Raise the price, lose some patients: the break-even churn

How many patients can you lose before the increase stops paying?

Divide the dollar size of the increase by your contribution margin per visit at the new fee. That share of your visits can go before you are back where you started. On a $150 visit raised to $165, with $20 of cost that occurs only when the visit occurs, the margin is $145 and the increase covers itself until you lose about 10 percent of your volume.

The formula underneath is the one the Small Business Administration publishes for any business: fixed costs divided by price minus variable costs gives the number of units you must sell to break even 1Ref 1U.S. Small Business Administration (2026).Write your business plan.Supplies the break-even formula (fixed costs divided by price minus variable cost per unit) that this article adapts from generic units to patient visits in order to size how much volume loss a fee increase can absorb.. Units are visits here. Variable costs are what a visit itself consumes, and price minus variable cost is what each visit contributes to everything else.

Run it at both fees. The break-even count drops from fixed costs over the old margin to fixed costs over the new one, and when you ask what share of your visits that drop represents, the fixed costs divide out of both sides.

Your rent never enters the percentage.

It sets the count. A practice carrying $12,000 a month of fixed costs and a $145 margin needs about 83 visits a month to stand still; one carrying $6,000 needs about 42. Both can lose the same tenth of their volume to the price change before the increase stops paying for itself.

How much room the increase buys depends on how much of each visit is variable cost. A 10 percent increase on a $150 fee, at four different per-visit costs:

Every figure in that table is arithmetic on the two numbers to its left. None of it is a measurement of anybody's practice, and no federal or small-business source publishes one.

Sort the costs before you run it

Two buckets, and a third the SBA names for the costs that refuse both. Fixed costs are defined as those that do not change with the increase or decrease in production or services 2Ref 2U.S. Small Business Administration (2026).Calculate your startup costs.Supplies the definition of fixed costs as those that do not change with the increase or decrease in production or services, and of semi-variable costs as a mixture of both components, which set up the fixed/variable sort the break-even math requires.: rent, the malpractice premium, the EHR subscription, and your own draw when you are sizing the increase against take-home. Variable costs occur only when a visit occurs, and semi-variable costs mix the two.

Semi-variable is the one that costs people time. The SBA's definition is costs composed of a mixture of both fixed and variable components 2Ref 2U.S. Small Business Administration (2026).Calculate your startup costs.Supplies the definition of fixed costs as those that do not change with the increase or decrease in production or services, and of semi-variable costs as a mixture of both components, which set up the fixed/variable sort the break-even math requires., and in a solo practice that is the phone plan with overage, the utilities, the billing service charging a base plus a percentage. Split each at your best estimate and write the estimate beside it, so next year's version of this exercise starts from something.

Most of the raw material already exists on a form you file. IRS Publication 334's business expenses chapter names the categories: bad debts, car and truck expenses, depreciation, employees' pay, insurance, interest, legal and professional fees, pension plans, rent, taxes, travel and meals, and business use of your home 3Ref 3Internal Revenue Service (2025).Tax Guide for Small Business (For Individuals Who Use Schedule C).Supports that the IRS's own small-business guide names the deductible business-expense categories a solo practice already tracks, which is the raw material the reader re-sorts into fixed and variable buckets.. Schedule C puts them on numbered lines, contract labor at line 11 among them 4Ref 4Internal Revenue Service (2025).Instructions for Schedule C (Form 1040), Profit or Loss From Business.Supports that Schedule C already itemizes a solo practice's costs onto specific numbered lines, contract labor at line 11 among them, so the cost data the formula needs sits on a form the reader already files..

But neither document does the sort. The IRS groups costs by deductibility, which is a different question from whether a cost moves with volume, and one category can land in either bucket depending on how you buy it. A covering clinician paid by the session is variable. The same clinician on a salary is fixed.

Write down which bucket you put each line in. The answer you get is only as defensible as that list, and in six months you will not remember why the phone bill was split the way it was.

What the increase has to clear before it is a raise

Some of the increase only holds you level. Prices for medical care services rose about 2.7 percent over the year to July 2026 in the BLS consumer price index, from 637.425 to 654.347 on the not-seasonally-adjusted series 5Ref 5U.S. Bureau of Labor Statistics (2026).Medical care services in U.S. city average, all urban consumers, not seasonally adjusted (Series ID: CUUR0000SAM2).Supports the medical-care-services index moving from 637.425 in July 2025 to 654.347 in July 2026, roughly 2.7 percent, as the published anchor for the standstill portion of a fee increase, and the October 2025 value marked unavailable due to the 2025 lapse in appropriations.. Medicare's own annual move is smaller still. Whatever share of your new fee matches what the last year already took is a standstill, and only the rest buys the tolerance in the table above.

That index measures what urban consumers pay for medical care services. It does not measure what your practice pays for rent, malpractice cover or a covering clinician's hours, so read it as a direction and not as your own cost line.

The series also has a hole in it. October 2025 carries no value at all, marked unavailable due to the 2025 lapse in appropriations 5Ref 5U.S. Bureau of Labor Statistics (2026).Medical care services in U.S. city average, all urban consumers, not seasonally adjusted (Series ID: CUUR0000SAM2).Supports the medical-care-services index moving from 637.425 in July 2025 to 654.347 in July 2026, roughly 2.7 percent, as the published anchor for the standstill portion of a fee increase, and the October 2025 value marked unavailable due to the 2025 lapse in appropriations., so any twelve-month read spanning that month is reading across a gap rather than through one.

