Second-Generation Antipsychotic Use Among Stimulant-Using Children, by Organization of Medicaid Mental Health
Abstract
Objective:
Reducing overuse of second-generation antipsychotics among Medicaid-enrolled children is a national priority, yet little is known about how service organization affects use. This study compared differences in second-generation antipsychotic utilization among Medicaid-enrolled children across fee-for-service, integrated managed care, and managed behavioral health carve-out organizational structures.
Methods:
Organizational structures of Medicaid programs in 82 diverse counties in 34 states were categorized and linked to child-level cross-sectional claims data from the Medicaid Analytic Extract covering fiscal years 2004, 2006, and 2008. To approximate the population at risk of antipsychotic treatment, the sample was restricted to stimulant-using children ages three to 18 (N=419,226). The sample was stratified by Medicaid eligibility group, and logistic regression models were estimated for probability of second-generation antipsychotic use. Models included indicators of county-level organizational structure as main predictors, with sequential adjustment for personal and county-level covariates.
Results:
With adjustment for person-level covariates, second-generation antipsychotic use was 31% higher among youths in foster care in fee-for-service counties than for youths in counties with carve-outs (odds ratio [OR]=1.69, 95% confidence interval [CI]=1.26–2.27). Foster care youths in integrated counties had the second highest adjusted odds (OR=1.31, CI=1.08–1.58). Similar patterns of use also were found for youths eligible for Supplemental Security Income but not for those eligible for Temporary Assistance for Needy Families. Differences persisted after adjustment for county-level characteristics.
Conclusions:
Carve-outs, versus other arrangements, were associated with lower second-generation antipsychotic use. Future research should explore carve-out features (for example, tighter management of inpatient or restricted access, as well as care coordination) contributing to lower second-generation antipsychotic use.
Use of second-generation antipsychotics substantially contributes to rising expenditures for Medicaid-enrolled children (1–4). Although primarily indicated for treatment of bipolar disorder and schizophrenia, most second-generation antipsychotic use among children is for off-label treatment of other mood and behavioral disorders (1,5). Second-generation antipsychotic use is controversial because efficacy has not been established for these off-label treatments and because there are side effects, including metabolic disorders (6) and elevated risk of cardiovascular disease and diabetes (7–9).
In recent years, second-generation antipsychotic use has grown considerably across the United States, but rates of increase and levels of use vary widely across states and regions (10,11). One possible influence on diverging utilization rates may be differences in the organization and delivery of Medicaid mental health services across localities. Some programs reimburse providers on a fee-for-service basis (that is, based on a defined fee schedule). Others use managed care—either integrated plans, in which a single plan receives a capitated payment to cover both general medical and mental health benefits, or carve-outs, in which mental health benefits are paid for with separate capitated payments that cover services delivered by managed behavioral health organizations or local mental health authorities (12). Figure 1 illustrates these three structures.
Figure 1 County-level behavioral health organizational structures
aIntegrated managed care: a county Medicaid program administers a single per-member-per-month (PMPM) payment to a managed care organization (MCO). The MCO contracts with individual providers.
bThe Medicaid program administers separate PMPM payments to an MCO and a managed behavioral health organization (MBHO). By definition, plans and providers in counties with integrated and carve-out structures receive payment on a capitated basis.
cThe county directly contracts with providers or uses an administrative services organization (ASO) to manage providers, who are then paid on a fee-for-service basis.
Our objective with this study was to examine differences in second-generation antipsychotic use across Medicaid organizational structures. We hypothesized that second-generation antipsychotic use would be lower under a capitation structure than a fee-for-service structure because theoretical and empirical research has found that capitated plans generally reduce the use of high-cost psychotropic medications (13,14). This reduction can operate either indirectly through selective contracting by Medicaid programs with cost-conscious managed care organizations (15) or more directly through financial incentives for individual providers (13).
