’ Spillover Effects of Adult Medicaid ExpansionsMaya Venkataramani, MD, MPH,a​ Craig on Evan Pollack, Children MD, MHS,a​ Eric T. Roberts,s PhDUseb

BACKGROUND: of Preventive Services abstract ∼ Since the passage of the Affordable Care Act, Medicaid enrollment has increased by 17 million adults, including many low-income parents. One potentially important, ’ but little studied, consequence of expanding health insurance for parents is its effect on METHODS: children s receipt of preventive services. – By using state Medicaid eligibility thresholds linked to the 2001 2013 Medical ’ Expenditure Panel Surveys, we assessed the relationship between changes in adult Medicaid eligibility and children s likelihood of receiving annual well-child visits (WCVs). ’ In instrumental variable analyses, we used these changes in Medicaid eligibility to estimate RESULTS: the relationship between parental enrollment in Medicaid and children s receipt of WCVs. Our analytic sample consisted of 50622 parent-child dyads in families with ’ incomes <200% of the federal poverty level, surveyed from 2001 to 2013. On average, a 10-point increase in a state s parental Medicaid eligibility (measured relative to the Pfederal poverty level) was associated with a 0.27 percentage point higher probability that a child received an annual WCV (95% confidence interval: 0.058 to 0.48 percentage points, = .012). Instrumental variable analyses revealed that parentalP enrollment in Medicaid was associated with a 29 percentage point higher probability that their child received an annual CONCLUSIONS: WCV (95% confidence interval: 11 to 47 percentage points, = .002). In our study, we demonstrate that Medicaid expansions targeted at low-income adults are associated with increased receipt of recommended pediatric preventive care for their children. This finding reveals an important spillover effect of parental insurance coverage that should be considered in future policy decisions surrounding adult Medicaid eligibility.

aDivision of General Internal Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland; and What’sw Kno n on This Subject: Cross-sectional bDepartment of Health Care Policy, Harvard Medical School, Boston, Massachusetts studies have revealed inconsistent relationships between parental insurance coverage and children s Dr Venkataramani conceptualized and designed the study, performed the initial analyses, and ’ drafted the initial manuscript; Drs Pollack and Roberts conceptualized and designed the study health care use. To our knowledge, no studies have and critically reviewed and revised the manuscript; and all authors approved the final manuscript investigated the link between parental insurance as submitted. and pediatric care by using more robust quasi- experimental methods. DOI: https://​doi.​org/​10.​1542/​peds.​2017-​0953 Accepted for publication Aug 8, 2017 What This Study Adds: We found that parental Medicaid enrollment is associated with a 29 Address correspondence to Maya Venkataramani, MD, MPH, Division of General Internal Medicine, percentage point higher probability that low- Johns Hopkins University School of Medicine, Suite 2-502, 2024 E Monument St, Baltimore, MD 21287. E-mail: [email protected] income children received annual well-child visits, highlighting a link between parents’ Medicaid PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). coverage and their children’s health care use. Copyright © 2017 by the American Academy of Pediatrics To cite: Venkataramani M, Pollack CE, Roberts ET. Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services. Pediatrics. 2017;140(6):e20170953

Downloaded from www.aappublications.org/news by guest on September 26, 2021 PEDIATRICS Volume 140, number 6, December 2017:e20170953 Article

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF ’ Since the passage of the Affordable in unobserved ways that may be met inclusion criteria but were Care Act (ACA), 31 states and the related to their child s use of care, pregnant at any point during the District of Columbia have expanded potentially biasing the estimated calendar year. Because pregnancy

