This first evaluation report on the Center for Medicare & Medicaid Innovation’s Integrated Care for Kids (InCK) Model identified several challenges, including the need to navigate complex legal and regulatory environments to establish data use agreements with organizations that provide services to children. Model participants also report issues with developing new data platforms to share information to support service integration.
CMS awarded InCK Model funding to implement locally designed child-and-family-centered delivery models and pediatric alternative payment models. Delivery models intend to expand care coordination beyond healthcare to include Core Child Services (such as schools, housing, and food services) and address unmet service needs. Other goals include incentivizing and facilitating quality improvements in care, reducing Medicaid expenditures, and reducing avoidable out-of-home placements among children.
At the start of the pre-implementation period, eight lead organizations received funding to implement the InCK Model. By the end of the pre-implementation period, seven organizations progressed to the five-year implementation period: Ann & Robert H. Lurie Children’s Hospital (Chicago, Illinois); Montefiore Medical Center (Bronx, New York); Clifford W. Beers Guidance Clinic (New Haven, Connecticut); Duke University, in partnership with University of North Carolina (select counties in North Carolina); Hackensack Meridian Health, in partnership with Visiting Nurse Association of Central New Jersey and the New Jersey Health Care Quality Institute (Central New Jersey); Nationwide Children’s Hospital (Eastern Ohio); and Egyptian Health Department (Southern Illinois).
CMS contracted with Abt Associates Inc. and its partners, Bailit Health and Insight Policy Research, to evaluate the implementation and impact of the InCK Model for each of the model’s award recipients and across all award recipients.
During the pre-implementation period, the evaluation team characterized model pre- implementation activities and provider, staff, patient, and caregiver experiences; captured information about local context; provided tailored support on model requirements; created measure specification and data templates for data collection; and determined a comparison group for each award recipient.
All award recipients (ARs) cited the need to improve systems of care for children and caregivers as the primary reason for applying to the model, and each created individualized approaches based on needs in their communities and their local context.
The report noted that the lead organizations have struggled to establish data use agreements (DUAs) with the state agencies that manage child core services (CCS) data. The challenges most often cited were legal barriers to share data and staff bandwidth issues related to the COVID-19 public health emergency (PHE). “Leaders at several organizations reported that program staff at the state CCS agencies supported data sharing for InCK Model purposes and were even excited about the model; however, lawyers at those agencies reported that statutory regulations made sharing individual-level, identifiable data difficult and sometimes impossible,” the report said.
For many groups, state CCS agencies could share data with the state Medicaid agency for the purposes of implementing the model. However, many agencies cited concerns about sharing individually identifiable data that would be submitted to CMS. Groups in North Carolina and Ohio had the most success negotiating with partners by the end of the pre-implementation period.
Several executives reported CCS organizations and other state CCS agencies raised the most concerns about data for behavioral health, substance use, and child welfare. Some ARs reported that establishing data sharing agreements with child welfare agencies was particularly challenging, and their respective agencies might not be able to share identifiable data at all.
Across the award recipients, behavioral health providers reported they typically have little information about a patient’s other health or social needs. Likewise, physical health providers do not know about services their patients may be receiving from other providers or organizations. For example, one adolescent health provider in the NC InCK region reported the only way he knows if a patient is receiving behavioral health services through school is if the patient tells him.
Despite having data sharing systems in place and pre-existing closed-loop referral systems, physical and behavioral health, and social service providers across the award recipients described having little awareness of services that patients may receive in other settings.
The award recipients planned to use existing or newly created data systems for service integration coordinators to track their work; store care plans; and facilitate information sharing between stakeholders. For example, OH InCK contracted with a third-party vendor to build a platform called Apricot 360, which will allow for a family’s single point of contact to invite members of the patient’s care team to share information with one another.
Developing new virtual platforms and integrating them into existing workflows and information sharing processes took significant effort in the pre-implementation period, the report said. For some organizations, delays in procuring a vendor to develop these platforms, managing that vendor, and then making sure the vendor can produce a tool with the functionality originally promised was a significant challenge, the report said.
For some of them, procurement took longer than anticipated and design work was slow, as their planned approach evolved over the course of the pre-implementation period, the report said. “Toward the end of 2021, two of the common vendors—Unite Us and NowPow—merged. ARs reported that staff turnover and other changes resulting from the merger led to further delays. CT InCK Embrace New Haven initially planned to use Unite Us but decided to move forward with a different vendor for some activities after difficulties executing their planned approach.”
“Finally, late in 2021, some ARs raised concerns that these virtual platforms would not have the robust functionality that vendors originally promised. At the end of the pre- implementation period, most of the ARs were still working to finalize these systems and integrate them into existing workflows. It remains unclear whether system functionality allows ARs to implement as planned.”