It is unclear, however, which type of capitated arrangement—integrated or carved out—might have a larger impact on second-generation antipsychotic use. On the one hand, integrated programs may lower second-generation antipsychotic use through greater coordination between primary care and specialty mental health providers, which could foster greater use of alternative therapies or reduce the use of hospitalizations where antipsychotic use is often initiated (16,17). On the other hand, carve-out systems usually rely on only one specialty managed behavioral health provider, which, compared with other providers, may possess greater case management and utilization review capabilities as well as a less fragmented system in which to target interventions toward medication management (18). Although more specialty service coordination might increase access to some services (19), it could also reduce perceived overuse of treatments such as second-generation antipsychotics.
Methods
Classifying behavioral health organizational structures
We conducted a comprehensive review of Medicaid policies in 82 counties in 34 states covering the period 2004–2008. Counties were the unit of analysis because managed care arrangements are often established at the county level and because there is county-level heterogeneity within large states such as Texas. The predominantly urban counties in this study represented different Medicaid systems and provider environments (Table 1). Study counties were part of a prior national study of Medicaid-enrolled children (20). We began with 97 counties but excluded 12 because they were not contained in our child-level Medicaid data; we excluded three additional counties because of poor data quality for psychotropic prescriptions (defined as rates below 1% of all children within a county). Inclusion of these three counties did not alter our main findings, however.
| Counties (%) | |||
|---|---|---|---|
| Carve-out | Fee for service | Integrated | |
| Characteristic | (N=25 counties) | (N=36 counties) | (N=21 counties) |
| Capitated or integrated pharmacyb | 35 | 28 | 38 |
| Large metropolitan area | 57 | 42 | 43 |
| Small metropolitan area | 35 | 39 | 29 |
| Nonmetropolitan area | 9 | 19 | 29 |
| High pediatrician density (>20 per 100,000 pop.) | 35 | 39 | 33 |
| High psychiatrist density (>15 per 100,000 pop.) | 26 | 31 | 29 |
| High poverty rate (>15%)b,c | 30 | 47 | 57 |
| High unemployment (>7%)b | 13 | 19 | 48 |
| High uninsured (>18%)b,d | 57 | 39 | 19 |
| Majority voted Democratb | 52 | 42 | 48 |
a
Capitated or integrated pharmacy was determined via review of policy documents. Other measures were based on analysis of data from the Area Resource File for years 2002–2008. Data represent the average within each category across years and reflect the closest calendar year for measures such as physician supply that are not reported annually.
b
p<.05, carve-out versus fee for service, by pairwise t test
c
p<.05, carve-out versus integrated, by pairwise t test
d
p<.001, carve-out versus integrated, by pairwise t test
To measure Medicaid organizational structures, we reviewed policy documents from state Medicaid programs, managed care providers, and county agencies. We supplemented our review with government and research reports. Coding was conducted independently by two team members using common criteria. Initial intercoder agreement was attained in 66 cases (80%). Ambiguous cases were resolved with consultation with key informants (academic researchers and clinicians) in states where study counties resided. Study approval was granted by the University of Pennsylvania Institutional Review Board.
Of note, some counties with capitated structures allowed high-need populations to remain in fee-for-service arrangements during the study period. We defined the predominant organizational structures at the county level rather than the child level, both because we did not have a comprehensive measure of which counties allow specific groups of children to remain in fee-for-service plans in otherwise capitated counties and because such children are likely not comparable with those in the general population in counties with predominantly fee-for-service plans.
We also examined whether pharmacy benefits were capitated versus reimbursed as fee for service (defined independently of the behavioral health plan structure). Capitated arrangements include drugs in a per-member-per-month payment to a managed care organization or mean that pharmacy is handled by a third-party pharmacy benefits manager.
Medicaid Analytic Extract
We linked our county-level measures to cross-sectional, child-level data from the Medicaid Analytic Extract (MAX) national claims database, covering the 2004, 2006, and 2008 fiscal years, during which use of second-generation antipsychotic use escalated among Medicaid-enrolled youths (10). Demographic, eligibility, and pharmacy data were extracted from the personal summary and pharmacy files.