Medicaid to adults earning <138%1 relationship19 between these status was not uniformly reported in of the federal poverty level (FPL). variables. all years of the MEPS, we identified Medicaid expansions have been pregnant women as having either a To address this bias, we used a shown to increase preventive care child born in the same calendar year quasi-experimental design that use, improve health outcomes, or by September of the next calendar – leveraged state-level variation in and reduce the financial burden of year, or as having any pregnancy- 2 5 adult Medicaid eligibility over the obtaining care. ‍ The authors of ’ related inpatient visits over the same period from 2001 to 2013 to assess several studies have also identified ’ period. This measure was highly the relationship between parents potential spillover effects in which correlated with pregnancy status for Medicaid coverage and children s increases in adult Medicaid coverage the years in which this variable was receipt of annual WCVs. Our findings indirectly affect health care use in included in the MEPS (Supplemental reveal how changes in federal and other populations, for example, by Table 5). state policies that affect Medicaid – increasing insurance enrollment ’ – coverage for adults could indirectly Consistent with the methods in among children and improving family 16 18 5 7 affect low-income children s use of previous studies,​ ‍ ‍ our primary financial standing. ‍‍ recommended primary care services. analyses were focused on low-income One important but understudied Methods families, defined as those with effect of expanding access to health incomesMeasures <200% of the FPL. Data Sources insurance for parents is its impact Outcome: WCV Use on well-child visit (WCV) use – among children. Recommended annually for children 3 years of age We analyzed data from the 2001 We assessed whether a child received and older, and more frequently for 2013 Medical Expenditure Panel 8 at least 1 WCV in the calendar year. infants and toddlers,​ WCVs serve Surveys (MEPS) linked to state “ ” WCVs were defined as outpatient as the primary platform for growth Medicaid eligibility criteria from “ ” visits for a well-child examination,​ and developmental screening, the Kaiser Family Foundation and “ ” for a general checkup,​ or for vaccination, and provision of county-level characteristics from immunization or shots for children anticipatory guidance. Children who the Area Health Resources File. The 21 <17. We controlled for changes in receive WCVs are more likely to MEPS is a nationally representative the recommended WCV schedule complete immunization schedules survey of the noninstitutionalized over time (such as introduction of and are less likely to have avoidable US population and includes detailed 9,10​ annual visit requirements for 7- and hospitalizations. ‍ WCV use in the information about family structure 9-year-olds in 2007) by including United States has been persistently and demographic characteristics, year fixed effects in regression suboptimal, particularly among racial health insurance20 status, and health 8 care use. State-Levelanalyses. Medicaid Eligibility for and ethnic minorities11 and in low- Parents income families. Our analytic sample consisted of children ages 2 through 17 linked Although the authors of previous to their biological, step, or adoptive studies have shown that parental– We obtained state Medicaid ’ parents living in the same household health care use is correlated with eligibility thresholds from 2000 12 15 during the calendar year. We first ’ children s receipt of care,​ ‍ ‍ through 2012 from the– Kaiser selected children ages 2 through 17 evidence regarding the relationship Family Foundation s surveys of state living in the household the entire 22 32 between parental insurance coverage – Medicaid programs. ‍ ‍ Because year. We then linked each child to a and pediatric care use is mixed. The states may change their Medicaid parent or guardian (age 20 64) living authors of several cross-sectional eligibility limits at different points in ’ in the household for the full year. studies found a positive correlation a year, we used income thresholds ’ Consistent with previous research, between parents insurance status– from the preceding year in all preference was given to the selection and children s health care use in regression analyses. Thresholds 16 18 of the mother if more than 1 parent primarily low-income families,​ ‍‍ 15 were based on eligibility criteria for in the household was surveyed. whereas analyses in14, mixed15​ income unemployed parents with dependent populations did not. ‍ A limitation Because Medicaid eligibility rules children, reflecting (if applicable) the of cross-sectional analyses is that differ for pregnant women, we higher income limits allowed under insured and uninsured adults differ excluded children whose mothers federal expansion waivers, and were Downloaded from www.aappublications.org/news by guest on September 26, 2021 2 Venkataramani et al

