Custom physical therapy software in Baltimore: four clocks, one therapist

The clinics this is actually about

There is a version of this article that opens with the national outpatient rehabilitation market and a compound annual growth rate, and it would tell a Baltimore clinic owner nothing. So here is the local picture instead, counted rather than estimated.

The Census Bureau’s County Business Patterns file for 2023 — the most recent county‑level release — puts 1,029 therapy establishments with paid employees in Maryland under NAICS 621340, the code for offices of physical, occupational and speech therapists and audiologists. Between them they employ 12,294 people and carry an annual payroll of $762.3 million. That caveat about paid employees matters here more than in most trades: a solo therapist with a treatment room, a plinth and no staff does not appear in this file at all, and Maryland has a lot of them. What follows is therefore a count of practices that have made the jump to being an employer.

Baltimore City has 89 of them. They employ 1,586 people and pay out $81.1 million a year. And here is the part that reframes everything: 51 of those 89 have fewer than five employees, while the citywide average is 17.8 employees per establishment. Both of those facts are true at once because the city’s therapy employment is barbell‑shaped. A small number of large hospital‑affiliated outpatient networks — MedStar, Johns Hopkins Rehabilitation Network, University of Maryland — sit at one end, and at the other sit fifty‑odd independent practices where the owner is also a treating therapist and the front desk is one person who has been there eleven years.

The ring around the city is where it gets genuinely interesting. Montgomery County has 186 therapy establishments and Baltimore County has 175 — so Montgomery has eleven more practices. But Baltimore County’s practices employ 3,901 people against Montgomery’s 1,454. Eleven more businesses, and two and a half thousand fewer people inside them. Baltimore County’s annual payroll in this code is $344.4 million, more than four times Montgomery’s $77.2 million, which works out to roughly $88,000 per employee against Montgomery’s $53,000. Those are two completely different industries wearing the same NAICS code. Montgomery is a county of small independent and part‑time practices; Baltimore County is a county of large employed‑therapist networks. If you are reading this from a practice in Towson or Timonium, your competitor down the road is probably not another owner‑operator.

The people inside those buildings are counted separately, and better, by the profession’s own regulator. The Maryland Board of Physical Therapy Examiners published its 2025 Maryland Physical Therapy Workforce Report from renewal and survey data collected between March and November 2025, and it is a genuinely useful document. As of 30 June 2025 the Board licensed 9,385 active practitioners — 7,091 physical therapists and 2,294 physical therapist assistants. Maryland has 150 practitioners per 100,000 people against a national 130, so the state is comparatively well covered.

Except that the coverage is lopsided in a way that ends up mattering to software. Maryland has 114 physical therapists per 100,000 against a national 92 — well ahead. It has 37 physical therapist assistants per 100,000 against a national 39 — slightly behind. Maryland leans on the more expensive licensee. The Board’s own wage table puts the 2024 annual mean at $107,690 for a physical therapist and $67,640 for an assistant. A state that staffs with proportionally more therapists and fewer assistants is a state where every hour of therapist time is more valuable and more constrained, which is exactly the condition under which a per‑head software meter behaves worst and a scheduling error costs most.

One more figure from the same report, because it tells you who this article is for: a third of surveyed practitioners work in private physical therapy offices not affiliated with a hospital or health system, and 88% of all practitioners are in direct clinical practice rather than administration, teaching or research. The independent sector is a third of the profession and almost all of it is treating patients. Nobody in it has a spare afternoon to be an EMR administrator.

Five meters for the same therapist

We went looking for what rehab therapy software actually costs on 24 August 2026, and the first thing worth reporting is how many of the URLs simply are not there any more.

Prompt Health — formerly Prompt EMR, and one of the fastest‑growing names in the category — returns a 404 on both /pricing and /plans. The old promptemr.com/pricing address redirects to the new domain and then 404s there too. The three‑tier per‑provider figures that circulate for Prompt come from a competitor’s blog, not from Prompt, and we are not going to reprint a rival’s numbers as though they were the company’s own. Raintree’s pricing URL: 404. ClinicSource’s: 404. TheraOffice’s redirects to a Netsmart product page, which is what an acquisition looks like from the outside. Fusion’s redirects to Ensora Health, the rebranded Therapy Brands. Three of the products on a typical 2023 shortlist now live inside two larger companies, a fourth has taken its prices down, and a clinic owner who spent a weekend comparing them last year is comparing ghosts.