Our sample was restricted to children ages three to 18 years with at least ten of 12 months of enrollment in the year. Among these children, we focused on those with any stimulant utilization in the year (N=419,226), a large population of children for whom the adjunctive use of antipsychotics is controversial (21). These youths constituted 3.8% of the total sample with ten months enrollment. Many of these children were being treated for disruptive behaviors related to a diagnosis of attention-deficit hyperactivity disorder (22). Because diagnostic classifications in claims data are likely to be unreliable, the identification of stimulant-using children allowed us to best approximate an at-risk population, and furthermore, select for a group in whom the possibility of missing prescription data across counties was less likely; children with recorded stimulant use would be more likely to have complete utilization histories for other psychotropic drugs, given that we had identified their stimulant use.
Child-level covariates
We classified children into eligibility groups by using the following hierarchy: if children had any enrollment in foster care, they were categorized as foster care; if they had any Supplemental Security Income (SSI) enrollment, they were categorized in the SSI group; any other children were categorized in the Temporary Assistance for Needy Families (TANF)/other group. Foster care and SSI accounted for many of the highest service-using children in the Medicaid program, and TANF was the largest and most heterogeneous group.
Age was categorized within calendar years as three to five years, six to 11 years, and 12 to 18 years. Race-ethnicity was coded as white, black or African American, Latino, or other. Children with race classified as unknown were excluded from the analysis (6.8% of the stimulant-using child population). Five counties had >10% of the child population with race unknown; sensitivity analyses for these counties showed equivalent trajectories of psychotropic use over the study period.
Contextual covariates
Using the Area Resource File, we selected county-level covariates representing contextual factors that could influence use of second-generation antipsychotics and mental health services more broadly, independent of county organizational structure. These included the county supply of providers (pediatrician-to-population and psychiatrist-to-population ratios [23,24]), the urbanicity of the county (core urban, smaller urban, and suburban or rural [25]), and sociodemographic characteristics (the poverty rate and uninsured rate [20]). We also included the percentage of voters in the county voting for the Democratic candidate in the 2008 Presidential election, a proxy for public attitudes influencing social service regulation and provision (26,27).
Statistical analysis
We first calculated the unadjusted rates of second-generation antipsychotic use across the organizational structures among sample children by eligibility group. In multivariable analysis, we used logistic regression models, stratified by eligibility group, to compare the odds of second-generation antipsychotic use across county-level organizational structures. Our first set of models adjusted for indicators for pharmacy structure, calendar year, and individual-level covariates. To assess whether these differences were attenuated by other features of county environment, our second set of models further adjusted for county-level contextual characteristics. A robust standard error estimator was used to accommodate clustering of children within counties. Using predicted margins, we calculated regression-adjusted probabilities of second-generation antipsychotic use within each eligibility group and organizational structure, setting the other covariates at the sample means. All analyses were conducted with Stata statistical software, version 12 (28).
Results
Covariate differences by organizational structure
As shown in Table 1, a greater proportion of counties with carve-out systems were located in large metropolitan areas, and a greater proportion of counties with integrated systems were located in nonmetropolitan areas. A higher proportion of counties with integrated systems had high poverty rates (57.1% compared with 30.4% of counties with carve-outs), and a higher proportion of counties with integrated Medicaid had elevated unemployment rates (47.6% versus 13.0% of counties with carve-outs). By contrast a higher proportion of counties with carve-outs had higher uninsured rates. A lower proportion of counties using fee for service had a pharmacy that was capitated or integrated with behavioral health services, as opposed to being paid out separately.