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF expressed in percentage points of the models for our full sample of parent- status to determine if our findings 33 ’ FPL. We used linear interpolation child dyads (the unit of analysis) could be explained by changes in to approximate thresholds in years in and stratified by category of family children s health insurance status. which data were not reported (2002, income (<100% of the FPL and 100% Second, because states may also ’ Parental2007, and Enrollment 2010). in Medicaid of the FPL to <200% of the FPL). We undergo concurrent expansions in used these regression analyses to children s Medicaid eligibility over estimate the change in probability time, which could in turn affect We constructed a binary indicator of receiving a WCV because of state child insurance coverage status, we for whether an adult had Medicaid Medicaid expansions for adults, repeated our analyses for a subgroup coverage at any point in the study holding all other covariates at their of 29 large states (with observable Cyear.ovariates sample means. state identifiers in the MEPS) whose ’ income thresholds for child Medicaid We then examined the relationship ≥ or CHIP programs were consistently between a parent s enrollment 200% of the FPL during the study in Medicaid and WCV usage by In our primary analyses, we period. In these states, low-income controlled for parental, family, child, using an instrumental variable (IV) ’ children in our cohort would have analysis. This approach addresses and county-level characteristics– ’ remained consistently eligible for bias from unmeasured factors that that are associated11,16​ with18,21​ children s ’ Medicaid, further isolating the impact could impact a parent s insurance health care use. ‍ ‍‍ Parental of parental Medicaid expansions on status and their children s health characteristics included age, sex, WCVs. race, Hispanic ethnicity, comfort care use by using within-state Third, we limited the study period speaking English, education, and changes in Medicaid eligibility ’ smoking status. Family level variables for parents as an instrument for to 2001 to 2009 to ensure that our were total income, size, and parental parental Medicaid enrollment. We results were not driven by the ACA s structure (mother and father both estimated a binary probit model for introduction of regulations requiring ’ ’ present in household versus not). We the probability that a child received a insurer coverage of preventive additionally controlled for the child s WCV as a function of his/her parent s health care services. Fourth, we ’ ≥ age and sex, the density of physicians Medicaid enrollment status, which reestimated our models for families in the patient s county (total active we instrumented by using within- with incomes 400% of the FPL. MD physicians divided by the county state changes in Medicaid eligibility Because we expected parents in these population), the county-level poverty thresholds for parents. Our IV models families to be relatively unaffected by rate, and an indicator for whether the controlled for state fixed effects, Medicaid expansions, this serves as family lived in an urban area (defined year fixed effects, and the covariates a falsification test (eg, we would not expect to see an impact of increasing Sastatistical a Metropolitan Analyses Statistical Area). described above (see Supplemental ’ Information for details). By using parental Medicaid eligibility on these models, we calculated a children s WCV receipt for this higher ’ predicted change in the probability income sample). We first examined the association that a child would receive a WCV if between the changes in a state s We used family survey weights their parent enrolled in Medicaid, parental Medicaid eligibility provided in the MEPS and variance holding all covariates at the sample thresholds and WCV receipt. estimates that accounted for means. The IV models were run on Specifically, we estimated a clustering of observations at the state our full sample of families and by multivariable logistic regression level. This study was approved by stratum of family income. the institutional review board of the model in which our outcome was ’ the probability that a child received We performed 4 sensitivity analyses. Johns Hopkins School of Medicine. at least 1 WCV during the year by First, because a child s insurance Results using parental Medicaid eligibility status may change in response “ ” thresholds (lagged by 1 year) as the to changes in parental access to 34 ’ main independent variable. In this insurance (eg, welcome mat Our analytic sample consisted of intention-to-treat framework, we effects), and because children s 50622 parent-child dyads across

