Safe and Judicious Use of Antipsychotic Medications in Children and Adolescents Toolkit

Change Ideas to Improve Target Measure Performance

Table 2 below provides change idea details, resources, and tools associated with the secondary drivers presented above in the key driver diagrams. These change ideas provide actionable pathways to support QI activities aimed at improving target measure performance in the safe and judicious use of antipsychotic medications in children and adolescents. The table is organized at the level of implementation (e.g., plan, plan/practice, youth/family) indicated in the key driver diagrams and then grouped by secondary driver.

Table 2. Change Ideas to Improve Target Measure Performance

Quality Improvement in Action: The NCINQ Antipsychotics in Children Learning Collaborative

From August 2017 to December 2018, NCINQ led a quality improvement (QI) learning collaborative with five New York-based Medicaid health plans. Participants in the collaborative used data gathered through national Healthcare Effectiveness Data and Information Set (HEDIS) reporting and the Medicaid Child Core Set to examine performance and improvement on several quality measures that address the use of antipsychotics in children and adolescents. The plans in this collaborative convened for two in-person meetings, bi-monthly webinars, monthly individual calls, and office hours as needed.

Plans that participated in the collaborative used and tested the key drivers and change ideas presented in this toolkit.ⁱⁱ Case studies and examples from the collaborative are highlighted below and in the section on QI Best Practices for the Safe and Judicious Use of Antipsychotic in Children and Adolescents. Plans also tested strategies to engage youth and family in their QI work. For additional information, strategies, and resources on engaging youth and family to support QI work, please see the standalone toolkit: Youth and Family Engagement in Quality Improvement.

The sections below describe measure-specific performance benchmarks, improvement strategies implemented by the participating plans, and performance results from the collaborative. Information on considerations for selecting appropriate improvement targets is also presented, as well as a summary of the contextual factors found to facilitate progress, key components of successful strategies, and identified challenges and barriers to improvement.

____________________

^ii. The strategies included in this toolkit were implemented prior to the COVID-19 pandemic and may need to be adapted to account for changes in care access and delivery during the pandemic.

Considerations for Selecting Appropriate Improvement Targets

To identify benchmarks that could be used to set improvement goals for each measure and to subsequently monitor performance during the course of the collaborative, the team undertook the following activities in a stepwise fashion:

1. Reviewed the HEDIS national performance benchmarks as well as the NY state benchmarks (see tables below in each measure-specific section for details).
2. Conducted an “improvement analysis” using 2015 and 2016 HEDIS reporting data to calculate the amount of improvement at the plan level seen year to year. “High improving” plans were defined as those that exhibited improvement rates between 2015 and 2016 in the top 25^th percentile.
3. Worked individually with the plans participating in the collaborative to compare baseline performance rates to the national and state benchmarks, as well as the findings from the “improvement analysis.”
4. Coached plans to track the amount of improvement by comparing current performance rates with where the performance was at one year prior (i.e., internal benchmarking).

Most plans selected modest improvement targets for the Psychosocial Care and Metabolic Monitoring measures, equating to around a 10% increase in performance rates compared to baseline. The Multiple Concurrent Antipsychotics measure performance was already very good for most plans with little room for improvement. Two plans chose to implement focused strategies for improvement that targeted this measure; however, only one of the plans was able to realize an improvement in performance.

Use of First-Line Psychosocial Care for Children and Adolescents on Antipsychotics

Table 3 below provides HEDIS performance benchmarks at the national and NY state levels for the Medicaid product line for the Psychosocial Care measure from 2016 through 2019. Voluntary state reporting data for this measure is available for federal fiscal year 2018 is available in the Medicaid Child Core Set.  

Table 3. Medicaid Performance Benchmarks, First-Line Psychosocial Care

Specific strategies implemented and tested during the collaborative to improve first-line psychosocial care in children and adolescents are included in Table 4 below. References to case study examples are provided when applicable and notable lessons learned specific to a particular strategy are presented where relevant. Strategies were implemented at the plan level.

