Antipsychotic Medication Use in Children and Adolescents: A Resource Guide and Workbook

1. Antipsychotic Medication Use in Children andAdolescents: A Resource Guide and Workbook Jeffery Thompson, MD, MPH Washington State Medicaid Chief Medical Officer

2. How States Worked Together on this Project: The search for truth – or something close. MMDLN invited a small number of States to discuss AP medication management and whether comparing systems and policies was possible between June of 2007 and August of 2009: • A series of informal discussions which generated the Project Plan • A shared discussion among a subset of State participants to develop a standardized Data Dictionary • An informal data sharing agreement between 16 States to pull and share utilization and demographic data • A survey of State programs and practices to categorize polices, statutes, and utilization controls related to mental health medications • A process to share State practices based on a systematic process (best, promising, emerging, and unknown) • A publication committee to write a Resource Guide • Close connections with AHRQ, NASMD and NASMHD

3. What were project challenges? The search for truth – or something close. Centralized Model: • Research idea responding to an RFP • Null hypothesis and project plan in an RO1 • Data is centralized and analyzed (typically line level) • Some mechanism is used to validate data • Report or publication comes out • The end Distributed Model: • Some issues arises (political, budget or other) that has a common theme across states • Consensus building defines the project plan • Common data dictionary and data pull at each state • Discussions and surveys are used to ensure validity • Data is aggregated • The beginning

4. What were project challenges? The search for truth – or something close. Centralized Model: • Pros • The project has a narrow focus • Definitions setting is not egalitarian • Line level data allows for validity checks • Cons • Data and trends may be confounded scalar changes and unknowns • The research question may not address the community need Distributed Model: • Cons • The project can morph • Definitions is set by committee • Aggregate data relies on state analyst understanding • Pros • Each state knows their community and structures to address trend issues • States gain a unique understanding through the process

5. What were project challenges? Both models are needed and need coordination! Centralized Model: • Can define the confounders and influencing issues • Can test the intended and unintended consequences • Can assist in merging costs and outcomes Distributed Model: • Defines the problem(s) • Each state may have differing issues • Integrates research into business/program model • Sets the next step agenda for research

6. What were the key findings in 16 States CY 2004-2007 Data? • Growth trends in Antipsychotic (AP) medication varied widely (48 percent decrease to a 39 percent increase) among the 16 States • More children in foster care (12.4 percent) were prescribed AP medications than non-foster care children (1.7 percent) • Just over 11 percent of children using AP medications were prescribed four or more during a calendar year (non-concurrent) • Over one-third of children (36 percent) had a greater than 20-day gap in AP medication • 16 States, generated 26 best, promising and emerging quality practices to assist states in improving quality and reducing variation

7. How can we measure issues? Red flags can assist in Measures of Program Performance The project identified a common set of quality concerns (red flags) to raise safety concerns • Children less than 5 years of age receiving AP medications • Children and adolescents being prescribed three or more mental health medications in a calendar year • Children and adolescents being prescribed two or more AP medications in a calendar year • Doses of antipsychotic medications exceeding recommended range • Gaps in therapy that exceed 20 days States can use these “red flags” to highlight variation in performance and trends as well as review the best, promising, and emerging practices to work with the prescribing community to improve quality.

8. How is Washington State using project data?Executive Summary CY 2007 • Total users: 4,978 (1.4%) of enrolled kids used an AAP • Growth trends: AAP users grew 25% and unit costs grew 38% comparing 2004 to 2007 • AAP use in the very young: 187 AAPs users are less than 5 years old • High dose: 499 (10%) of users are prescribed high doses of AAPs • Multiple AAP use: 896 (19%) of users had two or more AAP prescriptions • Multiple Mental Health Drug use: 621 (2.3%) of kids are prescribed 5 or more mental health drugs • Adherence: 1588 (39%) had a gap in AAP use of greater than 20 days. • Provider types: AAP were prescribed by Psychiatrists (45%), PCP (34%) and ARNP (21%)

9. How is Washington State using project data?CY 2004 – 2007 Trends A look at the trend data WA State found: Next Steps: Provider groups are reviewing chart data and reasons for trends

10. How is Washington State using project data?CY 2004 Regional Variation in Clients using 4 or More Mental Health Drugs by County Showing the regional variations engaged the provider community in solutions and that quality issues are present.

11. How is Washington State using project data?Because of variation a statewide call line was setup. • PAL is a Call in Line for mental health questions by a Ped Psychiatrist • The UW will be doing a program review Data Drives Programs, Statutes, Funding and Quality

12. Summary: This Project will Assist States in Ensuring Mental Health Drugs are Given the right care at the right time for the right reason • Questions?