used within-state changes in the insurance status is known to be the 13 study years, representing

Medicaid eligibility threshold for an independent11, determinant16,​ 18​ of 266557804 weighted pairs through parents as the exposure of interest. health care use,​ ‍ ‍ we ran the study period (Table 1). Slightly ’ ’ The models controlled for state and models additionally controlling for less than half of the dyads (44.7%) year fixed effects, in addition to the the child s Medicaid or Children s had incomes <100% of the FPL. The covariates described above. We ran Health Insurance Program (CHIP) mean age of children was 9.3 years, Downloaded from www.aappublications.org/news by guest on September 26, 2021 PEDIATRICS Volume 140, number 6, December 2017 3

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF TABLE 1 Primary Sample Characteristics % of Na,​b and the mean age of linked parents Child sex was 35.8 years. The majority of Male 51.5 parents were white (72.5%), non- Female 48.5 Hispanic (67.4%), and had earned Child age in y (mean) 9.30 at least a high school diploma or Child age in y, categories general education diploma (67.0%). 2–6 31.8 7 11 32.9 The majority of dyads consisted of – 12–17 35.3 children linked to mothers (95.4%). Parent sex At the beginning of the study period Male 4.6 Female 95.4 (2001), mean adult Medicaid Parent age in y (mean) 35.8 eligibility (weighted for the Parental education population across states in our Less than HS 33.0 sample and lagged by 1 year) for HS diploma or GED 35.6 unemployed parents was 73.5% of Any college 30.5 Not specified 0.9 the FPL, and increased to 107.7% Parental race of the FPL by 2013 (Supplemental White 72.5 Table 6). The proportion of states ≥ Black 19.9 with parental Medicaid eligibility Asian or Pacific Islander 4.7 limits 200% of the FPL increased Other 2.9 Parental ethnicity Hispanic 32.6 from 7.8% to 29.4% over the study Parent comfortable with English language 85.3 period (Supplemental Table 6). The Parental smoking status percentage of children in our sample Nonsmoker 70.0 receiving an annual WCV increased Current smoker 24.3 from 32.7% in 2001 to 47.9% in 2013 Not specified 5.7 Family income (in dollars; mean) 24 593.61 (Table 2). Family income, percentage of the FPL In the intention-to-treat analysis, <100% 44.7 ’ we found that a 10-point absolute 100% to <200% 55.3 Family size (no. of members) increase in a state s adult Medicaid 2 6.8 eligibility threshold (relative to the 3 15.6 FPL) was associated with a 0.27 4 27.6 percentage-point increase in the 5 24.3 probability that low income children 6 13.4 7 or more 12.3 received an annual PWCV (95% Parental structure confidence interval [CI]: 0.058 to 0.48 Single parent 42.5 percentage points, = .012; Table 3). Two parents 57.5 In stratified analyses, changes in GED, general education diploma; HS, high school. adult Medicaid eligibility thresholds a N = 266 557 804 weighted dyads (50 622 unweighted dyads). b were positively associated with Or mean as otherwise specified. WCV use for families with incomes 100% of the FPL to <200% of the P FPL (0.38 percentage points, 95% P 4). This relationship was strongest to 30 percentage points, = .237) CI: 0.10 to 0.66 percentage points, in families with incomes 100% to (Table 4). = .008) and was positively but not significantly related to WCVs in the <200% of the FPL. For these families, there was a 45 percentage point In sensitivity analyses (summarized lowest (<100% of the FPL) incomeP in Supplemental Table 9), controlling − higher probability that a child would group (0.13 percentage points, 95% for child Medicaid and/or CHIP have a WCV if a parent was enrolled CI: 0.11% to 0.36% points, = .284) enrollment did not significantly in Medicaid compared with the (Table 3). P affect our results, revealing that In IV analyses, we found that parental parent not being enrolled (95% CI: parental Medicaid enrollment 17 to 73 percentage points, = .002). ’ Medicaid enrollment was associatedP affects WCV use independently of with a 29 percentage-point (95% The relationship remained positive, children s insurance status. Limiting CI: 11 to 47 percentage points; = but was not statistically significant, in our analysis to large states in which − ≥ .002) increase in the probability that the <100% of the FPL income group child eligibility for Medicaid and their child would have a WCV (Table (11 percentage points, 95% CI: 7.4 CHIP remained 200% through the Downloaded from www.aappublications.org/news by guest on September 26, 2021 4 Venkataramani et al

