Topic bucket: Nursing shortage
The 2026 NSI staffing report landed and the number everybody quoted was the one that went the wrong direction. RN turnover ticked back up to 17.6 percent after dipping to 16.4 percent the year before. Becker's ran it under the usual headline about a workforce at a breaking point. Every CFO I know read it and reached for the same lever they always reach for. Spend more. More retention bonuses, more travel contracts, more recruiters, a sign-on package that would make a tech company blush.
I'll say the thing most of your executive team is thinking and won't put in a board deck. You are not short on nurses. You are wasting the ones you have.
Walk the floor of almost any practice or system we work with and watch what a registered nurse actually does in a day. A real chunk of it does not require a nursing license. Reading a blood pressure log. Reminding a diabetic patient to refill a prescription. Calling to check whether someone picked up their meds. Chasing a patient who missed a follow-up. That work matters. It keeps people out of the ER. But you are paying a clinician who took the NCLEX and carries real liability to do tasks a trained medical assistant, or in a lot of cases a well-built outreach agent, can do just as well. Then you act surprised when she burns out and leaves, and you book another $60,090 to replace her. That's the NSI figure for the cost of losing a single bedside RN. Multiply it across the 43 unfilled RN positions the report says the average hospital is carrying, and you see why the typical system bleeds somewhere between four and six million dollars a year on this. Verify those exact numbers against the full NSI report before you put them in front of your board, but the order of magnitude is right and you already feel it in your labor line.
Here's what nobody wants to say out loud because it sounds like you don't value nurses. Loading an RN with non-RN work is the opposite of valuing her. It's the fastest way to lose her.
So stop solving a design problem with a recruiting budget.
We redesigned our own model around exactly this, and it's the single biggest reason our clinician churn is a fraction of what these reports describe. Nurses do nurse work. Clinical judgment and escalation. The conversation that actually needs a license and a brain trained to catch the thing that's about to go wrong. Medical assistants run the routine touchpoints. And our AI agents handle the top of the funnel, the outreach and the reminders and the did-you-pick-up-your-prescription calls that used to eat a nurse's afternoon. The agent never makes a clinical decision. That line does not move at Welby and it never will. A human is always in the loop on anything that touches care. But an agent calling four hundred patients to get them scheduled is not a clinical decision. It's a phone call, and your nurse hates making it.
The math is not complicated. A travel RN runs you somewhere around $85 to $115 an hour all in once you count the agency markup, based on the 2026 bill rates floating around the staffing reports. A medical assistant doing the work that didn't need an RN in the first place costs a fraction of that. An outreach agent doing the work that didn't need a human at all costs almost nothing per contact and doesn't quit in eight months. You are not lowering the quality of care when you do this. You are raising it, because the nurse you kept now has the time and the headspace to do the part only she can do.
There's a second move that goes with the first one, and it's where most systems freeze up. Push more of this care into the home. The whole reason your nurses are drowning is that you've concentrated chronic-care management inside a building designed around acute visits and a twelve-minute appointment slot. Hypertension, diabetes, kidney disease, heart failure. None of that gets managed in twelve minutes in a clinic. It gets managed in the weeks between visits, in the patient's house, through monitoring and check-ins and somebody catching the blood pressure trend before it becomes a stroke. When you move that work home and staff it with the right person at each step, your in-clinic nurses stop being the overflow valve for everything the system can't otherwise absorb.
I know the objection. Scope of practice, state rules, the worry that you're handing clinical work to people who shouldn't have it. Good. You should worry about that. But there's a difference between practicing at the top of a license and practicing outside of one, and most systems have it backwards. They're so afraid of the second that they force everyone to practice at the bottom. Your RNs are doing MA work, your MAs are doing front-desk work, and your most expensive, hardest-to-replace people are the ones you're burning out first.
So what do you do Monday morning. Pull one report. Take your highest-volume chronic condition, probably hypertension, and have someone map every task in that care pathway against the lowest level of credential that can legally and safely do it. You will be uncomfortable with the gap. A big share of what your RNs touch every day is going to come back marked MA, or marked automatable. That gap is your shortage. Not the number in the NSI report. The gap between what you're paying for and what you're actually using.
Then start moving work down a level and out of the building, deliberately, one pathway at a time. You'll keep more nurses than any retention bonus will ever buy you.
The systems that win the next five years won't be the ones who recruited hardest. They'll be the ones who figured out their best nurses were never the constraint. The way they used them was.