Then there is WebPT, which is the biggest name in the category and does publish a pricing page. It has three tiers — Starter, Enhanced and Ultimate — each broken into Care, Engage, Monetize and Plus modules, with a complete feature matrix underneath. It contains no dollar figure anywhere. Every call to action is “Book a Demo.” And in its own FAQ, on its own pricing page, it answers the question “How much does WebPT cost?” like this:

“Pricing is based on your selected plan and any customization you may need, based on your practice’s unique needs. You can set pricing up as per provider per month or per visit.”

Read that twice. The largest vendor in outpatient rehab offers two structurally different meters for the same product — one that charges by headcount and one that charges by throughput — and publishes the rate for neither. Those two meters produce wildly different bills for the same clinic. A practice with six part‑time therapists and 900 visits a month is on the wrong side of one of them and the right side of the other, and there is no way to know which without a sales call. I do not think that is a trick; per‑visit pricing is a genuinely useful option for a growing practice with a lot of part‑timers. But a meter you cannot see is a meter you cannot plan against.

My favorite finding of the sweep is SPRY, which publishes a widely‑read blog post titled PT EMR Pricing Comparison 2026 comparing its competitors’ prices, and whose own /pricing page contains no dollar figure at all. That is not a scandal. It is just a very clear illustration of what the category rewards.

Four companies do publish real ladders, and they are worth studying closely, because between them they meter four completely different nouns.

I want to be fair to all four of the companies that publish, because publishing is the harder choice and each of these ladders is defensible on its own terms. PtEverywhere’s is the most honest structure in the category and it is not close. It publishes every number, it declines to charge for the front desk at all, and it caps non‑therapist accounts at the number of therapists so the bill cannot quietly inflate on administrative headcount. If you want a clean answer to “what will this cost me at nine therapists” you can get it without speaking to anyone: nine therapists is four accounts at $75 and four at $60, so $540 a month for the eight additional accounts, on top of the base plan.

Jane is the only vendor in this sweep that has noticed that therapists are not interchangeable units. Its half license is a real accommodation for a real staffing pattern, and its definition is admirably specific: a part‑time practitioner is one who books fewer than 24 hours a week. Jane is also careful to say what it means by the word, in a parenthesis I appreciated:

“This has nothing to do with your professional license — it’s just a software term for a practitioner profile that can take bookings.”

Which is precisely the problem in one sentence. In this industry the word license means two entirely different things: a seat in a database, and a document from the Maryland Board of Physical Therapy Examiners that determines whether a human being may lawfully touch a patient. One of them costs $35 a month. The other costs $325 every two years and takes a doctorate. Your software tracks the first one perfectly and, in every product we looked at, tracks the second one as a text field with an expiry date, if at all.

And HENO deserves credit for publishing a rate card most of its competitors will not, even though the rate card itself is the most complicated in the category — seven different meters running at once. One line on it is worth pulling out, because it connects to the most important number in this article. HENO’s published add‑on list meters e‑claims for exactly one payer — workers’ compensation. It is $0.75 per workers’ comp e‑claim on the $149 plan and $0.35 on the $225 plan, unlimited only on the $299 plan, and no other payer’s claims carry a per‑claim charge at all. Hold that thought for two sections.

The number the meter never sees

Every meter above counts a noun. A provider. A license. A concurrent seat. A location. An active patient. A visit.

Medicare counts something else entirely, and it is the only count that determines what the clinic gets paid. Under the eight‑minute rule — the methodology in Chapter 5 of the Medicare Claims Processing Manual, which CMS left unchanged for 2026 — you total the minutes of timed, one‑on‑one treatment, divide by fifteen, and bill an additional unit whenever eight or more minutes remain. That produces a ladder every therapist knows by heart:

A per‑provider subscription is completely blind to that ladder. It costs exactly the same whether the therapist stops at twenty‑two minutes or twenty‑three. The difference between those two minutes is about $30 under Medicare and $50.20 under Maryland workers’ compensation, and it recurs on every timed visit, every day, in a clinic where the schedule is built by a front desk person looking at a grid of colored rectangles that knows nothing about any of this.