Table 2 shows demographic characteristics of stimulant-using children in counties by organizational structure. Within counties with fee-for-service Medicaid plans, children in the TANF eligibility group comprised 60.0% of all sample children, compared with 53.3% of children in counties with carve-out plans and 56.3% of children in counties with integrated Medicaid plans. At least half of all children in each organizational structure were in the six- to 11-year-old age group, with the largest percentage in counties using fee for service (52.6%). Counties using fee for service also had the largest percentage of children younger than five (5.1%). Counties with carve-out and fee-for-service systems had similar racial-ethnic distributions, but counties with integrated Medicaid had more white and black children and substantially fewer Latino children (3.0% compared with about 25% in other counties).
| Carve-out | Fee for service | Integrated | |
|---|---|---|---|
| Characteristic | (N=152,750) | (N=172,124) | (N=94,352) |
| Eligibility category | |||
| Foster careb | 28.3 | 15.0 | 21.0 |
| Supplemental Security Incomec | 18.4 | 24.9 | 22.7 |
| Temporary Assistance for Needy Families | 53.3 | 60.0 | 56.3 |
| Male | 74.0 | 71.9 | 72.7 |
| Age group (years) | |||
| 3–5b,d | 3.3 | 5.1 | 2.8 |
| 6–11 | 50.3 | 52.6 | 50.3 |
| 12–18 | 46.4 | 42.3 | 46.9 |
| Race-ethnicity | |||
| White | 42.1 | 41.6 | 52.0 |
| Black | 31.2 | 34.1 | 43.4 |
| Latinoe | 25.0 | 23.4 | 3.0 |
| Otherc | 1.8 | .9 | 1.6 |
a
Authors’ analysis of Medicaid Analytic Extract data, for fiscal years 2004, 2006, and 2008. Values are expressed as percentages. The study sample comprised youths ages three to 18 with any stimulant prescription and ten months of Medicaid enrollment in one of the study counties.
b
p<.001, carve-out versus fee for service, by pairwise t test
c
p<.05, carve-out versus fee for service, by pairwise t test
d
p<.05, carve-out versus integrated, by pairwise t test
e
p<.001, carve-out versus integrated, by pairwise t test
Regression estimates
Table 3 presents odds ratios (ORs) for second-generation antipsychotic use by organizational structure for each eligibility group. Child-level estimates were adjusted for demographic characteristics, study year, and pharmacy structure. Compared with children in carve-out counties, children in foster care with prescriptions for stimulants in fee-for-service counties had the highest adjusted odds of second-generation antipsychotic use (OR=1.69), and children in counties with integrated Medicaid structures had the second highest adjusted odds (OR=1.31). Expressed as a predicted probability for a typical child in foster care using a stimulant, one-third (33.6%) of children in foster care in counties using carve-outs would be predicted to use a second-generation antipsychotic, compared with 45.9% of such children in fee-for-service counties and 39.8% of children in foster care in integrated Medicaid programs (Figure 2).
| Adjusted for child-level variables | Adjusted for child- and county-level variables | |||||
|---|---|---|---|---|---|---|
| Medicaid eligibility and delivery structure | OR | 95% CI | p | OR | 95% CI | p |
| Foster care | ||||||
| Fee for service | 1.69 | 1.26–2.27 | <.001 | 1.55 | 1.28–1.88 | <.001 |
| Integrated | 1.31 | 1.08–1.58 | .006 | 1.71 | 1.28–2.27 | <.001 |
| Supplemental Security Income | ||||||
| Fee for service | 1.30 | 1.10–1.55 | .003 | 1.30 | 1.09–1.54 | .003 |
| Integrated | 1.19 | 1.01–1.41 | .036 | 1.51 | 1.15–1.97 | .003 |
| Temporary Assistance for Needy Families | ||||||
| Fee for service | 1.10 | .86–1.41 | .430 | 1.13 | .90–1.40 | .300 |
| Integrated | 1.14 | .86–1.51 | .370 | 1.45 | 1.05–2.01 | .024 |
a
Child-level regression models adjusted for age, race-ethnicity, sex, year, and pharmacy structure; county-level models further adjusted for provider supply, urbanicity, poverty rate, uninsured rate, unemployment, and proportion of county that voted Democrat in the 2008 election. The study sample comprised youths ages three to 18 with any stimulant prescription and ten months of Medicaid enrollment in one of the study counties. The reference group for all eligibility categories was carve-out.