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF TABLE 2 Proportion of Children Who Received a WCV, Overall and by Year ≥ Year % Children With WCVa P b higher-income families ( 400% 2001 32.7 <.001 of the FPL), demonstrating that 2002 34.3 2003 34.6 the effects of changes in Medicaid 2004 35.9 coverage were concentrated, as 2005 35.8 expected, among lower-income 2006 33.5 families. 2007 35.1 2008 35.8 Discussion 2009 39.5 2010 38.2 2011 41.1 Leveraging 13 years of changes in 2012 40.7 2013 47.9 state Medicaid eligibility for adults All years 37.5 and performing an IV analysis, we a Weighted percentage. found that increases in the income b Pearson’s χ2 test comparing outcome over years. threshold for adult Medicaid eligibility were associated with a greater likelihood that children TABLE 3 The Adjusted Association Between Changes in a State’s Medicaid Eligibility Threshold for Parents and Child’s Receipt of a WCV in low-income families received at least 1 annual WCV. With our Probability (95% CI)a Stratified Analysis by FPL Income Categories results, we provide evidence of an <100% Probability (95% CI)a 100% to <200% independent relationship between Probability (95% CI)a ’ parental Medicaid enrollment and Parental Medicaid 0.27 (0.058 to 0.48) 0.13 ( 0.11 to 0.36) 0.38 (0.10 to 0.66) − children s primary care use in low- eligibility thresholdb income families, and we illustrate the potential for adult Medicaid Results represent the change in predicted probability of a WCV for every 10 percentage-point increase in the state parental ’ eligibility threshold. Estimates are from a multivariable logistic regression of WCV receipt on Medicaid eligibility threshold. expansions to have positive spillover In addition to state and year fixed effects, covariates in the model included child sex, child age (y), parent sex, parent effects on children s health care use. age (y), parent race, parent ethnicity, parent English-speaking status, parental smoking status, parental education level, parental structure of household, family income, family size, county poverty rate, county physician density, and county We found the strongest relationship Metropolitan Statistical Area Status. between adult Medicaid eligibility a Marginal predicted probability expressed in percentage points, calculated at the means of all other variables in model. b In 10 percentage-point increments of FPL. and WCVs in near-poor families (100% to <200% of the FPL). This likely reflects the fact that increases TABLE 4 The Adjusted Association of Parental Medicaid Enrollment With Child’s Receipt of WCV (IV Analysis): Results Represent the Increased Probability of a WCV for a Child Whose Parent is in parental Medicaid eligibility during Enrolled in Medicaid the 2000s primarily affected families with incomes slightly >100% of the Probability (95% CI)a Stratified Analysis by FPL Income Category FPL, whereas states with the least a <100% Probability (95% CI) 100%–<200% Probability generous Medicaid coverage for (95% CI)a nonpregnant adults generally did Parental Medicaid 29 (11 to 47) 11 (−7.4 to 30) 45 (17 to 73) not expand parental eligibility over enrollment the study period. Our analyses were Results represent the increased probability of a WCV for a child whose parent is enrolled in Medicaid. Estimates are from multivariable probit regression of WCV receipt on instrumental parental Medicaid enrollment. In addition to state therefore less able to detect effects and year fixed effects, covariates in the model included child sex, child age (y), parent sex, parent age (y), parent race, of eligibility changes in the lowest- parent ethnicity, parent English-speaking status, parental smoking status, parental education level, parental structure of income families. household, family income, family size, county poverty rate, county physician density, and county Metropolitan Statistical Area Status. Several mechanisms may underlie a Marginal predicted probability expressed in percentage points, calculated at the means of all other variables in model. this spillover effect of parental Medicaid coverage on WCV receipt. One hypothesis, supported by the – Behavioral Model of Health Services 35 ’ study period revealed a positive 125% 200% of the FPL). Limiting Use,​ is that insurance enhances – relationship between parental our analysis to the pre-ACA parents ability to navigate the health Medicaid eligibility thresholds and period (2001 2008) also did not care system for themselves and WCVs for our primary analytic significantly affect our results. for family members. This may lead sample and a significantly positive Finally, we did not find a significant to an increase in parental health- relationship for a subset of this association between parental seeking behaviors for their children sample (families with incomes Medicaid eligibility and WCVs in (eg, scheduling WCVs). It may also Downloaded from www.aappublications.org/news by guest on September 26, 2021 PEDIATRICS Volume 140, number 6, December 2017 5