Now put the second clock next to it, because this is where Maryland stops being like everywhere else.

The part no national platform models

Three patients per clinical treatment hour, averaged across a calendar day

Maryland regulates therapist throughput directly. COMAR 10.38.03.02(A)(2)(b)(i) requires that a physical therapist provide:

“Physical therapy services to not more than an average of three patients per clinical treatment hour per calendar day, excluding group therapy”

The identical sentence appears again at 10.38.03.02(B)(1)(h)(i) for physical therapist assistants, so the cap follows the treating clinician rather than the license class. Most states do not have this. Maryland does, and it is written into the standards of practice rather than into a payer contract, which means it is a licensing obligation rather than a billing one. Two details in that sentence do most of the work. It is an average across a calendar day, not a hard ceiling on any single hour — so a busy 10am does not violate it if a quiet 3pm balances it out, and knowing whether you are compliant requires a calculation over an entire day’s schedule that no scheduling grid performs. And group therapy is expressly excluded, which quietly makes group programming the single legitimate lever for increasing throughput without touching the cap at all.

Here is what happens when you put Maryland’s cap and Medicare’s ladder in the same room. Take one therapist and one hour of one‑on‑one treatment time — sixty therapist minutes, identical in every scenario — and vary only how those minutes are divided between patients.

Read the fourth column again. Maryland’s legal maximum is arithmetically the worst place to sit. Three twenty‑minute patients and two thirty‑minute patients consume the same sixty minutes of the same therapist, and one of them bills three units while the other bills four — because twenty minutes divides into fifteen once with five minutes left over, and five is less than eight, so five minutes of real treatment on every one of those three patients is simply not billable. Fifteen billable minutes an hour, evaporated by the shape of the schedule.

On one therapist working thirty‑five clinical hours a week, that one unit an hour is about 35 units, or roughly $1,060 of Medicare allowed charges a week, decided entirely by how the front desk chose to divide the day. It is not fraud, it is not upcoding, and it does not involve treating anybody differently — the patients in the thirty‑minute column are getting more care, not less. It is a scheduling shape, and it is invisible to every product we priced, because the product is counting therapists and Medicare is counting minutes.

I want to be careful here, because this is the kind of number that gets misused. No clinic runs at a hundred percent timed one‑on‑one codes for thirty‑five hours; there are evaluations, untimed modalities, no‑shows and documentation time in every real week. The point is not that a practice is losing exactly $1,060 a week. The point is that the largest single variable in a rehab clinic’s revenue per therapist hour is a scheduling decision that no rehab scheduling product surfaces, computes or warns about — and in Maryland there is a regulation sitting on top of it that pushes practices toward the least efficient shape.

Thirty days, or a written reason

Two subsections further down the same regulation, at 10.38.03.02(A)(2)(g), is an obligation that is far more operationally awkward than it looks. A Maryland physical therapist must reevaluate the patient as the condition requires:

“but at least every 30 days, unless the physical therapist, consistent with accepted standards of physical therapy care, documents in the treatment record an appropriate rationale for not reevaluating the patient.”

That is a rolling thirty‑day clock, running independently for every active patient, with an escape hatch that is itself a documentation event. It is not the Medicare recertification clock, which is a different period on a different trigger. It is not the plan‑of‑care expiry your EMR probably does track. It is a third clock, it is Maryland‑specific, and the compliant behavior when you miss it is not “do the reevaluation late” but “write down why you did not.”

A practice with 400 active patients has 400 of these clocks running at once. The report that would answer “who is at day 27 today and has no documented rationale” takes about ninety minutes to build against data you already have. We have never seen a clinic that had it.

