Written by Chad Miller, Chief Intelligence Officer. Interested in learning more? Stay connected — follow Chad on LinkedIn
Those of us working in Medicaid have seen a fundamental truth: social factors play a foundational role in clinical outcomes and cost of care. Housing instability, food insecurity, transportation barriers – these are not peripheral issues. They are central to succeeding in value‑based arrangements.
At Care Compass Network (CCN) we are working to advance value-based care across our region. By connecting healthcare, behavioral health, and community-based organizations, we leverage data and analytics to improve outcomes, address social drivers of health, and strengthen the communities it serves.
A few years ago, we acted on that belief by launching a social care pilot. Over a two‑year period, we coordinated a social and clinical integrated care team by embedding Community Health Workers (CHWs) employed by Community Based Organizations (CBOs) directly into primary care practices to support a health system operating under value‑based contracts.
The intent was straightforward: to address social needs upstream and help the system achieve better outcomes and financial performance.
Even as our data infrastructure underwent a complete rebuild, we recognized how critical this work was and launched the program ahead of the rebuild. To innovate you cannot wait for perfection.
As a result, during the pilot phase our focus was operational and relational. We concentrated on developing population health and data strategies, management support, integrating CHWs into clinical workflows, building trust with practices, and ensuring that referrals to community resources were happening.
What we lacked was clear, consistent data on what was occurring day to day. We couldn’t easily answer basic questions: How many patients on our roster were being seen each day? What insurance product/payer were they affiliated with? What needs were being addressed? And how did that activity connect back to the value‑based populations we were trying to support?
The end of the pilot phase of the social care program coincided with the standing up of our expanded analytics department. When we finally looked closely at the data, the picture was sobering. From a volume perspective, daily patient encounters per CHW were lower than expected. Additionally, of those engagements nearly half of the CHWs interventions were provided to patients who were not even on the targeted cohort panel. Documentation around referral reasons and interventions was scattered and inconsistent.
From the standpoint of efficiency, the model was clearly underperforming.
But the total cost of care data told a different story. When we analyzed total cost of care, we found a statistically significant impact for patients who had seen a CHW in the first quarter of each year. Despite all the operational shortcomings, those patients’ clinical costs had been reduced. The signal was real – even if the delivery model was far from optimized.
The data forced us to be more precise. It showed us that value was being created despite inefficiencies and inconsistencies. And it made clear that if we wanted to see deeper results, scale, and ultimately sustain the program under value‑based arrangements, we needed to lead with analytics, letting our data drive our decisions on how to pivot.
“Rather than jumping immediately to address more sophisticated outcome measures, we needed to take a deliberate step back.” – Chad Miller, Care Compass Chief Business Intelligence Officer
The priority became clear; it was not to prove impact again – it was to fix the program structure. We changed the leadership of the program to a dedicated program manager who could focus on converting it from a pilot to an operational program. On the analytics side, we focused on process metrics as a foundation: daily CHW encounters, alignment to attributed populations, and consistent documentation of referral reasons and activities. These process metrics were tools for operational discipline.
That shift mattered.
Once expectations were clear and activity was visible, behavior changed. CHW encounters increased significantly. Now equipped with standardized data reporting and workflows, we could see that overall outreach attempts increased by 38% and completed outreaches for panel members increased by 98%. Work became more intentional and better aligned with the populations tied to value‑based contracts. Leaders gained confidence that resources were being deployed where they mattered most. Importantly, none of this required asking CHWs to work harder. It required the system to work smarter.
There is a broader lesson here. Social care programs often underperform not because they lack value, but because they lack structure. Without analytics, everything looks equally urgent, resources get diluted, and success is defined by activity rather than impact. With analytics, social care can be managed like any other clinical strategy: targeted, measured, and continuously improved.
In Medicaid, social care does not fail because it lacks heart. It fails when it lacks visibility. Data provides that visibility. And when compassion is paired with operational rigor, social care doesn’t just feel right—it works.
About the Author
