Every figure on this page links to the organization that produced it, and each was re-checked against that source before publishing. How U.S. physical therapy clinics use video and market themselves, and the evidence behind it: our own read of clinic homepages; how many therapists, assistants and clinics there are and what they earn; how far chains have consolidated the business; who gets physical therapy and how many drop out; what Medicare pays and how often it pays in error; what exercise videos and telehealth do for patients and what physical therapy videos on YouTube and TikTok contain; the direct-access, advertising, testimonial and review rules; and where six popular numbers came from.
31.2% of 298 U.S. physical therapy clinic homepages we read on November 1, 2026 showed a video player; 16.1% played a muted background clip and 9.1% embedded a YouTube player (our measurement). (Mark Studios)See details
80.9% of those homepages linked to Facebook and 68.1% showed testimonials or reviews, but only 15.8% told visitors they could come without a referral. (Mark Studios)See details
Physical therapy clinics showed video as often as chiropractic offices (31.2% each) and more often than veterinary practices in same-day reads with the same method. (Mark Studios)See details
BLS counted 267,330 wage and salary physical therapist jobs in May 2025, 14.6% more than in 2019 (our calculation), and projects 12 percent growth to 2035. (BLS; our calculation)See details
Firms with 25 or more locations ran 14,348 U.S. therapy offices in 2022, 28.4% of the total, and multi-location firms took 54.9% of receipts (our calculations). (U.S. Census Bureau; our calculations)See details
CMS estimated a 19.9% improper payment rate for physical therapists in private practice in its 2025 report, about three times Medicare's overall rate; 88.3% of the errors were missing or insufficient documentation. (CMS; our calculation)See details
Medicare paid for 152,184 months of musculoskeletal remote therapeutic monitoring on office claims in 2024, 7.1 times the 2022 count (our calculation). (CMS; our calculation)See details
The "70% of patients never finish their plan of care" line traces to an unpublished vendor analysis; the best measurement found 73% of 444,995 patients missed at least one visit. (PLOS ONE; WebPT)See details
How many physical therapy clinic websites show video?
Fewer than one in three. On November 1, 2026 we opened the homepages of 400 U.S. physical therapy clinics drawn at random from OpenStreetMap, the open map database, and 31.2% of the 298 we could read showed a video player (our measurement). The frame was every U.S. physical therapy clinic mapped in OpenStreetMap with a website, 956 distinct clinic websites. Of the 400 sampled, 3 asked crawlers to stay away in robots.txt and were not visited, 25 showed a bot wall, 33 failed to load and 41 were no longer physical therapy sites. With the same method on the same day, clinic homepages showed video as often as chiropractic offices' and more often than veterinary practices' (31.2% against 31.2% and 24.1%; dental practices 35.0%, our comparison).
Much of that video is a background loop. 16.1% of homepages played a muted video that starts on its own, 51.6% of the pages with any video (our calculation), and 18.8% carried a video file hosted on the clinic's own site. 12.8% embedded a video from YouTube, Vimeo or a similar host, 9.1% a YouTube player. Of the pages with video, 54.8% used only self-hosted files.
On the homepage (298 readable U.S. physical therapy clinic sites, Nov 1, 2026)
Third-party scheduling tool (Prompt, Zocdoc, Jane and others)
22.1%
Patient testimonials or reviews
68.1%
Careers or jobs link
53.4%
Insurance, Medicare or other payers mentioned
32.9%
Home exercise program, patient portal or app
22.8%
No referral needed (direct access)
15.8%
Telehealth, virtual visits or remote monitoring
15.1%
Free screening, consultation or assessment
12.4%
Names a corporate chain, platform or franchise
6.0%
Clinics lean on social proof and Facebook more than on video. Homepages linked to Facebook 8.9 times as often as they embedded a YouTube player (our calculation), and two in three showed patient testimonials or reviews. Few told visitors they could come without a referral, although every state now allows it in some form. Prompt Health was the most common third-party scheduler, on 18 homepages. Only 18 homepages named a chain or franchise owner; most chain-owned clinics do not say so on the homepage, so that share is a floor and too small to compare. A hand check of 30 random sites agreed with the automatic video classification on all of them (100.0%); misses in the booking and scheduler rules were fixed before counting.
