Motivation Mental Health Disorders in Childhood: Assessing the Burden on Families Susan Busch Colleen Barry Yale University We gratefully acknowledge support from the RWJ Foundation through the HCFO initiative (Grant no 56465). Research Question Do families of children with mental health disorders experience a greater burden than families with general health conditions? – Financial outcomes (OOP spending) – Labor Market (participation, hours) – Parent time devoted to child’ child’s health care needs (arranging or providing care) Differential burden (cont.) Fragmented MH system may require more time arranging care (New Freedom Commission Report) MH diagnosis may be viewed as more subjective – Reduces acceptability of time away from work Stigma – May reduce inin-kind support from family members In 2001 approximately 13 percent of US children had special health care needs (van Dyck, Dyck, 2004) Caring for a sick child places a burden on families – The full cost of Special Health Care Needs may not be covered by health insurance – ReRe-allocation of parent time to sick child – Labor market participation/hours may increase or decrease (Salkever (Salkever,, 1982; Powers, 2002) Little evidence on differential burden by mental/general health diagnoses Physical and mental health conditions may create differential burden Private health insurance covers mental health conditions less generously – Lack of parity in benefit design Children with MH disorders may have less predictable needs (Gould, 2004) – Difficulty planning for expenses or scheduling time away from work Data 20002000-2002 SLAITS National Survey of CSHCN Large nationally representative sample of children (N=38,856) – Limited to children whose parents report the child has more health care needs than average, and that the condition is expected to last at least 12 months – All SHCN children Single crosscross-section 1 Analytic Approach Methods Concern that children with mental disorders may be different than children with general health disorders on other dimensions – Use propensity score methods to model the effect of mental versus general health care need on family burden – Use propensity score (rather than regression) due to concern that some data is ‘off support’ support’ Run logistic regression to estimate propensity to need mental health care Use these coefficients to predict the propensity to need mental health care for each child (yhat (yhat)) Match children with need for mental health care with SHCN child with no need for mental health care (by propensity score) Compare on outcome measures Measures Outcome Measures Mental health – Did child need mental health care in past 12 months? Financial – OutOut-ofof-pocket spending > $500 during past 12 months – Whether a family reported : that a child’ child’s health care has caused financial problems needing additional income for a child’ child’s medical expenses Labor Market – Any family member quit work to care for child? – Any family member reduce work hours to care for child? Time burden – Any family member spend more than 4 hours per week arranging care? – Any family member spend more than 4 hours providing care? Descriptive Statistics (Unadjusted) CSHCN needing mental health care CSHCN not needing mental health care Demographics – Age, gender, race, language, mother’ mother’s education, family income, number of adults in the household Insurance coverage (private, public, uninsured) Disease severity – Parents ranking of child’ child’s disability (1(1-10) – Parent report of time the child is affected by the condition (never, sometimes, usually, always) State fixed effects – Account for statestate-level differences in labor market opportunities or public benefits Unadjusted Outcomes (cont.) CSHCN needing mental health care CSHCN not needing mental health care Cut work hours 37 % 22 % Quit work 17 % 10 % Children with private coverage OOP spending >$500 42 % 26 % Child’ Child’s HC caused financial problems 30 15 Family needed add’ add’l inc to care for child 25 12 Children with public coverage OOP spending >$500 Labor Market Time Burden 12 % 10 % Child’ Child’s HC caused financial problems 29 22 >4 hrs/wk providing care 19 % 15 % Family needed add’ add’l inc to care for child 23 19 >4 hrs/wk arranging care 17 % 10 % 2 Adjusted Financial Outcomes Adjusted Labor Market Outcomes 45 42 40 40 Private Coverage Public Coverage 35.8 35 35 30 30 *** 29 28 25 20 30 27 26.5 23 20 23 22 *** Mental Health Care Need No Mental Health Care Need 17 *** 25 *** Mental Health Care Need 20 15 11 12 No Mental Health Care Need 16.2 10 15 12.1 *** 5 10 0 Out-ofpocket spending greater than $500 Child's health care has caused financial problems Family Out-ofneeded pocket additional spending income to greater than care for child $500 Child's health care has caused financial problems Fam ily needed additional income to care for child 5 0 Family members h ave cut w ork Family members h ave stopped hours to care for child w orking due to child’s h ealth Adjusted Time Burden Limitations 25 20.1 20 18 16.2 15 12.2 *** Mental Health Care Need No Mental Health Care Need Making causal inferences is difficult with cross sectional data Measures are selfself-report No information on specific diagnosis 10 5 0 Spends greater than 4 hours providing care Spends greater than 4 hours arranging care Conclusion/Implications On average, families of children with mental disorders bear a greater burden Design of public policies – Children’ Children’s SSI eligibility for those with mental health condition has changed over time – Multiple outcomes considered Ideal program would mitigate all these concerns Just providing services will not alleviate all the burden – argues for other types of assistance Parity – Requiring equivalent insurance coverage currently being debated in US Congress 3