Table 4. First-Line Psychosocial Care Improvement Strategies

Chart 1. First-Line Psychosocial Care Performance Results for Collaborative Plans that Targeted the Measure for Improvement, 2017-19

The run chart above (Chart 1) provides measure performance data for the three plans in the learning collaborative that implemented strategies to improve performance on the Psychosocial Care measure. The chart has been annotated to identify when particular strategies of note were implemented. For reference and context, the various strategies each plan implemented are noted below, as each plan implemented more than one strategy.

• Plan A: Prior-authorization pharmacy edits (Case Study 1); member and family outreach, including calls from clinical staff to potentially at-risk members to assess whether the member has had psychosocial care, needs to be connected with services, or is experiencing any barriers to receiving care.
• Plan B: Provider outreach and engagement to understand missed assessments and to work to improve access and compliance (Case Study 2); youth empowerment program to provide member programming on mental health treatment options and healthy lifestyle management.
• Plan C: Member and family outreach through the use of care managers and predictive reports to identify at-risk children and adolescents and address barriers to receiving psychosocial care; PSYCKES lookups and medical record data collection to identify psychosocial care not found in claims data; mental health awareness campaign to educate families on antipsychotics.

Metabolic Monitoring for Children and Adolescents on Antipsychotics

Table 5 below provides HEDIS performance benchmarks at the national and NY state levels for the Medicaid product line for the Metabolic Monitoring measure from 2016 through 2019.

Table 5. Medicaid Performance Benchmarks, Metabolic Monitoring

Specific strategies implemented and tested during the collaborative to improve appropriate metabolic monitoring in children and adolescents prescribed an antipsychotic are included in Table 6 below. References to case study examples are provided when applicable and notable lessons learned specific to a particular strategy are presented where relevant. Strategies were implemented at the plan level.

Table 6. Metabolic Monitoring Improvement Strategies

Chart 2. Metabolic Monitoring Performance Results for Collaborative Plans that Targeted the Measure for Improvement, 2017-19

The run chart above (Chart 2) provides measure performance data for the four plans in the learning collaborative that implemented strategies to improve performance on the Metabolic Monitoring measure. The chart has been annotated to identify when improvement strategies were implemented. For reference and context, a complete list of the strategies implemented are included below, since each plan implemented a multi-pronged improvement approach.

• Plan B: Standing lab orders; empowerment program with community partner to host member programming on mental health treatment options and health lifestyle management; incentive program to encourage appropriate metabolic monitoring.
• Plan C: Provider outreach, including care gap reports and discussion; data exchange with hospitals to capture lab test data not captured in claims; member and family education through care management.
• Plan D: Provider outreach and education (Case Study 3); member and family outreach (data team identified members past due on monitoring and case managers targeted these members through education and lab appointment reminders).
• Plan E: Provider outreach and engagement via mailing campaign.

Use of Multiple Concurrent Antipsychotics in Children and Adolescents

Table 7 below provides HEDIS performance benchmarks at the national and NY state levels for the Medicaid product line for the Multiple Concurrent Antipsychotics measure from 2016 through 2018. Voluntary state reporting data for this measure is also available from fiscal year 2016 through 2018 at the Medicaid Child Core Set.  

Table 7. Medicaid Performance Benchmarks, Multiple Concurrent Antipsychotics

Specific strategies implemented and tested during the collaborative to improve performance on the Multiple Concurrent Antipsychotics measure are included in Table 8 below. References to case study examples are provided when applicable and notable lessons learned specific to a particular strategy are presented where relevant. Strategies were implemented at the plan level.