Two conversion factors for the same fifteen minutes

Now the number that reframes the entire economics of a Baltimore rehab practice, and it comes from a one‑page PDF the Maryland Workers’ Compensation Commission published on 1 November 2025.

Under COMAR 14.9.08, the Commission sets its own Maryland Specific Conversion Factor — the multiplier applied to relative value units to produce a maximum reimbursable amount. For services on or after 1 January 2026, the factor for medical services and treatment is $55.20. Medicare’s 2026 conversion factor, finalized in the CY 2026 Physician Fee Schedule rule, is $33.40 for clinicians who are not qualifying alternative payment model participants and $33.57 for those who are.

These are not different fee schedules built on different logic. We downloaded the Commission’s own 2026 fee guide workbook, and its single worksheet is named PPRRVU2026_Jan_nonQPP — it is the Medicare relative value unit file. Maryland workers’ compensation prices exactly the same work, with exactly the same relative values, and multiplies by a number 65% larger. Here is what that does to the codes an outpatient orthopedic clinic actually bills, straight out of the Commission’s published maximum reimbursable amount column:

Two things in that table deserve a paragraph each.

The first is the evaluation row. Maryland workers’ compensation pays exactly $169.38 for a low‑complexity, a moderate‑complexity and a high‑complexity physical therapy evaluation. Identical to the cent. CPT split the evaluation code into three tiers in 2017 precisely so that a complex patient could be billed differently from a simple one, and every EMR in this category now walks the therapist through a decision tree of body regions, comorbidities and clinical presentation stability to determine which of the three to pick. Against this payer, that decision tree is worth nothing. The clinician time it consumes is real and the revenue difference is zero.

The second is the arithmetic on a whole visit. A common orthopedic pattern — two units of therapeutic exercise plus one of manual therapy — comes to $89.73 under Medicare and $148.31 under Maryland workers’ compensation. Identical work, identical forty‑five minutes, identical room, identical therapist. A 65% difference that depends only on why the patient got hurt.

And now the line I asked you to hold. The only claims HENO meters per claim are workers’ compensation claims — $0.75 each on the $149 plan, $0.35 on the $225 plan, unlimited on the $299 plan — while no other payer’s claims carry a per‑claim charge at all. It is the exact inverse of where the value is: the one payer worth roughly $58 more a visit is the only one the software charges you to invoice. That is not malice; it reflects a genuinely more manual clearinghouse path for comp claims. It is simply what happens when a product is priced on its own costs rather than on your economics, which is what every subscription is.

The practical version of this section is short and I would say it on a call. If you take workers’ compensation work in Maryland, the payer mix on your schedule is worth more than any software decision you will make this year, and the report that shows you next week’s booked hours split by payer with expected reimbursement attached is a report almost nobody has.

The 499 therapists whose license is not Maryland’s

As of 30 June 2025, 499 physical therapy practitioners licensed in another state held a compact privilege to practice in Maryland, issued by the Physical Therapy Compact Commission, which now has 34 member states actively issuing. They came from 31 different states; Virginia alone accounts for 22% of them. Running the other way, 678 Maryland licensees — 7.2% of the workforce — hold a compact privilege somewhere else, most often Virginia (24%), Washington DC (13%) and Delaware (9%).

A compact privilege is not a Maryland license. It is an authorization that derives from a license in another state and lapses when that home‑state license does, on a renewal cycle your Maryland practice does not control and cannot see. Which means the credentialing question for a Baltimore clinic employing a compact‑privilege therapist is not “when does their license expire” but “when does a license in Pennsylvania that I have never seen expire, and did they renew the privilege afterwards?”

We have not found a rehab EMR that models this at all. They model a license: a number, a state, an expiry date, sometimes an upload. A compact privilege is a different object with a different parent, and it matters most in exactly the situation where an error is worst — a metro area twenty minutes from the DC line and forty from Delaware, where cross‑border practice is ordinary and telehealth makes it invisible. Only 18% of Maryland licensees report providing care by telehealth, and telehealth has only been in COMAR since 2023, so this is a young problem that is going to get bigger rather than smaller.

Twenty‑four hours, thirty hours, and the Affordable Care Act

One last clock, and it is my favorite because it is so small and so revealing.

Jane defines a part‑time practitioner as one who books fewer than 24 hours a week, and charges a half license accordingly. The Maryland Board’s own 2025 workforce survey defines full‑time as 30 hours or more, and part‑time as 15 to 29 — explicitly following the Affordable Care Act definition. So a therapist booking 26 hours a week is full‑time to your software, part‑time to the state’s workforce statistics, and part‑time under the federal rule that governs whether you owe them health coverage. Three definitions, three thresholds, one person.

None of these definitions is wrong. Jane’s is a fair proxy for how much a practitioner uses the product; the ACA’s is a benefits rule; the Board’s is a research convention. But if you are staffing a clinic around part‑time therapists — and in Montgomery County, at 7.8 employees per establishment, almost everyone is — you are managing to three different lines at once, and the only one your software will ever remind you about is the one that determines its own invoice.

Where the cash side actually lives

Everything above is about the insured side of a rehab practice, and it is the part that gets all the attention. The commercially interesting half is somewhere else, and Maryland is unusually well set up for it.

Maryland passed the first direct access law in the United States in 1979. A licensed physical therapist here may evaluate and treat without a physician referral, subject to the referral obligations that arise if the condition does not respond as expected. Insurers can still impose their own requirements, so direct access is a licensing answer rather than a payment answer — but commercially it means something specific and underused: the front door of a Maryland physical therapy practice is its own website, not a physician’s fax machine. A patient with a sore shoulder in Hampden can decide, at 9pm, to start care, and the only thing standing between that decision and a booked evaluation is whether you own a page that can take it.

Then there is what happens at the other end of the episode. The CY 2026 KX modifier threshold is $2,480 for physical therapy and speech‑language pathology combined, with a second threshold at $3,000 where claims may be pulled for targeted medical review. At roughly $90 of Medicare allowed charges for a typical orthopedic visit, $2,480 arrives at about the twenty‑eighth visit — three visits a week for nine weeks, which is one rotator cuff or one total knee. It is not a coverage cliff; you keep billing, you attest with the modifier, and the documentation carries it. But it is the point at which the conversation with the patient changes, and it is also the point at which a great many patients who still want to train stop being a clinical question and become a commercial one.

What sits in that space is a real business, and almost none of it touches a payer:

Cash‑pay clinical services. Dry needling is the obvious one, and Maryland treats it as a distinct credential — the Board charges a one‑time $100 dry needling registration fee under COMAR 10.38.07.02, so it is a licensed, differentiated, defensible service rather than an upsell. Running gait analysis, return‑to‑sport testing, bike fitting, pre‑operative conditioning and post‑discharge maintenance programs all sit in the same category: priced by you, sold by you, delivered in a room you are already paying rent on.

Group programming, which is worth calling out twice. COMAR expressly excludes group therapy from the three‑patients‑per‑hour average, which makes it the one throughput lever Maryland regulation leaves open. It is clinically well supported for the post‑acute population, and sold as a cash membership rather than billed as 97150 it escapes both the cap and the fee schedule at once.

Retail that you are uniquely qualified to sell. A rehab clinic knows things about its customers that no other retailer of the same products knows: the diagnosis, the exact home exercise program, the resistance level of the band that was prescribed, the brace size that was fitted, the date the plan of care ends and the exercises the patient is supposed to still be doing six weeks later. Bands, blocks, rollers, straps, braces, ice and heat, and the home equipment for whatever program you just wrote. Every one of those is currently being bought from a national retailer that knows none of it.

This is the argument I make most often on calls with rehab owners, and it is not really about e‑commerce. It is about the fact that a physical therapy practice has, sitting in its EMR, the single best‑qualified customer list in retail — people whose exact needs are documented, whose compliance you can measure, and who trust you more than they trust any brand — and the EMR is architecturally incapable of selling anything to it. That is not a criticism of the EMR. Clinical documentation systems should be clinical documentation systems. It just means the commercial layer has to be somewhere else, and it may as well be somewhere you own.

