Part 2 of a series on value-based care
In the first post, I explained why healthcare is splitting into two systems—one that pays for volume, another that pays for value. But here's where it gets interesting: value-based care isn't one thing. It's a collection of different models, each with its own rules, risk levels, and reward structures.
If you're building technology for healthcare, you need to understand these models. Not at a conceptual level—at a mechanical level. Because the model determines what data matters, what workflows exist, who makes decisions, and where the leverage points are.
This is the guide I wish I'd had when we started Pear. Let's break down how value-based care actually works.
The Landscape: Who's Playing and How
Value-based care exists primarily in four domains:
- Medicare Shared Savings Program (MSSP) - The foundation, serving 11M+ beneficiaries
- ACO REACH - The advanced version, with full capitation options
- Medicare Advantage (MA) - Private insurance for Medicare with built-in VBC incentives
- Commercial VBC - Employer and private insurance value-based contracts
Each has different mechanics, different risk profiles, and different opportunities. Let's walk through them.
Medicare Shared Savings Program (MSSP): VBC Training Wheels
MSSP is where most organizations start with value-based care. It's the largest ACO (Accountable Care Organization) program, launched by CMS in 2012.
The basic mechanic:
You form an ACO—a legal entity that brings together providers (typically primary care physicians, specialists, hospitals) who agree to coordinate care for a defined patient population. CMS assigns you Medicare beneficiaries based on where they receive primary care. You continue getting paid normal fee-for-service rates for every service you provide, but now CMS also sets a spending benchmark based on your population's historical costs.
At the end of the year, CMS compares your actual spending against the benchmark. If you spent less while meeting quality standards, you share in the savings (typically 40-50% of the difference). This is called "one-sided risk"—you can gain from savings but don't lose money if you exceed the benchmark.
After a few years, most ACOs must move to "two-sided risk"—you share in savings (up to 75%) but also share in losses if you exceed the benchmark. This is where real alignment happens.
Example in practice:
Say your ACO has 10,000 attributed beneficiaries with a benchmark of $10,000 per beneficiary per year (total benchmark: $100M). You implement better care coordination, prevent unnecessary hospitalizations, and manage chronic conditions more effectively. At year end, actual spending is $95M.
You saved $5M. You hit quality benchmarks. You earn 50% of the savings: $2.5M to split among your ACO's providers.
Why this model exists:
MSSP gives traditional providers a low-risk way to experiment with value-based care. You don't have to restructure your entire organization or take major financial risk upfront. You can learn population health management, build care coordination teams, and develop the infrastructure while your existing fee-for-service revenue continues.
It's a bridge—not the destination, but a reasonable path to get there.
The challenge:
The shared savings model creates a strange dynamic. You're still maximizing fee-for-service revenue (more visits, more procedures, more tests) while trying to reduce total cost of care. These can work against each other. An ACO that's too successful at prevention might see fee-for-service revenue drop faster than shared savings can make up for it.
This is why many successful MSSP ACOs eventually move to more advanced models where the incentives fully align.
ACO REACH: Full Risk, Full Alignment
If MSSP is training wheels, ACO REACH (Realizing Equity, Access, and Community Health) is the advanced course. Launched in 2022, REACH allows ACOs to take on significantly more financial risk—and opportunity.
The key difference: capitation
In REACH, you can receive capitated payments—a fixed amount per patient per month (PMPM) to cover their care. Instead of billing fee-for-service and hoping for shared savings at year end, you get predictable monthly revenue and keep whatever you don't spend on care.
REACH offers different capitation options:
- Primary Care Capitation: Fixed PMPM for primary care services only
- Partial Capitation: Fixed PMPM for primary care plus Part B services (outpatient care)
- Total Care Capitation: Fixed PMPM for ALL Medicare Part A and B services (including hospital care)
The more services you capitate, the more risk you take—but also the more your incentives align with keeping patients healthy.
Example in practice:
Your ACO chooses total care capitation for 5,000 beneficiaries. CMS calculates a risk-adjusted payment (say $900 PMPM, or $10,800 per beneficiary per year). You receive $4.5M monthly ($900 × 5,000 patients).
Now prevention becomes directly profitable. Every ER visit you prevent, every hospital readmission you avoid, every complication you catch early—that's money that stays in your organization instead of going to hospitals or specialists.
If you keep patients healthy and spend $9,500 per beneficiary, you've generated $1,300 per patient in margin—$6.5M total. That funds better care coordination, more care managers, proactive interventions, and better technology.
Why this model exists:
REACH is designed for organizations ready to fully commit to population health management. The capitated model completely flips the incentive structure—profit comes from health, not from volume.
It also allows for more innovation. In fee-for-service, if you want to hire care coordinators or build remote monitoring programs, that's overhead reducing your margin. In capitation, those investments directly improve your bottom line by keeping patients out of expensive settings.
The challenge:
Total care capitation is genuinely risky. If you miscalculate, if you have unexpected high-cost patients, if you don't have the infrastructure to manage population health effectively, you can lose significant money. This isn't for organizations just starting their VBC journey.
But for those who master it, the model is incredibly powerful. You're not just sharing in savings—you're building a fundamentally different kind of healthcare organization.
Medicare Advantage: VBC Built Into Insurance
Medicare Advantage is different from ACOs. Instead of traditional Medicare, beneficiaries choose a private insurance plan (offered by UnitedHealth, Humana, CVS/Aetna, etc.) that contracts with CMS to provide all Medicare benefits.
How MA creates VBC incentives:
MA plans receive a capitated payment from CMS for each enrollee. They keep what they don't spend on care, so they have direct incentive to keep members healthy. But there's a crucial additional mechanism: Star Ratings.
CMS rates every MA plan on a 1-to-5 star scale based on roughly 40 quality measures across four domains:
- Staying Healthy: Preventive screenings, vaccines, annual wellness visits
- Managing Chronic Conditions: Diabetes care, blood pressure control, medication adherence
- Member Experience: CAHPS surveys measuring access, care quality, and customer service
- Plan Responsiveness: Complaints, appeals, customer service metrics
Plans with 4+ stars get bonus payments from CMS (up to 5% of revenue). For a large MA plan, that's hundreds of millions of dollars. These bonuses can be used to offer richer benefits (lower premiums, dental/vision coverage, gym memberships) that attract more members, creating a flywheel.
What this means in practice:
An MA plan managing 100,000 beneficiaries gets around $12,000 PMPM from CMS (risk-adjusted). That's $1.2B annual revenue. A 4-star rating versus 3.5-star might mean an extra $50M in bonus payments.
Those quality metrics become everything. Did patients get their annual wellness visits? Are diabetic patients getting A1C tests? Are members refilling medications on time? Are unnecessary ER visits being prevented?
The plan directly benefits from investing in care coordination, chronic disease management, and preventive care. This is why MA plans employ armies of care managers, offer free gym memberships, and call members to schedule preventive screenings.
Why this model dominates:
Medicare Advantage has grown from 31% to over 50% of Medicare beneficiaries. The combination of capitation (aligning incentives) and Star Ratings (ensuring quality) creates powerful forces for better care.
For providers, MA plans often offer higher reimbursement rates than traditional Medicare in exchange for more rigorous quality reporting and care coordination requirements. Many ACOs also contract with MA plans, creating additional VBC revenue streams.
The challenge:
MA plans are intensely competitive. Members can switch plans annually during open enrollment, so you need to continuously deliver value. Star Ratings can swing based on small changes in performance, creating pressure to optimize every measure.
Some critics argue MA plans cherry-pick healthy patients or upcode diagnoses to increase risk-adjusted payments. There's legitimate debate about how much of MA's success comes from better care versus better patient selection. But the directional incentive—getting paid more for keeping people healthy—is fundamentally right.
Commercial VBC: The Emerging Frontier
While Medicare leads, commercial payers (employer-sponsored insurance, private plans) are increasingly adopting VBC models.
Why commercial is different:
Commercial populations are generally younger and healthier than Medicare, so chronic disease management is less central. The opportunity is more about preventing expensive complications, managing high-cost conditions, and reducing unnecessary utilization.
Commercial VBC contracts vary widely:
- Shared savings arrangements similar to MSSP
- Bundled payments for specific procedures (joint replacements, cardiac surgery)
- Condition-specific capitation (diabetes, oncology)
- Full-risk contracts similar to ACO REACH
Example: Employer-sponsored VBC
Large employers increasingly contract directly with health systems for value-based care. Boeing, Walmart, and Intel have pioneered "centers of excellence" models where they send employees to high-quality providers who accept fixed bundled payments for complex procedures.
A joint replacement might cost $30,000 in traditional fee-for-service. In a bundled payment model, the provider gets a fixed $20,000 to cover everything—surgery, hospital stay, physical therapy, any complications. If they deliver high-quality care efficiently, they profit. If complications arise, they absorb the cost.
Why this matters:
Commercial VBC is growing faster than Medicare VBC in some markets. For providers, commercial contracts are more flexible and potentially more lucrative. For employers, VBC reduces costs while improving care quality for their workforce.
The challenge is fragmentation—every commercial payer has different contracts, different quality metrics, different reporting requirements. While Medicare programs have standardized measures and processes, commercial VBC requires navigating dozens of different arrangements.
The Infrastructure Gap
Here's the reality that jumps out when you understand these models: existing healthcare IT wasn't built for this.
Traditional EHR systems (Epic, Cerner, etc.) were designed for fee-for-service workflows: documentation, billing, order management. They bolt on VBC capabilities as afterthoughts—analytics dashboards, registry tools, care gap reports—but they're not architected around population health management.
What VBC organizations actually need:
1. Real-time risk stratification Which of your 50,000 attributed lives are at highest risk for expensive outcomes in the next 90 days? This requires integrating claims data, EHR data, social determinants, and predictive models—continuously, not in quarterly reports.
2. Care orchestration at scale How do you coordinate proactive interventions for thousands of high-risk patients? Who calls them? What gets documented? How do you track outcomes? This is a new workflow, not something that exists in traditional EHRs.
3. Quality measure automation MA Star Ratings have 40+ measures. MSSP has 23 quality metrics. Commercial contracts each have their own. Manually pulling data for quality reporting is unsustainable. You need automated tracking, gap closure workflows, and performance monitoring.
4. Financial visibility In fee-for-service, your revenue is your claims. In VBC, you need to understand: What's your current run rate against benchmark? Which providers or patient segments are driving costs? Where should you invest in interventions? Traditional financial systems can't answer these questions.
5. Patient engagement infrastructure VBC requires ongoing relationships with patients, not episodic encounters. Remote monitoring, medication adherence, appointment scheduling, health coaching—these need scalable infrastructure, not manual phone trees.
The organizations succeeding in VBC have built or bought new infrastructure. They've created dedicated population health teams, implemented specialized analytics platforms, and developed workflows that don't exist in traditional healthcare IT.
This is the opportunity space. Not replacing Epic, but building the VBC-native infrastructure that sits on top of existing systems and enables organizations to actually execute on population health management.
Where the Money Flows
Let's make this concrete. Where do VBC organizations actually generate margin?
The highest-leverage interventions:
-
Preventing avoidable hospitalizations - A single hospitalization costs $15,000-$50,000+. Preventing even a handful through better chronic disease management or proactive care pays for an entire care coordination team.
-
Reducing readmissions - 30-day hospital readmissions cost $15,000+ and often indicate poor care coordination. Transitional care management, medication reconciliation, and follow-up calls have enormous ROI.
-
Managing high-risk patients - The top 5% of patients drive 50% of costs. Intensive case management for this cohort—medication adherence, transportation assistance, behavioral health support—generates outsized returns.
-
Closing care gaps - Missed preventive screenings, annual wellness visits, diabetic eye exams—these aren't just quality metrics, they prevent expensive complications. A $200 screening might prevent a $50,000 emergency.
-
Optimizing medication adherence - Non-adherence to chronic disease medications leads to complications, hospitalizations, and ER visits. Medication therapy management programs cost pennies on the dollar compared to the complications they prevent.
The math is straightforward: invest proactively in the 5-20% of patients driving most costs, and the savings dwarf the investment.
Why This Matters for Builders
If you're building technology for healthcare, understanding these models changes what you build:
For MSSP ACOs: Tools that identify high-risk patients, coordinate care, and close quality gaps—but that integrate with fee-for-service workflows since providers still run on both systems.
For ACO REACH: Infrastructure for managing total cost of care—predictive analytics, care orchestration platforms, financial tracking—since these organizations are running real population health programs.
For Medicare Advantage: Star Ratings optimization tools, member engagement platforms, and quality measure automation—since these plans live and die by those metrics.
For commercial VBC: Flexible platforms that can adapt to varying contract structures, bundled payment analytics, and employer reporting—since commercial arrangements are more heterogeneous.
The model determines the problem. The problem determines the solution.
What's Next
We've covered how VBC models work mechanically. But here's what I keep coming back to: most healthcare AI companies are building for fee-for-service workflows.
Ambient scribes, prior authorization automation, revenue cycle management—these tools optimize the old system. They make fee-for-service less painful, but they don't change the fundamental game.
Value-based care creates completely different opportunities. Not automation of existing workflows, but entirely new capabilities that only matter when you're managing population health.
In the next post, I'll explain why this distinction matters—why building AI for fee-for-service is a different (and much smaller) game than building AI for value-based care.
Because if you're building for the wrong model, you're optimizing for a system that's already being replaced.