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF ’

function through a welcome mat the introduction of the ACA s support have raised the concern that effect in which eligible but previously requirement that insurance plans states could curtail Medicaid benefits

uninsured children enroll in Medicaid6,7,​ 34​ cover preventive services for or eligibility, thereby significantly39,40​ after their parents gain coverage. ‍ ‍ children, and other efforts11 to increase reducing parental enrollment. ‍ However, because our estimates pediatric preventive care. We Given the evidence that increased ’ were substantively unchanged incorporated year fixed effects into access to pediatric care early in life– is

after we controlled for children s our models to control for temporal associated with improved health41 and43 Medicaid and/or /State CHIP status, trends that may have resulted from lower hospital use in adulthood,​ ‍ our analyses indicate that such a these changes and verified that our changes in parental coverage may woodwork effect was not primarily results were unchanged when we have long-term impacts on children responsible for changes in WCV use. limited the study period to years that will be important to consider – preceding these ACA mandates when modifying the Medicaid Another potential mechanism is that ’ for pediatric care (2001 2009). program. parental Medicaid coverage may Finally, many measures in the Conclusions improve families financial standing, MEPS, including insurance status freeing up resources to provide and health care use, are self- preventive services for children. reported and subject to recall and In our study, we demonstrate that Low-income families who enroll in social desirability bias. However, parental Medicaid enrollment is public insurance have decreased out- respondents are aware that health associated with increases in pediatric of-pocket medical spending and a 5,36​’ care providers may be contacted to primary care use in low-income reduced likelihood of bankruptcy. ‍ verify self-reported use, which may families. Given the suboptimal rates Studies have revealed that children s 15 mitigate reporting error. of WCV use in low-income families, health care use is sensitive to out- our findings suggest that efforts to of-pocket costs, particularly in low- These findings are of great ’ 37,38​ expand Medicaid for parents may income families. ‍ significance given the current help to promote their children s uncertainty surrounding the future Our study has several limitations. receipt of recommended preventive of the ACA and Medicaid expansions First, our conclusions are not derived care. authorized by the law. Our work from a randomized controlled trial Acknowledgments highlights the potential for Medicaid of Medicaid enrollment; however, expansions targeting low-income we use a quasi-experimental design adults to mitigate disparities in the that leverages plausibly exogenous We thank Ray F. Kuntz, Center for receipt of WCVs between low- and state-level policy changes to isolate Financing, Access and Cost Trends high-income families. Currently, the effect of parental Medicaid Data Center Coordinator at the 19 states have not expanded adult enrollment on WCVs from other Agency for Healthcare Research Medicaid coverage to 138% of the family and person-level determinants ∼ and Quality, for his assistance with FPL under the ACA. According to of this relationship. Second, our data access, and Martin Andersen, Current Population Survey data, 5.5 analyses may not isolate the impact the University of North Carolina at million children in these 19 states of changes in parental Medicaid Greensboro, for his helpful comments live in families in which a parent eligibility and coverage on WCVs if on an earlier draft of this article. would qualify for expanded Medicaid states contemporaneously expanded coverage. Our intention-to-treat Medicaid eligibility or increased estimates imply that the spillover coverage generosity for children. To ∼ effect of Medicaid expansion would address this concern, we conducted result in 135 000 additional annual Abbreviations a sensitivity analysis in which we ’ WCVs for low-income children in limited our analytic sample to states these 19 states (see Supplemental with CHIP or children s Medicaid ’ Information for calculation). ACA: Affordable Care Act eligibility thresholds consistently CHIP: Children s Health >200% of the FPL. In this sensitivity Likewise, our results reveal Insurance Program analysis, we continued to find a the potential for reductions in CI: confidence interval positive relationship between adult Medicaid coverage to have ’ FPL: federal poverty level parental eligibility for Medicaid and unintended spillover effects on IV: instrumental variable WCVs, particularly in near-poor children s health care use. Recent MEPS: Medical Expenditure families. Third, our study period proposals to reform the Medicaid “ ” Panel Survey encompasses several changes in program by using block grants or WCV: well-child visit the recommended WCV schedule, per-capita caps on federal financial Downloaded from www.aappublications.org/news by guest on September 26, 2021 6 Venkataramani et al