What custom actually costs

We publish our prices, which after a sweep that produced four 404s and three redirects feels almost combative. Here they are against the problem I have just described.

What we would actually build

If a four‑therapist practice in Canton or Towson called us tomorrow, we would not propose replacing the EMR. We would propose building the layer beside it, and we would build it in this order.

First, the schedule‑shape report, because it pays for itself fastest and because it is the one thing nobody has. It reads appointments, therapists and treatment durations out of whatever system you already run, and it answers two questions every morning: what is each therapist’s average patients per clinical treatment hour for today under COMAR, and how many billable units will this shape produce against how many it could produce. That is a report, not a platform. It takes a week.

Second, the payer‑mix overlay, which is the same schedule with money on it. Workers’ compensation, Medicare, commercial and cash slots stop being interchangeable rectangles and start carrying their expected reimbursement, so that a cancellation in one is visibly not the same event as a cancellation in another, and so that the person filling a gap at 2pm knows which waiting‑list patient to call first.

Third, the clocks. Thirty‑day reevaluations with the documented‑rationale exception. The KX running total per patient per calendar year. Plan‑of‑care expiry. Credential expiry, with compact privileges modeled as derived from a foreign license rather than as a license of their own. Every one of these is a date arithmetic problem, every one of them is currently being managed in somebody’s head or in a spreadsheet called Recerts FINAL v3, and every one of them is trivially automatable once the data has somewhere to live.

Fourth, the store, and by then you know exactly what belongs in it, because the first three have told you which services your patients actually want after the insured episode ends.

None of that replaces WebPT or Jane or Prompt. All of it sits beside them, reads from them, and is yours.

When you should not call us

There are several situations where the honest answer is that you should keep renting, and we would rather say so here than on an invoice.

• You are a solo therapist or a two‑person practice. At that size a published ladder like PtEverywhere’s or Jane’s is genuinely good value, the scheduling arithmetic is something you can hold in your head, and a $3,500 sprint is better spent on marketing than on software.
• You are unhappy with your EMR and think custom means replacing it. It does not, and you should not want it to. A certified clinical documentation and claims system is a decade of regulatory compliance work; nobody should rebuild that for a four‑therapist clinic and we will decline if you ask.
• Your problem is billing performance, not software. If your days in accounts receivable are ninety and your denial rate is fifteen percent, a better report will not fix that. A better biller will.
• You are hospital‑affiliated. If you are one of Baltimore County’s twenty‑two‑employees‑per‑site network clinics, your systems decisions are made two levels above you and no amount of good arithmetic changes that.
• You have not yet asked your current vendor for its rate in writing. Do that first. It is free, it takes one email, and for the two largest names in this category the answer is genuinely not on the internet.

How we work

We are a small studio and we are deliberate about it. You talk to the people writing the code — there is no account manager between you and the build. The price is fixed and public before we start, you pay half to begin and the balance when it ships, and the source code, the repositories, the keys and every account are yours at the end. We build on a deliberately boring stack, React and TypeScript over Node and Postgres with Stripe for payments, deployed on infrastructure you own, so that any competent developer can pick it up after we hand it over. If you want to see what that looks like before you talk to us, the demos are real applications rather than screenshots, and the pricing page has the numbers on it.

And if the honest answer on the call is “keep your EMR, fix the shape of your schedule and put a store on the front of it,” that is what you will hear. We have said a version of it to a florist, a towing company and two dental practices this year.

Questions we get from Baltimore rehab practices

How much does physical therapy EMR software cost in 2026?