Figure 2 Predicted probability of second-generation antipsychotic use among stimulant-using youths, by Medicaid eligibility group and county organizational structurea
aVertical lines indicate confidence intervals. SSI, Supplemental Security Income; TANF, Temporary Assistance for Needy Families
Among stimulant-using children in the SSI eligibility group, use of second-generation antipsychotics was again significantly higher in counties with fee-for-service (OR=1.30) and integrated (OR=1.19) structures compared with counties with carve-outs. The relative difference between counties with fee-for-service and integrated systems was smaller among children receiving SSI than for children in foster care. One third (33.3%) of children receiving SSI in counties with carve-outs were predicted to use a second-generation antipsychotic (similar to foster care youths in those counties), compared with 39.2% of children with SSI eligibility in counties with fee-for-service structures and 37.2% of children with SSI eligibility in counties with integrated Medicaid. Finally, there were no statistically significant differences in second-generation antipsychotic use across organizational structures for stimulant-using children in the TANF eligibility group. Use of second-generation antipsychotics was lowest overall in this population, ranging from 15.0% in counties with carve-out systems to 16.7% in counties with integrated systems.
Further adjustment for county-level variables (Table 3) generally did not alter relative differences in second-generation antipsychotic use between counties with carve-out versus fee-for-service structures. They did, however, substantially increase the relative differences between counties with integrated versus carve-out structures and the predicted probabilities of using a second-generation antipsychotic in those counties. For example, adjustment for these variables increased the predicted probability of second-generation antipsychotic use for children in foster care from 39.8% to 45.3%. This was likely due to adjustment for county-level variables such as the poverty rate, which was associated with higher second-generation antipsychotic use on average (Figure 2). [Additional analysis, including sensitivity tests for our main models, is available in an online data supplement to this article.]
Discussion
We systematically compared use of second-generation antipsychotic medications among Medicaid-enrolled children in 82 counties, focusing on differences across three organizational structures (carve-outs, fee for service, and integrated). We found marked differences in second-generation antipsychotic use across organizational structures. For example, one-third of stimulant-using children in foster care in counties with carve-outs used second-generation antipsychotics, compared with almost half of such children in counties with fee-for-service plans—a 31% relative difference. Second-generation antipsychotic use was relatively higher among children in foster care in counties with integrated versus carve-out structures. Differences across organizational structures were smaller, but still significant, for children in the SSI eligibility group. These patterns persisted after adjustment for other sociodemographic and service system features of counties. Rates of use did not significantly vary across organizational structures for stimulant-using children in the TANF eligibility group, however.
Wider variation in second-generation antipsychotic use across policy structures among children in foster care, relative to other eligibility groups, likely reflects the fact that these children are a particularly vulnerable population with high service use needs. In the absence of other institutional constraints, children in foster care may be targeted more often for second-generation antipsychotic treatment, with less oversight from a consenting guardian or parent (29). Children in foster care may also experience changes in placement, which can increase disruptions in treatment and reduce oversight (30).
Our study findings contribute to an emerging literature on regional variation in mental health treatment. Small area differences in the intensity and cost of treatment recently have been identified for psychotropic use among children (31). Although contextual factors, such as physician supply and poverty that predispose populations to receive more treatment, are undoubtedly contributors (32,33), differences across systems persisted after models adjusted for the factors we hypothesized could independently influence use of antipsychotics.
Our study underscores the need to focus on the prescribing behavior of clinicians, which in turn is likely to be influenced by incentives and constraints in different organizational structures. We hypothesized that second-generation antipsychotic use would be lower in both types of capitated systems compared with fee-for-service systems because capitated programs are more likely to contain costs by limiting access to high-intensity services and managing a network of providers. Partially confirming this hypothesis, we found that use of second-generation antipsychotics was consistently higher among children in foster care and SSI populations in fee-for-service versus carve-out systems, but it was not necessarily higher in fee-for-service versus integrated systems.
