> ## Content Index > Fetch the complete content index at: https://agility.nexcurve.com/llms.txt > Use this file to discover other available public pages before exploring further. # The Sufficiency Razor - URL: https://agility.nexcurve.com/the-sufficiency-razor/ - Published: 2026-09-10T18:46:06.000Z - Updated: 2026-09-10T18:46:06.000Z - Description: A Healthier Population Is Not the Same as a More Affordable Health System - Author: J. Michael Eaton - Tags: health-system-executives, capital-efficiency There is an increasingly attractive argument in health care: obesity drives diabetes, cardiovascular disease, sleep apnea, joint disease and a long list of other expensive conditions. GLP-1 medications can materially reduce obesity. Therefore, broader use of GLP-1s could radically reduce health care spending. There is enough truth in that argument to make it compelling. But it does not resolve the underlying question of affordability. GLP-1s may prove to be one of the most important population health interventions of our time. For decades, we have tried to address obesity largely through behavioral modification, asking people to eat differently, exercise more and sustain those changes indefinitely. At population scale, the results have been limited. Now we have therapies capable of altering the biological mechanisms governing appetite and metabolism, with results that are difficult to dismiss. If those benefits hold, we should be much more aggressive about prescribing these medications to people for whom they are clinically appropriate. We may eventually think about obesity treatment less as an expensive specialty intervention and more as a form of population health infrastructure. But making Americans healthier is not the same thing as making American health care affordable. That distinction matters. ## There Are Two Cost Curves There are two dominant levers behind health care spending. 1. The first is how much illness people develop and how much care they require. 2. The second is what it costs to produce and deliver that care. GLP-1s can have a powerful effect on the first. If fewer people develop diabetes, suffer cardiovascular events, progress to severe sleep apnea or experience obesity-related disability, they should require less medical care. That is a meaningful clinical and economic benefit as it drives down utilization. But it does not explain why American health care is so expensive when care actually occurs. The United States spends almost $15,000 per person on health care each year, roughly twice what comparable wealthy countries spend. The difference is not simply that Americans consume twice as much care. In many categories, we do not. We do, however, pay substantially more for the services we receive. Obesity can therefore be a major contributor to disease burden without being the principal explanation for the cost structure of American health care. The economics of GLP-1s themselves make the point. The Congressional Budget Office has projected that broader coverage of anti-obesity medications would improve health and reduce spending on other medical services. It also found that those savings would not, under its assumptions, offset the cost of the drugs over the period studied. That is not an argument against using them. We do not require every effective treatment to generate a positive financial return. We treat cancer because people live longer. We replace hips because people can walk. We control hypertension because preventing strokes matters. If GLP-1s materially improve health, we should use them. We should simply stop asking them to solve a problem they were never designed to solve. > **What to Remember:** Making Americans healthier is not the same thing as making American health care affordable. ## The Sufficiency Razor Imagine the best-case scenario. GLP-1 use expands dramatically. Obesity falls. Diabetes declines. Cardiovascular events decrease. Billions of dollars in avoidable utilization disappear. That would be a remarkable achievement. People will still develop cancer, break hips, have babies, suffer infections, need joint replacements, experience strokes and require treatment for thousands of conditions unrelated or only partially related to obesity. The affordability question then becomes much more direct: > *What is the lowest level of clinical and resource intensity that can safely and reliably resolve the need?* That is where health care leaders need to become much more demanding. For every clinical need, we should be asking: 1. What actually needs to be done? 2. Who is the lowest sufficient level of clinical expertise required? 3. Where is the lowest sufficient setting in which the need can safely be resolved? Much of American health care has evolved around the reverse logic. We organize around hospitals, departments, specialists, professional boundaries and reimbursement rules, then fit the patient's problem into that architecture. The patient reaches the hospital, and the problem inherits the fixed cost of hospital care. The patient reaches a physician, and the problem consumes scarce physician time. An engineered system starts somewhere else. It starts with the problem and designs the least resource-intensive safe way to solve it. Increasingly, that should also mean using automation and artificial intelligence to remove administrative work that never required expensive human labor in the first place. I think of this as The Sufficiency Razor, an analog to Occam's Razor: use the lowest level of clinical intensity that can safely and reliably achieve the desired outcome. Health care is often built the opposite way. We design sites and services around the highest-intensity need that might arrive, while payment often rewards greater intensity and complexity. We have also become very good at escalating patients when more care is required and much less disciplined about de-escalating when it is not. The challenge at the front door is therefore not to under-triage. It is to detect the high-acuity patient without routing everyone into a high-acuity model. Escalate fast when needed. De-escalate just as fast when it is not. > **What to Remember:** Think of this as The Sufficiency Razor, an analog to Occam's Razor: use the lowest level of clinical intensity that can safely and reliably achieve the desired outcome. ## Cost per Resolved Clinical Need The Sufficiency Razor also exposes the limits of focusing only on unit price. If a service costs $1,000 today and we negotiate it down to $800, we have made progress. But if the patient's problem could have been safely resolved for $100 through a different pathway, we are still optimizing the wrong thing. The more useful economic unit may not be cost per encounter at all. It may be cost per resolved clinical need. That changes the conversation. Instead of asking how efficiently we perform today's transactions, we ask whether those transactions should exist in their current form. Instead of maximizing the productivity of every asset, health systems decide which asset should be used at all. The hospital becomes one node in the system rather than its organizing center. The physician becomes one level of expertise rather than the default labor unit. The office visit becomes one way to resolve a need rather than the default container for care. > **What to Remember:** The more useful economic unit may not be cost per encounter at all. It may be cost per resolved clinical need. That is not incremental cost management. It is re-engineering. And it allows us to hold two ideas at once: we should spend aggressively on interventions that materially improve health and reduce avoidable disease, including GLP-1s where clinically appropriate, while becoming far less tolerant of unnecessary clinical intensity in the care that remains. ## Where to Start Four questions a leadership team can assign, own and answer. **1\. Redesign the front door**: Can we detect the high-acuity patient without routing everyone into a high-acuity model? **2\. Build the de-escalation path**: We escalate well. Where is the explicit pathway, and whose job is it, to move a patient back down when the acuity does not hold? **3\. Change the unit of measurement**: Can we report cost per resolved clinical need alongside cost per encounter — and are we prepared for what that comparison shows? **4\. Take the administrative work out of clinical labor**: Which of our current tasks require expensive human judgment, and which are being done by expensive people only because no one has removed them? ## The Takeaway A healthier population should need less care. A more affordable system should use fewer resources to safely resolve the needs that remain. Those are different problems. We need to engineer both. Which one is your organization actually working on — and who owns the other one?