Anywhere from about $54 a month for a solo practitioner to $299 plus a stack of per‑therapist add‑ons for a multi‑location clinic — for the platforms that publish a figure at all, and the largest names in the category publish nothing. Checked on 24 August 2026: PtEverywhere publishes a complete ladder — additional therapist accounts at $75 each from two to five, $60 each from six to ten, $50 each at eleven and above; non‑therapist accounts at $39, capped so they cannot exceed the therapist count; administrators and receptionists free; revenue cycle management at $39 per therapist per month. Jane publishes three plans at $54, $79 and $99 a month with additional practitioners charged as a full or half license. HENO publishes three plans at $149, $225 and $299 plus per‑therapist, per‑location, per‑text, per‑fax, per‑claim and per‑active‑patient add‑ons and an optional $1,500 data import. Practice Perfect publishes a per‑concurrent‑user rate with declining tiers at $15.00 for four to ten licenses and $10.00 at eleven and above. WebPT publishes a three‑tier feature matrix with no dollar figure on it. Prompt Health’s pricing and plans URLs both returned 404s. SPRY’s own pricing page contains no dollar figure while its blog compares everyone else’s. Raintree’s and ClinicSource’s pricing URLs returned 404s, TheraOffice’s redirects to Netsmart, and Fusion’s redirects to Ensora Health.

Why do physical therapy software vendors charge per provider when Medicare pays per unit?

Because a head is easy to count and a minute is not. Every meter in this category counts a noun: a provider, a license, a concurrent seat, a location, an active patient, a visit band. Medicare counts something else. Under the eight‑minute rule in Chapter 5 of the Medicare Claims Processing Manual, timed one‑on‑one minutes are summed and divided by fifteen, with an extra unit whenever eight or more minutes remain — so 8 to 22 minutes is one unit and 23 to 37 is two. The twenty‑third minute of a treatment is worth an entire unit and the twenty‑fourth through thirty‑seventh are worth nothing. A per‑provider subscription is blind to that distinction: it costs the same whether the therapist finishes at 22 minutes or 23, and the difference between those two minutes, priced at CPT 97110 in 2026, is about $30 under Medicare and $50.20 under Maryland workers’ compensation.

How many patients per hour can a physical therapist treat in Maryland?

An average of three per clinical treatment hour per calendar day, excluding group therapy. The rule is COMAR 10.38.03.02(A)(2)(b)(i) for physical therapists, and 10.38.03.02(B)(1)(h)(i) repeats it word for word for physical therapist assistants. Two details matter operationally. It is an average across a calendar day, not a hard ceiling on any single hour, so knowing whether you are compliant requires a calculation over a whole schedule rather than a check at the moment of booking. And group therapy is expressly excluded, which makes group programming the one legitimate way to add throughput without touching the cap. No national EMR we checked computes this figure, for the understandable reason that forty‑nine other states do not have it.

Does Maryland pay more than Medicare for physical therapy?

Maryland workers’ compensation does, by a wide margin, and the reason is a single published number. The Maryland Workers’ Compensation Commission sets its own Maryland Specific Conversion Factor under COMAR 14.9.08; for services on or after 1 January 2026 the factor for medical services and treatment is $55.20. Medicare’s 2026 conversion factor is $33.40 for clinicians who are not qualifying alternative payment model participants and $33.57 for those who are. The Commission’s own 2026 fee guide worksheet is titled PPRRVU2026_Jan_nonQPP, so both schedules price the same relative value units and differ mainly in the multiplier. In published Commission dollars that puts therapeutic exercise (97110) at $50.20 per fifteen‑minute unit, manual therapy at $47.90, therapeutic activities at $60.90 and neuromuscular re‑education at $56.56. All three evaluation codes — 97161, 97162 and 97163 — pay exactly $169.38, so the complexity level your EMR makes you choose is worth nothing to this payer.

What is the KX modifier threshold for 2026 and how many visits is that?

The CY 2026 KX modifier threshold is $2,480 for physical therapy and speech‑language pathology combined, and a separate $2,480 for occupational therapy. Above it you keep billing, appending the KX modifier to attest that services remain medically necessary and that the documentation supports it. A second threshold at $3,000 is the point at which claims may be selected for targeted medical review. Turned into visits: a common orthopedic pattern of two units of therapeutic exercise plus one of manual therapy comes to roughly $90 of Medicare allowed charges a visit, so $2,480 arrives at about the twenty‑eighth visit — three visits a week for nine weeks. That is one rotator cuff or one total knee, not an unusual course of care, which is why the threshold is best treated as a documentation event rather than a coverage limit.

Do I need a referral for physical therapy in Maryland?

