abuse or neglect in prior year                    3633 (357) (***)
First out-of-home placement at age 0-3                1146 (352) (**)
First out-of-home placement at age 4-11               1437 (355) (***)
First out-of-home placement at age 12 or older        3674 (489) (***)
5 or more out-of-home placements                      9826 (460) (***)
First abuse or neglect at age 0-3                     -313 (440)
First abuse or neglect at age 4-11                    -404 (406)
First abuse or neglect at age 12-17                   -829 (432)
Parent risk factors
Parent has mental health condition                      29 (217)
Parent has a substance use disorder                   -221 (219)
Parent was homeless                                    619 (212) (**)
Parent was involved in criminal justice system        -648 (203) (**)
Parent was disabled                                    368 (228)
Parent was involved in domestic violence              -218 (214)
Parent used food assistance, not cash               - 1292 (248) (***)
Parent used cash assistance                         - 1945 (297) (***)
Parent has disabling condition                        -621 (212) (**)
Parent died                                            987 (404) (*)
Child and family demographics
Age in years                                          -243 (56) (***)
Female                                                -182 (169)
Hispanic                                              -650 (235) (**)
Non-Hispanic Black                                     415 (410)
Non-Hispanic Indian                                    984 (683)
Non-Hispanic Asian                                    -645 (807)
Non-Hispanic Pacific Islander                          604 (1520)
Multiple races (non-Hispanic)                         -105 (210)
Reference is non-Hispanic White
Parent was limited English proficient                 -949 (329) (**)
Medical coverage
Count of Medicaid eligibility months                   295 (47) (***)
Count of third-party liability coverage months          89 (25) (***)
Child social and health risk factors
Chronic noncomplex health conditions                  1024 (216) (***)
Chronic complex health conditions                     3115 (264) (***)
Reference is nonchronic noncomplex conditions
Child has mental health condition                      636 (231) (**)
Child has substance use disorder                       388 (239)
Child is involved in criminal justice system          3438 (287) (***)
Child is a parent                                    -3298 (703) (***)
Child medical risk score (CDPS)                        479 (99) (***)

                                                    Medical system
                                                    Coef. (SE)

Intercept                                           -2311 (282)
Maltreatment history
Any lifetime physical abuse (excluding prior year)     26 (85)
Any lifetime sexual abuse (excluding prior year)      336 (109) (**)
Any lifetime neglect (excluding prior year)           264 (131) (*)
Any abuse or neglect in prior year                    466 (147) (**)
First out-of-home placement at age 0-3                444 (140) (**)
First out-of-home placement at age 4-11                 1 (146)
First out-of-home placement at age 12 or older         -4 (260)
5 or more out-of-home placements                      737 (232) (**)
First abuse or neglect at age 0-3                    -232 (160)
First abuse or neglect at age 4-11                   -139 (144)
First abuse or neglect at age 12-17                  -280 (163)
Parent risk factors
Parent has mental health condition                     10 (66)
Parent has a substance use disorder                  -141 (75)
Parent was homeless                                   -54 (72)
Parent was involved in criminal justice system         -1 (67)
Parent was disabled                                    52 (83)
Parent was involved in domestic violence             -184 (78) (*)
Parent used food assistance, not cash                -175 (65) (**)
Parent used cash assistance                          -209 (86) (*)
Parent has disabling condition                        141 (70) (*)
Parent died                                           -75 (143)
Child and family demographics
Age in years                                          145 (15) (***)
Female                                                -16 (50)
Hispanic                                              -37 (66)
Non-Hispanic Black                                   -227 (119)
Non-Hispanic Indian                                   723 (193) (***)
Non-Hispanic Asian                                   -158 (159)
Non-Hispanic Pacific Islander                         -27 (252)
Multiple races (non-Hispanic)                         -11 (75)
Reference is non-Hispanic White
Parent was limited English proficient                 -86 (78)
Medical coverage
Count of Medicaid eligibility months                   98 (14) (***)
Count of third-party liability coverage months        -91 (8) (***)
Child social and health risk factors
Chronic noncomplex health conditions                  468 (68) (***)
Chronic complex health conditions                    3053 (91) (***)
Reference is nonchronic noncomplex conditions
Child has mental health condition                     112 (62)
Child has substance use disorder                     -210 (116)
Child is involved in criminal justice system         -136 (148)
Child is a parent                                    3446 (239) (***)
Child medical risk score (CDPS)                      2869 (24) (***)

(***) P < 0.001; (**) P < 0.01; (*) P < 0.05.

