Fifty-five percent of healthcare employees said they intend to look for, interview for, or switch jobs in 2026. HRSA projects the country will be short nearly 700,000 physicians, registered nurses, and licensed practical nurses by 2037. That is not a recruitment problem. That is a structural problem.
If you run a health system today, the temptation is to keep solving it the way you always have. Signing bonuses. Agency contracts. Retention stipends. More aggressive recruiting pipelines. None of that is wrong. All of it is insufficient.
The honest assessment is that the workforce math no longer closes. The chronic disease burden is rising, the population is aging, and the labor pool to care for that population is shrinking. You cannot hire your way out of that gap. The only realistic path forward is to extend the capacity of the clinicians you already have.
That is what AI is actually for in healthcare. Not chatbots. Not flashy demos. Capacity.
The nursing shortage gets the headlines, and it should. But the more dangerous trend underneath it is what burnout and turnover are doing to the patients you already manage.
The patients with the highest stakes are the ones with chronic conditions. Hypertension, diabetes, heart failure, COPD. These are conditions where outcomes are determined less by the doctor's visit and more by what happens in the 364 days between visits. When nurse case managers are stretched thin, that follow-up either does not happen at all or gets compressed into rushed phone calls that miss the early signals. Readmissions climb. Quality scores slip. Total cost of care goes the wrong direction.
This is not a theoretical risk. It shows up directly in your variable cost line and in your value-based contracts.
Three forces are pushing in the same direction, and none of them are going to reverse on their own.
First, the supply pipeline is constrained. Nursing programs are turning away qualified applicants because they cannot find enough faculty. The bedside-to-faculty pipeline is broken because experienced nurses are leaving the bedside before they ever consider teaching.
Second, the demand curve is going vertical. The population over 65 will grow significantly through 2040, and that is the population with the highest concentration of chronic disease and the highest utilization. More patients, more complexity, more touchpoints required.
Third, the economics of the bedside have changed. Hospitals are reporting elevated turnover with many professionals leaving bedside roles entirely rather than moving employers. They are not being recruited away. They are leaving the work.
You cannot recruit your way around any of those three forces, let alone all three at once.
Most of the AI conversation in healthcare has focused on documentation. AI scribes, ambient note generation, after-hours work reduction. Those tools are real and they matter. Research out of Duke University showed AI scribes producing a 20 percent reduction in note-taking time and a 30 percent drop in after-hours work.
But documentation is the easy part. The harder and more consequential opportunity is clinical workflow itself, and specifically the work of identifying which patients need attention right now versus which ones can wait until next week.
That is the unglamorous reality of chronic care. A nurse case manager with a panel of several hundred patients cannot actually look at every reading from every device every day. So patients who are starting to drift quietly, the ones whose blood pressure has crept up over the past ten days or whose glucose is trending poorly post-meal, do not get caught until the next scheduled touch. That is too late.
This is where AI changes the unit economics. When AI is actually paired with clinical workflow, it can flag rising-risk patients in real time, surface the readings that matter, and route them to the case manager with context. The clinician still makes the call. But she is making it on the patients who actually need her, not the ones whose data looks fine.
That is what augmentation looks like in practice. The clinician is still the clinician. The AI is the layer that decides who deserves her time today.
At Welby Health we have built our model around exactly this question. Licensed RN case managers paired with our MARKUS AI platform, monitoring chronic disease patients continuously between visits.
The outcomes data is clear. With cellular-enabled blood pressure cuffs and a clinical team responding to flagged readings, our hypertension patients see an average 20 percent decrease in blood pressure. With smart glucose monitors driving customized care plans, patients see over a 20 percent reduction in blood glucose in four weeks. Heart failure patients on our platform are 5.5 times more likely to adhere to their guideline-directed therapies.
Those are not numbers we hit by adding more nurses. Those are numbers we hit by giving every nurse the leverage of a system that watches every patient continuously and tells her where to spend her hour.
Equally important to the CFO sitting next to the CMO: this model generates new billable revenue under CCM, RPM, and TCM codes. The 2026 CMS Final Rule expanded those opportunities further, lowering time thresholds for monitoring and treatment management and creating new pathways for shorter monitoring periods to be billable. The ground has shifted in favor of organizations that have a real chronic care infrastructure in place.
If you are running a health system or a provider organization, you are going to be asked the same question from your board over the next twelve months. What is your plan to deliver more care with fewer clinicians? The answer is not going to be more recruiters.
The answer is a clinical operating model where AI is doing the work of triage, surveillance, and prioritization at scale, and your clinicians are spending their time on the patients and decisions that actually need a human. That is not a future state. That is buildable today, and the organizations that build it now will have a meaningful structural advantage over the organizations that wait.
The workforce shortage is not going to fix itself. The patients are not going to get less complex. The contracts are not going to get more forgiving. The only variable left to move is leverage. And leverage, in 2026, is AI plus a clinician.
Healthcare is heading into a structural workforce shortfall that recruiting will not solve. HRSA projects a shortage of nearly 700,000 physicians and nurses by 2037, while 55 percent of healthcare employees plan to job-search in 2026. The only realistic path forward is extending the capacity of the clinicians you already have through AI. The highest-leverage place to apply that AI is chronic care, where outcomes are decided between visits and workflow triage is the bottleneck. Welby Health's model of RN case managers paired with our MARKUS AI platform is delivering 20 percent blood pressure reductions, 20 percent glucose reductions in four weeks, and a 5.5x improvement in heart failure therapy adherence. Health system CEOs who build this clinical operating model now will have a structural advantage over those who try to hire their way out.