Higher use of second-generation antipsychotics in counties with fee-for-service versus carve-out structures could stem from a less restrictive prescribing environment. Reducing the barriers to access to mental health treatments is considered to be one positive feature of fee-for-service systems, particularly in light of concerns that Medicaid-enrolled children with mental health problems often receive limited treatment and delayed care (34). For instance, access to inpatient care is highly managed and may be more limited in carve-out systems (35). One unintended consequence of a less restrictive environment, however, could be broadened access to treatments such as second-generation antipsychotics, which are prone to overuse and may raise safety concerns. These competing objectives—reducing barriers to treatment and constraining prescribing of second-generation antipsychotics and similar medications—may require more complex reimbursement models and delivery structures. Hence, efforts to constrain use of second-generation antipsychotics should be evaluated within the broader context of efforts to increase contact with providers and increase use of nonmedication treatments (including cognitive-behavioral therapy), which could be particularly beneficial for children with emotional and behavioral problems.
It is less clear why use of second-generation antipsychotics among children in foster care and SSI populations might be higher in counties with integrated structures relative to carve-outs. We posit that some of this difference could be due to more aggressive management of inpatient and outpatient utilization and more coordinated programming of a range of therapeutic behavioral health services by specialized managed behavioral health organizations compared with integrated organizations with less experience and expertise in contracting for mental health services.
Cost management may reduce second-generation antipsychotic prescribing among primary care clinicians but also buffer use of inpatient care and care in other treatment settings where second-generation antipsychotic use is more prevalent. Relatedly, it is possible that a greater focus on provider credentialing and education could limit second-generation antipsychotic use, although it is unknown whether these efforts are greater within carve-out structures. Further investigating these differences in delivery and regulation of treatment is a goal for future research.
Several limitations should be considered in interpreting our findings. First, some service utilization may be underreported in the MAX for managed behavioral health care enrollees (36). Reassuringly, recent studies have found that patterns of prescription medication utilization in the MAX are consistent between managed care and fee for service (37). As noted, we also mitigated concerns about underreporting by restricting our sample to children who already had a recorded stimulant prescription fill. Second, we did not adjust for diagnoses or clinical symptoms because of concerns of systematic bias in underreporting across organizational structures. As such, we could not assess patterns of service use within particular diagnostic groups or levels of clinical need.
Third, as mentioned, we focused on the predominant organizational structures in the counties, but we could not systematically track cases where children in foster care or those in the SSI program may have been exempted from managed care. Because such children are sometimes allowed to remain in fee-for-service Medicaid in counties with managed care for other populations, our results for these groups may underestimate differences between fee-for-service and other structures. Finally, we emphasize that although our cross-sectional study design can identify important associations, it cannot establish causality. There may be other unmeasured features of counties that simultaneously influence choice of organizational structure and prevalence of second-generation antipsychotic prescribing. Future research, employing pre-post comparisons, might enable better identification of causal mechanisms. Research might also examine the contribution of organizational structure to regional variation in antipsychotic use relative to other contextual variables.
Conclusions
The differences across organizational structures we identified in our study warrant further research, both to understand potential pathways through which organizational structures shape provider behavior and to identify the long-term impact on safety and quality of care.
Efforts to constrain overuse of second-generation antipsychotics among Medicaid-enrolled children will require some providers to change their prescribing patterns. Although carved out Medicaid mental health benefits could reduce use of second-generation antipsychotics in populations of children, a recent trend toward more integrated services should prompt closer investigation of other tools for tailoring second-generation antipsychotic use across a variety of settings. These tools include formulary restrictions, prior authorization, provider education (that is, academic detailing), and expanded access to alternatives to pharmacotherapy (including psychotherapy).
Acknowledgments and disclosures
This project was supported by grant R01 HS01855001A1 from the Agency for Healthcare Research and Quality in the National Institutes of Health (NIH). Dr. Saloner received financial support from the Robert Wood Johnson Foundation Health and Society Scholars Program. Dr. Raghavan also acknowledges funding support from NIH grants R01 HS020269 and R01 MH092312.
Dr. French is a statistical editor for JAMA Pediatrics. The other authors report no competing interests.
Supplementary Material
File (appi.ps.201300574.ds001.pdf)
- View/Download
- 51.80 KB
File (appi.ps.201300574_ds001.pdf)
- View/Download
- 51.80 KB
References
1.