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose. FUNDING: Dr Venkataramani is supported by an institutional National Research Service Award (T32HP10025B0). POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose. COMPANION PAPER: A companion to this article can be found online at www.pediatrics.​ ​org/​cgi/​doi/​10.​1542/​peds.​2017-​3236.

References 1. kaiser Family Foundation. Estimates Medicaid beneficiaries. Pediatrics. review. Am J Prev Med. 2016;50 of eligibility for ACA coverage among 2001;108(1):90–97 (5 suppl 1):S27–S33 the uninsured in 2016. 2016. Available 10. Freed GL, Clark SJ, Pathman DE, 20. Cohen S. Sample Design of the 1996 at: http://​kff.​org/​health-​reform/​ Schectman R. Influences on the receipt issue-​brief/​estimates-​of-​eligibility-​for-​ Medical Expenditure Panel Survey of well-child visits in the first two years Household Component. Rockville, MD: aca-​coverage-​among-​the-​uninsured-​ of life. Pediatrics. 1999;103 in-​2016/​view/​footnotes/​. Accessed Agency for Health Care Policy and (4 pt 2):864–869 February 7, 2017 Research; 1996 11. Abdus S, Selden TM. Adherence with 21. Selden TM. Compliance with well-child 2. Sommers BD, Baicker K, Epstein AM. recommended well-child visits has visit recommendations: evidence from Mortality and access to care among grown, but large gaps persist among the Medical Expenditure Panel Survey, adults after state Medicaid expansions. various socioeconomic groups. Health 2000-2002. Pediatrics. 2006;118(6). N Engl J Med. 2012;367(11):1025–1034 Aff (Millwood). 2013;32(3):508 515 – Available at: www.​pediatrics.​org/​cgi/​ 3. Sommers BD, Blendon RJ, Orav EJ, 12. Bates AS, Fitzgerald JF, Dittus content/​full/​118/​6/​e1766 Epstein AM. Changes in utilization RS, Wolinsky FD. Risk factors for and health among low-income adults underimmunization in poor urban 22. Broaddus M, Blaney S, Dude A, Guyer J, Ku L, Peterson J. Expanding Family after Medicaid expansion or expanded infants. JAMA. 1994;272(14):1105–1110 private insurance. JAMA Intern Med. Coverage: States’ Medicaid Eligibility 13. kogan MD, Alexander GR, Jack BW, Policies for Working Families in the 2016;176(10):1501–1509 Allen MC. The association between Year 2000. Washington, DC: Center 4. Gross T, Notowidigdo M. Health adequacy of prenatal care utilization on Budget and Policy Priorities; insurance and the consumer and subsequent pediatric care 2001. Available at: www.​cbpp.or​ g/​ bankruptcy decision: evidence from utilization in the United States. archiveSite/​1-​2-​02health.​pdf expansions of Medicaid. J Public Econ. Pediatrics. 1998;102(1 pt 1):25–30 23. kaiser Commission on Medicaid and 2011;95(7):767–778 14. Hanson KL. Is insurance for children the Uninsured. Enrolling children enough? The link between parents’ 5. Wherry LR, Kenney GM, Sommers BD. and families in health coverage: the and children’s health care use The role of public health insurance in promise of doing more. 2002. Available revisited. Inquiry. 1998;35(3):294–302 reducing child poverty. Acad Pediatr. at: https://​kaiserfamilyfound​ation.​ 2016;16(suppl 3):S98–S104 15. Goedken AM, Urmie JM, Polgreen LA. files.​wordpress.​com/​2002/​05/​4046.​pdf. Factors related to receipt of well-child Accessed February 7, 2017 6. Dubay L, Kenney G. Expanding public visits in insured children. Matern Child health insurance to parents: effects on Health J. 2014;18(3):744–754 24. kaiser Commission on Medicaid and children’s coverage under Medicaid. the Uninsured. Preserving recent 16. Davidoff A, Dubay L, Kenney G, Health Serv Res. 2003;38(5):1283–1301 progress on health coverage for Yemane A. The effect of parents ’ children and families: new tensions 7. DeVoe JE, Tillotson CJ, Angier H, insurance coverage on access to emerge. A 50 state update on eligibility, Wallace LS. Predictors of children s care for low-income children. Inquiry. ’ enrollment, renewal and cost- health insurance coverage 2003;40(3):254 268 – sharing practices in Medicaid and discontinuity in 1998 versus 2009: 17. Gifford EJ, Weech-Maldonado R, Short SCHIP. 2003. Available at: www.​cbpp.​ parental coverage continuity plays a PF. Low-income children’s preventive org/​archiveSite/​7-​30-​03health.​pdf. major role. Matern Child Health services use: implications of parents’ Accessed February 7, 2017 J. 2015;19(4):889 896 – Medicaid status. Health Care Financ 25. kaiser Commission on Medicaid and 8. Hagan J, Shaw J, Duncan P, eds. Rev. 2005;26(4):81–94 the Uninsured. Beneath the surface: Bright Futures: Guidelines for Health 18. DeVoe JE, Tillotson CJ, Wallace barriers threaten to slow progress on Supervision of Infants, Children, and LS. Children’s receipt of health expanding health coverage of children Adolescents. 3rd ed. Elk Grove Village, care services and family health and families: a 50 state update on IL: American Academy of Pediatrics; insurance patterns. Ann Fam Med. eligibility, enrollment, renewal and 2008 2009;7(5):406–413 cost-sharing practices in Medicaid and 9. Hakim RB, Bye BV. Effectiveness 19. Wallace J, Sommers BD. Health SCHIP. 2004. Available at: www.​cbpp.​ of compliance with pediatric insurance effects on preventive org/​archiveSite/​10-​4-​04health.​pdf. preventive care guidelines among care and health: a methodologic Accessed February 7, 2017