Medicare's number is smaller and better documented. The CY2026 Physician Fee Schedule final rule projects the qualifying-APM conversion factor rising $0.39, a 1.2 percent increase, and the non-qualifying factor rising $0.23, or 0.7 percent 6Ref 6Centers for Medicare & Medicaid Services (HHS) (2025).Medicare and Medicaid Programs; CY 2026 Payment Policies Under the Physician Fee Schedule and Other Changes to Part B Payment and Coverage Policies; Medicare Shared Savings Program Requirements; and Medicare Prescription Drug Inflation Rebate Program.Supports the CY2026 conversion-factor updates of 1.2 percent ($0.39) and 0.7 percent ($0.23) as the scale of Medicare's own annual move, and that CMS kept the Medicare Economic Index's 2006-based cost-share weights unchanged for CY2026 ratesetting rather than publishing a freshly reweighted figure.. There is no clean index figure to set beside those, either: for CY2026 ratesetting CMS kept the Medicare Economic Index's 2006-based cost-share weights unchanged rather than publishing a freshly reweighted number 6Ref 6Centers for Medicare & Medicaid Services (HHS) (2025).Medicare and Medicaid Programs; CY 2026 Payment Policies Under the Physician Fee Schedule and Other Changes to Part B Payment and Coverage Policies; Medicare Shared Savings Program Requirements; and Medicare Prescription Drug Inflation Rebate Program.Supports the CY2026 conversion-factor updates of 1.2 percent ($0.39) and 0.7 percent ($0.23) as the scale of Medicare's own annual move, and that CMS kept the Medicare Economic Index's 2006-based cost-share weights unchanged for CY2026 ratesetting rather than publishing a freshly reweighted figure., so anyone quoting you a current-year MEI percentage should be asked where it came from.

But you are not on a fee schedule.

A cash or membership practice sets its own number and can move it by considerably more than 1.2 percent in a year, which is the whole reason the tolerance question exists. Subtract the standstill portion first. Whatever remains above it is the part that has to pay for the departures it causes.

Some of the churn was leaving anyway

Patients leave a panel for reasons that have nothing to do with your fee, and the increase gets blamed for all of it unless you counted first. Pull the number of patients who stopped coming in each of the last four quarters before the notice goes out. Without that baseline, every departure in the three months after the letter looks caused by the letter, and you will read your own increase as a failure.

One floor under that baseline is geographic. The Census Bureau put the share of people who moved in a year at 8.7 percent in 2022, reported as not statistically different from the 8.4 percent rate in 2021 7Ref 7U.S. Census Bureau (Kristin Kerns-D'Amore) (2023).Change in Marital Status Became More Common Reason for Moving from 2021 to 2022, Housing/Neighborhood Improvement Reasons Declined.Supports the general-population annual mover rate of 8.7 percent in 2022, reported as not statistically different from the 8.4 percent rate in 2021, used only as a floor-level reminder that relocation removes some share of any list each year and explicitly disclaimed in the prose as not a churn benchmark for a medical practice.. That figure is drawn from the whole population. It says nothing about medical practices, and it should not be borrowed as a churn benchmark for one.

What it does is set the expectation that some share of any list becomes unreachable each year for reasons no fee change touches. Relocation is one. A job change, a plan change and a move into a different phase of care are others, and every one of them will arrive in the quarter after your letter looking like a response to it.

So count the departures before the notice goes out, and record a reason on every cancellation for two quarters either side of the change. A reason field is thirty seconds at the front desk, and it is the only thing that separates the patients who left over the price from the ones who were already going.

A monthly-fee panel asks the same question in a different unit. The churn arithmetic there runs on members lost against members added each month, and the DPC equation multiplies members by fee where a visit practice multiplies visits by price. The tolerance formula does not change. Only what you are counting does.

Sizing the increase when the math runs out

The arithmetic sizes the tolerance for any increase you name. It will not name the increase. No federal or small-business authority publishes a safe ceiling for one move, so the size is your own judgment, made against your waiting list, your referral flow, how long it has been since the last change, and where the fee sits next to the price floor your costs already set.

The two shapes fail differently. A small annual increase can be repeated next year if this one goes badly, and it takes several years to close a gap that has already opened. A large single increase closes the gap at once, and you get exactly one reading of how the panel responds, because you cannot un-send the letter and try a smaller number.

The other ceiling is the clinical one: the caseload math sets how many visits a week you can hold without the quality of them dropping, and a practice already at that ceiling is trading a price increase against a waiting list rather than against lost revenue. If the departing visits are refilled, the loss never reaches the revenue line, and the tolerance number becomes a floor.

Timing is a separate lever, and it costs nothing in the arithmetic. A change dated to the start of a plan year, announced far enough ahead that patients decide before their next visit rather than at the desk, reads differently to the person paying it. Give written notice well before the first affected appointment, and say plainly what the new fee is.

Run it on your own numbers this week

Four figures and twenty minutes. Take last year's Schedule C, your visit count for the same year, the fee you charge now and the fee you are considering, and the answer falls out of one division. Do it before you draft the notice, because the tolerance is what tells you whether the increase you have in mind is worth the letters it will cost you.

1. Total the costs that would not change if you saw ten fewer patients next month. Add your own draw if the question is take-home pay. 2. Total the costs that occur only when a visit occurs, divide by last year's visit count, and you have variable cost per visit. Split anything semi-variable at your best estimate and record the estimate next to it. 3. Subtract that from the fee you are considering. The result is your contribution margin at the new fee. 4. Divide the dollar size of the increase by that margin. The result is the share of your visits you can lose before the increase stops paying for itself. 5. Multiply that share by your current visit count to get a number of people, then set it against the departures you already counted for the last four quarters.

What comes out is a threshold: the point at which the increase stops paying for itself. It is not a prediction of how many patients will leave, and nothing published will give you that. Take the number to whoever prepares your return along with the fixed and variable split you used, because how the draw in step one is treated depends on how the practice is organized, and that is a conversation to have rather than a rule to read off a page.