Crystal S, Olfson M, Huang C, et al: Broadened use of atypical antipsychotics: safety, effectiveness, and policy challenges. Health Affairs (Project Hope) 28:w770–w781, 2009
2.
Verdier J, Zlatinov A: Trends and Patterns in the Use of Prescription Drugs Among Medicaid Beneficiaries: 1999 to 2009. Princeton, NJ, Mathematica Policy Research, 2013
3.
Children’s Mental Health: Concerns Remain About Appropriate Services for Children in Medicaid and Foster Care. Report GAO-13-15. Washington, DC, Government Accountability Office, 2012. Available at www.gao.gov/assets/660/650716.pdf
4.
Lagnado L: US probes use of antipsychotic drugs on children. Wall Street Journal, 2013. Available at online.wsj.com/news/articles/SB10001424127887323477604578654130865747470
5.
Matone M, Localio R, Huang Y-S, et al: The relationship between mental health diagnosis and treatment with second-generation antipsychotics over time: a national study of US Medicaid-enrolled children. Health Services Research 47:1836–1860, 2012
6.
Correll CU, Manu P, Olshanskiy V, et al: Cardiometabolic risk of second-generation antipsychotic medications during first-time use in children and adolescents. JAMA 302:1765–1773, 2009
7.
Andrade SE, Lo JC, Roblin D, et al: Antipsychotic medication use among children and risk of diabetes mellitus. Pediatrics 128:1135–1141, 2011
8.
Correll CU: Antipsychotic use in children and adolescents: minimizing adverse effects to maximize outcomes. Journal of the American Academy of Child and Adolescent Psychiatry 47:9–20, 2008
9.
Cohen D, Bonnot O, Bodeau N, et al: Adverse effects of second-generation antipsychotics in children and adolescents: a Bayesian meta-analysis. Journal of Clinical Psychopharmacology 32:309–316, 2012
10.
Rubin D, Matone M, Huang Y-S, et al: Interstate variation in trends of psychotropic medication use among Medicaid-enrolled children in foster care. Children and Youth Services Review 34:1492–1499, 2012
11.
Rawal PH, Lyons JS, MacIntyre JC, II, et al: Regional variation and clinical indicators of antipsychotic use in residential treatment: a four-state comparison. Journal of Behavioral Health Services and Research 31:178–188, 2004
12.
Mechanic D: Managing behavioral health in Medicaid. New England Journal of Medicine 348:1914–1916, 2003
13.
Mechanic D, Schlesinger M, McAlpine DD: Management of mental health and substance abuse services: state of the art and early results. Milbank Quarterly 73:19–55, 1995
14.
Bloom JR, Cheng J-S, Hu TW, et al: Use of antipsychotic medications in treating schizophrenia among different financing and delivery systems. Journal of Mental Health Policy and Economics 6:163–171, 2003
15.
Hurley R, Zuckerman S: Medicaid managed care: state flexibility in action; in Federalism and Health Policy. Edited by Holahan J, Weil A, Wiener JM. Washington, DC, Urban Insitute, 2003
16.
Hutchinson AB, Foster EM: The effect of Medicaid managed care on mental health care for children: a review of the literature. Mental Health Services Research 5:39–54, 2003
17.
Tang MH, Hill KS, Boudreau AA, et al: Medicaid managed care and the unmet need for mental health care among children with special health care needs. Health Services Research 43:882–900, 2008
18.
Frank RG, Garfield RL: Managed behavioral health care carve-outs: past performance and future prospects. Annual Review of Public Health 28:303–320, 2007
19.
Inkelas M: Incentives in a Medicaid carve-out: impact on children with special health care needs. Health Services Research 40:79–99, 2005
20.
Hurlburt MS, Leslie LK, Landsverk J, et al: Contextual predictors of mental health service use among children open to child welfare. Archives of General Psychiatry 61:1217–1224, 2004
21.