TABLE 5 Estimated overall associations of selected risk factors with
behavioral health system and medical system costs

                                                  Behavioral health
                                                  system
                                                  Estimate  CI

Maltreatment history
Any lifetime physical abuse (excluding prior       159      [89, 232]
year)
Any lifetime sexual abuse (excluding prior year)   238      [135, 335]
Any lifetime neglect (excluding prior year)         99      [-1, 205]
Any abuse or neglect in prior year                1104      [934, 1293]
First out-of-home placement at age 0-3             154      [26, 308]
First out-of-home placement at age 4-11            168      [48, 291]
First out-of-home placement at age 12 or older     698      [410, 1001]
5 or more out-of-home placements                  1777      [1452, 2118]
First abuse or neglect at age 0-3                   -2      [-137, 120]
First abuse or neglect at age 4-11                   1      [-115, 112]
First abuse or neglect at age 12-17                -42      [-180, 84]
Parent risk factors
Parent has mental health condition                  46      [-8, 101]
Parent has a substance use disorder                  3      [-60, 65]
Parent was homeless                                109      [51, 168]
Parent was involved in criminal justice system     -82      [-138, -29]
Parent was disabled                                 91      [29, 154]
Parent was involved in domestic violence           -20      [-75, 44]
Parent used food assistance, not cash             -201      [-274, -129]
Parent used cash assistance                       -299      [-376, -221]
Parent has disabling condition                     -92      [-144, -41]
Parent died                                        197      [49, 357]

                                                  Medical system
                                                  Estimate  CI

Maltreatment history
Any lifetime physical abuse (excluding prior        22      [-114, 170]
year)
Any lifetime sexual abuse (excluding prior year)   302      [87, 567]
Any lifetime neglect (excluding prior year)        213      [37, 422]
Any abuse or neglect in prior year                 390      [161, 633]
First out-of-home placement at age 0-3             335      [109, 575]
First out-of-home placement at age 4-11            -29      [-220, 169]
First out-of-home placement at age 12 or older     -15      [-462, 507]
5 or more out-of-home placements                   624      [177, 1156]
First abuse or neglect at age 0-3                 -193      [-467, 45]
First abuse or neglect at age 4-11                -117      [-312, 60]
First abuse or neglect at age 12-17               -246      [-550, 15]
Parent risk factors
Parent has mental health condition                  17      [-80, 115]
Parent has a substance use disorder               -122      [-239, -3]
Parent was homeless                                -50      [-156, 39]
Parent was involved in criminal justice system       3      [-101, 117]
Parent was disabled                                 47      [-92, 182]
Parent was involved in domestic violence          -160      [-289, -37]
Parent used food assistance, not cash             -161      [-280, -37]
Parent used cash assistance                       -202      [-280, -47]
Parent has disabling condition                     134      [15, 255]
Parent died                                        -71      [-223, 118]
COPYRIGHT 2019 Health Research and Educational Trust
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:RESEARCH ARTICLE
Author:Patton, Deleena A.; Liu, Qinghua; Adelson, Jaimie D.; Lucenko, Barbara A.
Publication:Health Services Research
Article Type:Statistical data
Geographic Code:1U9WA
Date:Feb 1, 2019
Words:8851
Previous Article:An empiric approach to identifying physician peer groups from claims data: An example from breast cancer care.
Next Article:Do racial and ethnic disparities in health care use vary with health?
Topics:

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |