Primary Care Is Breaking Under Chronic Disease Burden. The Next Healthcare Platform Must Be Built for Longitudinal Care.

What the 2025 Primary Care Scorecard reveals about chronic care infrastructure — and why clinician-reviewed AI is the right response.

A national report on the state of U.S. primary care landed earlier this year — and if you work in healthcare, it deserves more than a quick scroll.

The 2025 Primary Care Scorecard, published by researchers at the Robert Graham Center and co-funded by the Milbank Memorial Fund and The Physicians Foundation, paints a clear picture: chronic diseases are rising, primary care investment remains disproportionately low, the physician workforce is shrinking, and patients are losing access to the continuous, coordinated care they need most.

The report’s core message is straightforward but sobering: primary care is being asked to manage more complexity with fewer resources. That is a structural problem — and it will not be solved by adding more dashboards or documentation tools.

It is the exact problem CareOS is being built to address.

CareOS Health is developing a clinician-reviewed chronic-care operating system — infrastructure designed to support care teams managing patients with multiple chronic conditions over time, not just at the point of the visit. Learn more at careoshealth.org.

The Chronic-Care Problem Is Operational, Not Just Clinical

When people discuss chronic disease, the focus tends to land on the clinical conditions themselves: hypertension, diabetes, obesity, sleep apnea, COPD, depression, heart failure, chronic pain.

But inside a working clinic, the challenge goes well beyond the diagnosis. It is operational.

A single patient may carry obstructive sleep apnea, obesity, hypertension, type 2 diabetes, anxiety, and chronic pain simultaneously. Their care touches a physician, a nurse, a care manager, a medical assistant, a behavioral health team, and sometimes remote monitoring data — all in separate systems, managed through fragmented workflows.

The problem is not that clinicians do not care. The problem is that the infrastructure was never built for longitudinal, multi-condition care at scale.

Most electronic health records were designed around documentation and billing. Most remote patient monitoring tools were built around device readings. Most care plans live as static documents. Most follow-up work still depends on staff manually remembering, tracking, and closing loops.

That is not sustainable as chronic disease burden grows.

What the 2025 Primary Care Scorecard Confirms

The 2025 Primary Care Scorecard makes clear that primary care carries a growing share of clinical complexity while receiving a small fraction of healthcare system investment. A few findings stand out for anyone working in chronic care or primary care technology.

Primary care remains underfunded relative to specialty and procedural care. The physician workforce is shrinking. More adults lack a consistent usual source of care. Chronic disease burden continues to rise. And physicians are increasingly being asked to manage not just diagnoses but complex overlapping conditions, patient messages, insurance reviews, mental health screening, social needs documentation, and follow-up gaps.

That last point is where the pressure is most acute. The modern clinician is no longer simply treating a diagnosis. They are managing a system of conditions, behaviors, risks, documents, tasks, messages, and care-team coordination — much of it without purpose-built infrastructure to support them.

CareOS is being built for that reality.

AI Should Not Replace Clinicians. It Should Remove Friction Around Them.

Healthcare AI tends to get discussed in sweeping terms: AI that diagnoses, AI that predicts, AI that replaces physicians. That framing misses where the real value lies.

The American Medical Association has reported that physicians see the biggest opportunity for AI in reducing administrative burden. That insight matters. The burden on clinicians is not just clinical decision-making — it is everything around the visit.

Reviewing scattered patient information. Summarizing history. Tracking care gaps. Preparing care plans. Coordinating tasks across care teams. Monitoring longitudinal progress. Identifying barriers to follow-through. Ensuring nothing goes to a patient without clinician sign-off.

This is where well-designed AI can help — if it is built correctly.

CareOS is not being built as an autonomous clinical system. It is being built as a clinician-reviewed chronic-care operating layer. The distinction is fundamental.

One Longitudinal Care Layer for Chronic Care Management

The CareOS vision is to give chronic-care clinics a single operating layer for managing their patient populations over time.

That means bringing together: patient registry and program enrollment, chronic condition context, structured questionnaires, care plans, remote monitoring signals, behavioral health overlays, medication-context review, patient barriers, social-risk factors, follow-up task management, clinician-reviewed summaries, and reporting support.

Instead of asking staff to manage chronic disease management through scattered screens and manual memory, CareOS aims to organize the work into a structured, clinician-reviewed care pathway.

Real patients do not present with one condition. A patient with obstructive sleep apnea often has obesity, hypertension, and depression. Poor sleep affects medication adherence. Medication non-adherence worsens chronic disease control. CareOS is designed around that connected reality, not the single-diagnosis visit model.

Why Chronic Care Needs an Operating System, Not Another Dashboard

Healthcare already has dashboards. Device dashboards. Billing dashboards. EHR dashboards. Quality reporting portals. Patient portals.

The problem is that dashboards do not automatically create workflow. Data sitting in a screen does not become a care plan, a follow-up task, or a coordinated handoff.

A true chronic care operating system should be able to answer practical questions that clinics already struggle with every day: Which patients are enrolled in which programs? Which have overdue follow-up? Which care plans need clinician review? Which patients have barriers affecting adherence? Which care-team member owns the next step?

These are not futuristic questions. They are the everyday operational realities that most healthcare software still does not answer in one place.

CareOS is being built to bring those pieces together into a care coordination and care plan automation layer — one that is organized around the whole patient, not just the last visit.

Safety Is Designed In, Not Added Later

Healthcare software should not casually automate clinical decisions. Any digital health platform operating in chronic care needs clear, consistent safety boundaries built in from the start.

CareOS is being designed with explicit constraints: no automatic diagnosis, no automatic medication changes, no automatic treatment escalation, no patient-facing release without clinician approval, no replacement of physician judgment.

AI in this context can draft, summarize, organize, and flag. The clinician reviews and approves. That is the difference between useful infrastructure and risky automation. In value-based care environments and chronic disease management programs, that approval gate is not a technicality — it is the product.

Why This Matters Now

The U.S. healthcare system is moving in a clear direction: more chronic disease, more multimorbidity, more behavioral health complexity, more remote patient monitoring, and more documentation pressure on already-stretched care teams.

At the same time, the primary care crisis documented in the 2025 Scorecard shows that those care teams are being asked to do more with less. The gap between what chronic care requires and what current infrastructure supports is widening.

That is where better infrastructure matters. Not AI hype. Not more screens. Practical, clinician-reviewed workflow infrastructure that helps care teams manage the whole patient across time.

The Direction Is Clear

The 2025 Primary Care Scorecard confirms what many clinicians already know: chronic care is becoming too complex to manage with fragmented systems alone.

The next generation of healthcare infrastructure needs to go beyond AI-generated notes, device dashboards, and static care plans. It needs longitudinal care infrastructure — designed around the way chronic disease actually works, not the way the last decade of health IT was built.

That is what CareOS is working toward: a clinician-reviewed chronic-care operating system that connects patients, programs, care plans, follow-up tasks, monitoring signals, and care-team coordination into one structured, safe workflow.

CareOS is still being built. But the direction is clear.

Primary care and chronic-care clinics do not need more noise. They need better coordination. Safer workflows. Tools that help care teams manage the whole patient over time.

Learn more about CareOS Health at careoshealth.org.

Sources

2025 Primary Care Scorecard — The Physicians Foundation / Milbank Memorial Fund / Robert Graham Center

AMA Survey: Physicians See AI’s Biggest Opportunity in Reducing Administrative Burden

[embed]CareOS | Care Management Software for Longitudinal Patient Follow-Up CareOS helps physician practices coordinate patient follow-up, care plans, documentation, billing-ready workflows, and…careoshealth.org