Downloaded from www.aappublications.org/news by guest on September 26, 2021 PEDIATRICS Volume 140, number 6, December 2017 7

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF 26. kaiser Commission on Medicaid and wordpress.​com/​2013/​01/​7855.​pdf. 36. McMorrow S, Kenney GM, Long SK, Goin the Uninsured. In a time of growing Accessed February 7, 2017 DE. Medicaid expansions from 1997 to need: state choices influence health 30. kaiser Commission on Medicaid and 2009 increased coverage and improved coverage access for children and the Uninsured. Holding steady, looking access and mental health outcomes families: a 50 state update on ahead: annual findings of a 50-state for low-income parents. Health Serv eligibility rules, enrollment and survey of eligibility rules, enrollment Res. 2016;51(4):1347–1367 renewal procedures, and cost-sharing and renewal procedures, and cost 37. Children's Health Fund. Unfinished practices in Medicaid and SCHIP for sharing practices in Medicaid and business: more than 20 million children and families. 2005. Available SCHIP, 2010-2011. 2011. Available at: children in U.S. still lack sufficient at: https://​kaiserfamilyfound​ation.​files.​ https://​kaiserfamilyfound​ation.​files.​ access to essential health wordpress.​com/​2013/​01/​in-​a-​time-​of-​ wordpress.​com/​2013/​01/​8130.​pdf. care. Available at: https://​www.​ growing-​need-​state-​choices-​influence-​ Accessed February 7, 2017 childrenshealthfu​nd.​org/​wp-​content/​ health-​coverage-​access-​for-​children-​ uploads/​2016/​11/​Unfinished-​Business-​ 31. k aiser Commission on Medicaid and-​families-​report.​pdf. Accessed Final_.​pdf. Accessed February 7, 2017 February 7, 2017 and the Uninsured. Performing under pressure: annual findings 38. Medicaid and CHIP Payment and 27. kaiser Commission on Medicaid of a 50-state survey of eligibility, Access Commission (MACPAC). The and the Uninsured. Resuming the enrollment, renewal and cost- effect of premiums and cost-sharing path to health coverage for children sharing policies in Medicaid and on access and outcomes for low- and parents: a 50 state update on CHIP, 2011-2012. 2012. Available at: income children. 2015. Available at: eligibility rules, enrollment and https://kaiserfamilyfound​ ation.​ files.​ ​ https://​www.​macpac.​gov/​wp-​content/​ renewal procedures, and cost- wordpress.com/​ 2013/​ 01/​ 8272.​ pdf.​ uploads/​2015/​07/​Effect-​of-​Cost-​ sharing practices in Medicaid and Accessed February 7, 2017 Sharing-​on-​Low-​Income-​Children.​pdf. SCHIP in 2006. 2007. Available at: Accessed February 7, 2017 https://​kaiserfamilyfound​ation.​files.​ 32. kaiser Commission on Medicaid and 39. Chatterjee P, Sommers BD. The wordpress.​com/​2013/​01/​7608.​pdf. the Uninsured. Getting into gear for economics of Medicaid reform and block Accessed February 7, 2017 2014: findings from a 50-state survey of eligibility, enrollment, renewal and grants.JAMA. 2017;317(10):1007–1008 28. kaiser Commission on Medicaid and cost-sharing policies in Medicaid and 40. Goodman-Bacon AJ, Nikpay SS. Per the Uninsured. Health coverage for CHIP, 2012-2013. 2013. Available at: capita caps in Medicaid - lessons children and families in Medicaid and https://​kaiserfamilyfound​ation.​files.​ from the past. N Engl J Med. SCHIP: state efforts face new hurdles wordpress.​com/​2013/​05/​8401.​pdf. 2017;376(11):1005–1007 a 50-state update on eligibility rules, Accessed February 7, 2017 enrollment and renewal procedures, 41. Currie J, Decker S, Lin W. Has public and cost-sharing practices in Medicaid 33. Rober ts ET, Gaskin DJ. Projecting health insurance for older children and SCHIP in 2008. 2008. Available at: primary care use in the Medicaid reduced disparities in access to care https://​kaiserfamilyfound​ation.​files.​ expansion population: evidence for and health outcomes? J Health Econ. wordpress.​com/​2013/​01/​7740.​pdf. providers and policy makers. Med Care 2008;27(6):1567–1581 Res Rev. 2015;72(5):515 561 Accessed February 7, 2017 – 42. Boudreaux MH, Golberstein E, McAlpine 34. DeVoe JE, Marino M, Angier H, et al. 29. kaiser Commission on Medicaid and DD. The long-term impacts of Medicaid Effect of expanding medicaid the Uninsured. Challenges of providing exposure in early childhood: evidence for parents on children s health health coverage for children and ’ from the program’s origin. J Health insurance coverage: lessons from the parents in a recession: a 50 state Econ. 2016;45:161–175 Oregon experiment. JAMA Pediatr. update on eligibility rules, enrollment 43. Miller S, Wherry LR. The long-term 2015;169(1):e143145 and renewal procedures, and cost- effects of early life Medicaid coverage. sharing practices in Medicaid and 35. Aday LA, Andersen R. A framework for 2016. Available at: http://​dx.​doi.​org/​10.​ SCHIP in 2009. 2009. Available at: the study of access to medical care. 2139/​ssrn.​2466691. Accessed February https://​kaiserfamilyfound​ation.​files.​ Health Serv Res. 1974;9(3):208–220 7, 2017

Downloaded from www.aappublications.org/news by guest on September 26, 2021 8 Venkataramani et al

Venkataramani et al https://doi.org/10.1542/peds.2017-0953 December 2017 Spillover Effects of Adult Medicaid Expansions on Children’s Use of Preventive Services 6 140 Pediatrics 2017 ROUGH GALLEY PROOF Spillover Effects of Adult Medicaid Expansions on Children's Use of Preventive Services Maya Venkataramani, Craig Evan Pollack and Eric T. Roberts Pediatrics 2017;140; DOI: 10.1542/peds.2017-0953 originally published online November 13, 2017;

Updated Information & including high resolution figures, can be found at: Services http://pediatrics.aappublications.org/content/140/6/e20170953 References This article cites 26 articles, 5 of which you can access for free at: http://pediatrics.aappublications.org/content/140/6/e20170953#BIBL Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): Advocacy http://www.aappublications.org/cgi/collection/advocacy_sub Federal Policy http://www.aappublications.org/cgi/collection/federal_policy_sub Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.aappublications.org/site/misc/Permissions.xhtml Reprints Information about ordering reprints can be found online: http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news by guest on September 26, 2021 Spillover Effects of Adult Medicaid Expansions on Children's Use of Preventive Services Maya Venkataramani, Craig Evan Pollack and Eric T. Roberts Pediatrics 2017;140; DOI: 10.1542/peds.2017-0953 originally published online November 13, 2017;

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://pediatrics.aappublications.org/content/140/6/e20170953

Data Supplement at: http://pediatrics.aappublications.org/content/suppl/2017/11/09/peds.2017-0953.DCSupplemental

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. Pediatrics is owned, published, and trademarked by the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2017 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news by guest on September 26, 2021