Table 8. Multiple Concurrent Antipsychotics Improvement Strategies

Chart 3. Multiple Concurrent Antipsychotics Performance Results for Collaborative Plans that Targeted the Measure for Improvement, 2017-19

The run chart above (Chart 3) provides measure performance data for the two plans in the learning collaborative that implemented strategies to improve performance on the Metabolic Monitoring measure. The chart has been annotated to identify when improvement strategies were implemented. For reference and context, the strategies implemented by each plan for improvement on this measure are noted below. 

• Plan A: Pharmacy edits to reject second prescription when a child or adolescent is already prescribed an antipsychotic medication (Case Study 4).
• Plan C: Monthly chart review and provider education.

Contextual Factors Facilitating Performance Improvement

Throughout the course of the collaborative, the team found that the following contextual factors facilitated performance improvement on the safe and judicious use of antipsychotics in children and adolescents.

• Availability of National HEDIS and State measure performance benchmarks helped plans set improvement goals.
• Alignment of measures across reporting programs allowed the plans to capitalize on data from national, state, plan, and provider levels to target QI efforts. For the NY Quality Assurance Reporting Requirements (QARR)[i], plans are required to use HEDIS specifications, which ensures alignment of data elements and coding across both programs. High levels of reportability and measure utilization stemming from measure alignment contributed to measure uptake and widely available performance results, enabling an examination and comparison of performance rates for NY plans participating in the collaborative, other plans in NY, and plans reporting nationally.
• Running monthly performance rates allowed plans to examine data monthly and compare against the previous year to determine if they were trending toward improvement.
• Having a range of expertise represented on the QI team allowed some plans to implement very specific strategies, such as pharmacy edits, because they had access to the relevant specific expertise (e.g., pharmacy consultant on staff).
• Involvement and buy-in from senior leadership facilitated key provider partner relationships.

Key Components of Successful Strategies

A handful of key components to successful strategies were identified during the collaborative. These key components were not specific to improvement on any specific measure, but rather appeared to have a positive impact regardless of the measure targeted for improvement.

• Using reports to highlight provider performance offered opportunities to educate and deliver resources to providers.
• Focused education and outreach to prescribers (e.g., psychiatrists) was more effective when led by medical directors (i.e., peer-to-peer education) versus QI team members.
• Regular in-person meetings at practice sites with clinical and front office staff at practices proved to be most effective in engaging provider teams and served as opportunities to educate and provide resources to provider teams that supported performance improvement.

Challenges and Barriers to Improvement

Throughout the collaborative, it became apparent that the following items presented challenges to plans when undertaking initiatives to improve performance on the pediatric antipsychotics use measures. To the extent that plans embarking on quality improvement campaigns for the antipsychotics measures can mitigate or control these factors, doing so may increase the likelihood of performance improvement. These challenges were not found to be specific to one particular measure but rather global in nature.

• Given the very small number of youths receiving antipsychotics, plans were met with resistance from providers to engage in QI for these measures. Providers were more focused on quality measures that impacted a greater number of their patients. In particular, the Multiple Concurrent Antipsychotics measure was challenging to engage providers on due to the very small number of youths receiving this therapy.
• There were challenges to coordination of care between behavioral health and physical health providers in settings where there was a lack of behavioral health integration in primary care.
• Due to concerns about sharing mental health information per federal regulations and the stigma associated with behavioral disorders, pediatricians were not always aware of services their patients received elsewhere.
• Plans found it challenging to determine which providers to hold accountable for managing the metabolic side effects of antipsychotics. While antipsychotics were often prescribed by psychiatrists, lab tests to monitor side effects were not typically shared with pediatricians.
• For the Metabolic Monitoring measure, the specifications did not highlight care gaps that existed. Plans found that examining whether children and adolescents received glucose testing and cholesterol testing as two separate rates provided more actionable information. Specifically, when these rates were reported separately, gaps in cholesterol testing became evident, which plans could then target to drive improvements.

Lack of Information on Deprescribing: Even when plans were able to engage providers, there was very little evidence-based guidance available on protocols for deprescribing antipsychotics in children and adolescents that QI teams could provide via provider outreach and education.