COLLEGE OF PUBLIC HEALTH AND HUMAN SCIENCES School of Social and Behavioral Health Sciences EFFECT OF THE ACA ON INSURANCE FOR CHILDREN WITH PRE-EXISTING CONDITIONS DIANA GOVIER, MPH JANGHO YOON, PHD JUNE 24 TH , 2017
Agenda • Background • Specific Aims • Methods • Results • Discussion • Future Directions • Limitations
Chronic conditions in U.S. children • Chronic condition (CC): condition lasting 3 mo.-lifetime 1 • ≈ ¼ US children have CCs 1 • 5 most frequently reported CCs 1 • Asthma (9%), learning disabilities (7.8%), ADD/ADHD (6.4%), speech problems (3.7%), oppositional defiant/conduct disorder (3/3%) • Vary by age, race/ethnicity, income, place of residence 1 • Childhood CCs are on the rise 2 • More likely insured than healthy counterparts 3
Why should we care? • Pre-existing conditions (PECs) insurance exclusions 4 • asthma; diabetes; cancers; mental health conditions; illnesses that put children at greater risk for needing healthcare 5 • >20,000 denied individual coverage due to PECs 6 • More likely underinsured than healthy counterparts 7 • Increased unmet medical needs, delays in care, use of ER, hospitalizations, co-morbidities, mortality, financial hardship • Decreased medication management, usual source of care, preventative services 8-13
ACA policy change Guaranteed issue : prohibition of exclusions and/or differential pricing/benefits based on PECs, health status, claims history, etc 14 • Estimated 31,000 - 72,000 uninsured children would be able to gain coverage who were previously denied 15 • Estimated 90,000 insured children would gain coverage for PECs previously excluded 15
Specific aims Aim I: Examine national trends in insurance for children 5-18 w/ and w/o PECs for years 2008 through 2013. Aim II: Examine effect of guaranteed issue on likelihood of having any insurance, public insurance, and private insurance for children 5-18 w/PECs
Methods – Data & study sample • 5 overlapping panels of Medical Expenditure Panel Survey (MEPS) years 2008 – 2013 • 2011 lag year • Oversampling of racial & ethnicity minorities • Oversampling of low-income individuals • Study sample: 18,701 children 5-18 yrs • No exclusion criteria other than age
Methods – Dependent variables • Any insurance (model 1) – binary variable indicating whether child has any type of insurance • Public insurance (model 2) – binary variable indicating whether child has any public insurance • Private insurance (model 3) – binary variable indicating if child has any private insurance
Methods – Independent variables • Policy implementation (Post) – binary variable indicating time period • Policy period on or after October 1 st , 2010 • Pre-existing condition (PEC) – binary variable indicating if child has PEC • Defined as having asthma, ADD/ADHD, or special healthcare needs • Policy effect (PEC*Post) – interaction term indicating post-policy time period and presence of PEC • Time fixed-effects (T) – month dummy variables for months Jan, 2008-Dec, 2013 to control for secular trends
Methods – Model specification • Difference-in-differences (DID) • Compares ∆ in treated to ∆ in controls as result of policy change 16 • Mimics experimental research design using observational study data 16 • Reduces confounding from unobserved factors 16 • Model specification: logit 𝑄𝑠 ( HI it =1 ) = β 0 β 1 PEC it β 2 T i β 3 Post i *PEC it X1 i β + X2 it β + ε it • Covariates • X1 i β = time in -varying covariates: race/ethnicity, sex • X2 it β = time -varying covariates: age, poverty category, US Census region, MSA • Average marginal effects (AME) to estimate effect of policy on children w/PECs
PEC prevalence increasing over time Data Source: MEPS HC, 2008-2013 Adjusted for complex survey design effect
Insured rate among children increased during study period
Publicly insured rate continued to increase after ACA implementation
Privately insured rate decreased before ACA then stabilized for children w/PECs
Uninsured rated decreased before and after ACA implementation
ACA resulted in significant increase in probability of insurance for children w/PECs Average marginal effect of guaranteed issue on probability of any insurance for chidren w/PECs Pr(any insurance) Avg. Marginal Effect 0.96** (0.32) Note: Boostrapped standard errors w/300 repetitions Note: Statistically significant covariates: month, poverty category, age, sex, race, ethnicity, US Census region, metropolitan statistical area * Statistically significant at the 95% level. ** Statistically significant at the 99% level. *** Statistically significant at the 99.9% level. • By ≈ 1.04% • = %pt change from policy / pre-ACA % children insured w/PECs
ACA resulted in sign. increase in probability of pub/priv insurance for children w/PECs Average marginal effect of guaranteed issue on probability of insurance for chidren w/PECs Pr(pub insurance) Pr(priv insurance) Avg. Marginal Effect 1.93* 1.11 (0.83) (0.86) Note: Boostrapped standard errors w/300 repetitions Note: Statistically significant covariates: month, poverty category, age, sex, race, ethnicity, US Census region, metropolitan statistical area * Statistically significant at the 95% level. ** Statistically significant at the 99% level. *** Statistically significant at the 99.9% level. • By ≈ 4.70% for pub ins & ≈ 2.02% for priv ins (not stat. sigificant.) • = % change from policy / pre-ACA % children insured w/PECs
Discussion • From 2008-2013 decrease in uninsurance for children 5-18 • Unexpected results for effect of guaranteed issue on public insurance • Protections for children w/PECs could improve further • Grandfathered plans are exempt from guaranteed issue • Previously 45-55% children w/special HC needs/disability on grandfathered plans 6 • American Health Care Act (AHCA) makes future uncertain • Could change insurance regulations for children w/PECs • Allow insurers in states with community rating waivers to vary premiums based on PEC-status for those with lapses in coverage >=63 days
Future Directions • Guaranteed issue increased access to insurance for children w/PECs, but important aspects uninvestigated • More research needed to comprehensively discuss best policy-practices for insurance for children w/PECs • Health insurance coverage shown to increase access to & utilization of healthcare services for children w/special healthcare needs 18 • Did guaranteed issue affect healthcare utilization? • Did guaranteed issue affect healthcare expenditures? • How do these effects vary by region, race/ethnicity, and family income? • How has continuity of coverage changed for children w/PECs?
Limitations • Definition of PECs limited in detail • Geographic indicators only include Census region • Does not account for condition-specific carve-outs prohibited by ACA’s guaranteed issue provision
Recommend
More recommend