Not to start. Maryland passed the first direct access legislation in the United States in 1979, and a licensed physical therapist may evaluate and treat without a physician referral, subject to the referral obligations that arise if the condition does not respond as expected or if evaluation reveals findings needing medical attention. Insurers may impose their own requirements, so direct access is a licensing answer rather than a payment answer. Commercially it matters more than most clinics act on: it means the front door of a Maryland practice is its own website rather than a physician’s fax machine, which is the strongest single argument for owning your intake, scheduling and storefront instead of renting a booking widget bolted to somebody else’s EMR.

Is custom software cheaper than WebPT, Prompt or Jane for a physical therapy clinic?

Not as a straight replacement, and we will say so on the call. A four‑therapist clinic on a published ladder is paying somewhere between $3,000 and $12,000 a year for documentation, scheduling and claims, and no honest build pays that back by replacing it — replacing a clinical system also means owning the claims interfaces, the compliance surface and the upgrade path forever. The arithmetic changes beside the subscription. One additional billable unit per therapist hour, which is what the difference between a twenty‑minute and a thirty‑minute visit structure produces under the eight‑minute rule, is worth about $30 under Medicare and $50 under Maryland workers’ compensation; on one therapist working thirty‑five clinical hours a week that is roughly $1,060 a week of Medicare units decided by schedule shape alone. Our fixed prices are a $3,500 Prototype Sprint, an online store from $6,000 and a custom app or operations system from $12,000, so the case is usually made on scheduling arithmetic and the cash‑pay storefront, not on the subscription.

Should a Baltimore physical therapy clinic build its own online store?

It is the strongest single argument in this trade, and it rests on two facts specific to physical therapy. The first is that Maryland’s direct access law lets a patient begin care without a referral, so a storefront is a genuine front door rather than a gift shop. The second is that a clinic knows more about its customers than any retailer selling them the same products: the diagnosis, the exact home exercise program, the bands and blocks and braces already prescribed, and the date the plan of care ends. Everything downstream of the KX threshold is a commercial question rather than a clinical one, and the cash‑pay lines that never touch a payer fit naturally there — dry needling, which Maryland registers separately for a one‑time $100 board fee; running gait analysis; return‑to‑sport testing; maintenance and performance programs; and small‑group classes, which are also the one throughput lever COMAR expressly excludes from the three‑patients‑per‑hour cap. Sold through your own store, on your own customer list, at your own margin.

Method and sources. Vendor pricing was read directly from each company’s own published pages on 24 August 2026; where a URL returned a 404 or redirected away, we have said so rather than quoting an aggregator, and prices change without notice. Jane serves its pricing page in the currency of the visitor’s location, and the plan structure rather than the currency is the point being made. Establishment, employment and payroll figures are from the US Census Bureau, County Business Patterns 2023 county file, NAICS 621340, extracted and totaled by us; CBP counts only establishments with paid employees, so solo practitioners are excluded. Workforce, licensure, compact‑privilege, wage and telehealth figures are from the Maryland Board of Physical Therapy Examiners, 2025 Maryland Physical Therapy Workforce Report, current as of 30 June 2025. Regulatory citations are to COMAR 10.38.03.02 and 10.38.07.02 and to COMAR 14.9.08. Maryland workers’ compensation figures are the Commission’s published Calendar Year 2026 Medical Fee Guide maximum reimbursable amounts and its Calendar Year 2026 Maryland Specific Conversion Factor notice dated 1 November 2025; the Medicare comparison column is our own arithmetic applying the 2026 non‑QP conversion factor to the same relative value units, and assumes the same geographic adjustment on both sides. Medicare conversion factors, the KX modifier threshold and the eight‑minute rule are from the CY 2026 Physician Fee Schedule final rule and Chapter 5 of the Medicare Claims Processing Manual. The composite clinic is ours, built from the shapes of practices we have talked to, and is illustrative rather than a real client. Nothing here is legal, billing, tax or clinical advice — verify your own position with the Maryland Board of Physical Therapy Examiners, the Maryland Workers’ Compensation Commission, your Medicare administrative contractor or your own advisers before relying on any of it.