Burcu M, Zito JM, Ibe A, et al: Atypical antipsychotic use among Medicaid-insured children and adolescents: duration, safety, and monitoring implications. Journal of Child and Adolescent Psychopharmacology 24:112–119, 2014
22.
Zito JM, Safer DJ, Sai D, et al: Psychotropic medication patterns among youth in foster care. Pediatrics 121:e157–e163, 2008
23.
Konrad TR, Ellis AR, Thomas KC, et al: County-level estimates of need for mental health professionals in the United States. Psychiatric Services 60:1307–1314, 2009
24.
Cook BL, Doksum T, Chen C-N, et al: The role of provider supply and organization in reducing racial/ethnic disparities in mental health care in the US. Social Science and Medicine 84:102–109, 2013
25.
Chevarley FM, Owens PL, Zodet MW, et al: Health care for children and youth in the United States: annual report on patterns of coverage, utilization, quality, and expenditures by a county level of urban influence. Ambulatory Pediatrics 6:241–264, 2006
26.
Snowden LR, Masland M, Wallace N, et al: Increasing California children’s Medicaid-financed mental health treatment by vigorously implementing Medicaid’s Early Periodic Screening, Diagnosis, and Treatment (EPSDT) Program. Medical Care 46:558–564, 2008
27.
Choi SO, Bae S-S, Kwon S-W, et al: County limits policy types and expenditure priorities. American Review of Public Administration 40:29–45, 2010
28.
Stata Statistical Software: Release 12. College Station, Tex, StataCorp, 2011. Available at www.stata.com/support/faqs/resources/citing-software-documentation-faqs
29.
Naylor MW, Davidson CV, Ortega-Piron DJ, et al: Psychotropic medication management for youth in state care: consent, oversight, and policy considerations. Child Welfare 86:175–192, 2007
30.
Mekonnen R, Noonan K, Rubin D: Achieving better health care outcomes for children in foster care. Pediatric Clinics of North America 56:405–415, 2009
31.
Leslie LK, Raghavan R, Hurley M, et al: Investigating geographic variation in use of psychotropic medications among youth in child welfare. Child Abuse and Neglect 35:333–342, 2011
32.
Bruckner TA, Hodgson A, Mahoney CB, et al: Health care supply and county-level variation in attention-deficit hyperactivity disorder prescription medications. Pharmacoepidemiology and Drug Safety 21:442–449, 2012
33.
Cox ER, Motheral BR, Henderson RR, et al: Geographic variation in the prevalence of stimulant medication use among children 5 to 14 years old: results from a commercially insured US sample. Pediatrics 111:237–243, 2003
34.
Sparer M: Medicaid Managed Care: Cost, Access, and Quality of Care. Princeton, NJ, Robert Wood Johnson Foundation, 2012
35.
Raghavan R, Leibowitz AA, Andersen RM, et al: Effects of Medicaid managed care policies on mental health service use among a national probability sample of children in the child welfare system. Children and Youth Services Review 28:1482–1496, 2006
36.
Nysenbaum J, Bouchery E, Malsberger R: The Availability and Usability of Behavioral Health Organization Encounter Data in MAX 2009. Princeton, NJ, Mathematica Policy Research, 2013
37.
Byrd VLH, Dodd AH: Assessing the usability of MAX 2008 encounter data for comprehensive managed care. Princeton, NJ, Mathematica Policy Research, 2013. Available at ideas.repec.org/p/mpr/mprres/7715.html. Accessed August 5, 2013
Information & Authors
Information
Published In
Cover: Two Children Watching the Clouds in a Field, by Elizabeth Shippen Green. © Copyright 2014 National Museum of American Illustration™, Newport, RI. Photos courtesy Archives of the American Illustrators Gallery™, New York City.
History
Published ahead of print: 17 November 2014
Published online: 1 December 2014
Published in print: December 01, 2014
Authors
Funding Information
NIHR Health Services and Delivery Research (HS&DR) programme10.13039/501100002001: R01 HS01855001A1
Robert Wood Johnson Foundation10.13039/100000867:
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).