The Cerebral Palsy Research Network (CPRN) is a comprehensive network of hospitals and community members working together to improve health outcomes for people with CP. It hosts both community and clinical registries to gather comprehensive data for research. In a recent presentation to the Learning Health Sciences Collaboratory at the University of Michigan, Paul Gross, CPRN’s CEO, detailed how the network is leveraging EHR data for research and quality improvement and adopting a learning health system model.
Gross, who has a child with hydrocephalus and CP, is described as a driving force in accelerating clinical and translational research in neuroscience for cerebral palsy and hydrocephalus. Besides his work with the CPRN, he is the co-founder of the Hydrocephalus Clinical Research Network, the past chairman of the Hydrocephalus Association (HA) , and co-founder of the Adult Hydrocephalus Clinical Research Network.
CPRN was founded in 2015 with a mission to optimize the lifelong health and wellness of people with CP and their families, and the core focus is high-quality research, education and community programming. The network has grown to 35 centers and includes over 15,000 patients and nearly 3,000 community members.
One of the first things CPRN’s founders did was run a three-month program funded by PCORI to set a patient-centered research agenda. “Our vision is that by engaging a community of stakeholders and a very large number of health systems all together in a learning health network for CP, we can collectively focus the research and the care,” Gross said. “We gather and share lived experience of community members, and we measure aspects of the healthcare process to continuously improve their health and well-being in ways that matter to them.”
Gross spoke about what is involved in growing the effort into a learning health network. They take standardized data collected for usual care, and integrate it into the hospital’s EHR so that usual care can be done with some structured data. “That allows us to capture data of all these different types. All the patient characteristics, all the measures and scales that are applied if they are getting treatments, either medical or surgical, and the outcomes,” he explained. What they are working toward is a local registry where all that data is accessible to do things about that cohort of patients at the hospital, but also to centralize it in a national registry. “Then, as we improve outcomes, we’re able to make changes to these forms and templates to actually do implementation and spread of the core outcomes that we create,” he added.
Turning this into a learning health network involves gathering all this data in a single data coordinating center. “Our sites sign a participation and data use agreement that refers to a set of standard operating procedures that are collectively developed within the network,” Gross added. “They sign that agreement with the CP Research Network, and then on the back end, the CP Research Network has a statement of work with the University of Pittsburgh, with Steve Wisniewski, who co-directs their epidemiology data center. That creates a way that our little nonprofit, can actually have this very big collection of very important medical data. This is the way that we get to a learning health network. We have all of these individual learning health systems working together to combine data and to to collaborate.”
Gross also spoke about their research and improvement strategy. “In other networks I’ve been involved with, we’ve tried to raise money to do lots of pilot studies and to make studies happen. Here, what we’re trying to do is invest in the infrastructure to let a thousand flowers bloom and really stimulate the ecosystem,” he said. “We set the patient-centered research agenda. We built the infrastructure of the registries, the network and that data coordinating center to help clinicians with the data they need to look at practice variation or to develop studies. We try to balance between generating evidence — answering unanswered questions — and assembling teams to improve care and improve outcomes. Then we use that registry and all that data to both support research for hypothesis generation or recruitment planning, and also to monitor the implementation of our improvements and the outcomes over time.”
He said this has led to a lot of activity, a lot of publications, a lot of ideas working their way through the pipeline. “I will say that a number of these quality improvement projects have actually produced manuscripts as well.”
One example is a focus on the improvement of the assessment of pain in adults. “We were able to grow that from 24% of the visits to 90% of the visits where pain was being assessed, and that was one of the top items that was in our patient-centered research agenda,” he said. “We were able to take that data and use it to apply for a grant for a study of pain classification, which is just entering its second year right now.” They also take some of those improvements and push them into the notes that physicians are using to be able to spread that improved assessment, and track that over time.
CPRN is investing in some new areas this year to try to grow the number of stakeholders that are involved in its work and double the the size of its community advisory group, so that it can involve them in every study, Gross said.
In summary, he identified several benefits for other sites of care, other hospital systems, other clinicians to participate with CPRN, including enhanced knowledge sharing. For example, they ran a series of meetings where each of the sites presented their model of care for CP, and shared that across 12 of the centers.
There’s also the improved care that comes out of the quality improvement methodology. “What I think is unique about what we do is this connection between the EMR and the registry being part of usual care means that we are not predominantly relying on duplicate data of clinical research assistants extracting data; we also have this big set of patient-reported outcome measures that can be linked to all this clinically validated data,” Gross said.