OpenStreetMap is volunteer-mapped, so this is a random sample of U.S. clinics that someone mapped with a website, not of all clinics. One chain website, such as ATI's or Athletico's, stands for many mapped clinics and counts once. The per-site results are in homepages.csv and the summary in crawl-summary.json.
How many physical therapists are there, and what do they earn?
About 267,000 on employer payrolls. The Bureau of Labor Statistics counted 267,330 wage and salary physical therapist jobs in May 2025 (BLS), with a median annual wage of $102,760 (BLS); the top tenth earned $135,140 or more (BLS). Counting the self-employed, the Occupational Outlook Handbook puts the total at about 283,700 in 2025 (BLS). Therapists work beside 112,430 physical therapist assistants, median wage $68,380 (BLS), and 48,150 aides, median $35,240 (BLS). That is 0.42 assistants per therapist (our calculation).
The workforce has grown faster than the pay. From May 2019, when BLS counted 233,350 therapist jobs (BLS) at a median of $89,440 (BLS), jobs rose 14.6% and the median wage 14.9% before inflation (our calculations). Assistant jobs, 96,840 in 2019 (BLS), rose 16.1%, while aide jobs, 49,270 in 2019 (BLS), fell slightly (our calculation).
BLS projects therapist employment to grow 12 percent from 2025 to 2035 (BLS), to 317,600 jobs (BLS), with about 13,400 openings a year (BLS). Assistant jobs are projected to reach 140,400, up 23 percent (BLS), and aide jobs 51,000 (BLS), with about 24,900 openings a year for the two together (BLS).
Where physical therapists work (May 2025, BLS)
Figure
Source
Share of therapist jobs in offices of physical, occupational and speech therapists and audiologists (2025)
Outpatient clinics are the biggest employer and pay therapists the least of the large settings, which is one reason clinics compete hard to recruit. By state, California had the most therapist jobs in May 2025, 26,580 at a median of $124,190 (BLS), then Texas with 21,870 (BLS). Florida had 16,290 therapist jobs (BLS) at a median of $101,040 (BLS), and 9,870 assistant jobs (BLS) at a median of $72,530, above the national figure (BLS).
How many physical therapy clinics are there, and who owns them?
About 54,000 private therapy offices with employees, plus more than 100,000 one-person businesses. Federal statistics do not separate physical therapy from occupational therapy, speech therapy and audiology: all four share one industry code, NAICS 621340, so every count in this section includes them. BLS counted 54,264 private establishments in 2025 (BLS QCEW), employing 518,334 people (BLS QCEW) at average pay of $52,955 per job (BLS QCEW). That is 36.5% more establishments than the 39,741 of 2019 (BLS QCEW; our calculation) and far above the 33,869 of 2015 (BLS QCEW). Florida had 4,614 of them in 2025 (BLS QCEW).
Most offices are small. Census County Business Patterns counted 52,058 offices with paid employees in 2023 (Census Bureau), employing 466,903 people (Census Bureau); 29,545 of them had fewer than five employees (Census Bureau). Another 116,639 therapy and audiology businesses had no employees at all in 2024 (Census Nonemployer Statistics), up 28.8% from 90,570 in 2019 (Census Bureau; our calculation); 17,520 of those were in Florida (Census Bureau).
Revenue has grown. Offices with employees took in $42.80 billion in 2022 (2022 Economic Census), against $31.51 billion in 2017 (2017 Economic Census), a rise of 35.8% before inflation (our calculation). The Census Bureau's annual business survey puts 2024 revenue at $46.79 billion (Census AIES), after a pandemic dip from $36.09 billion in 2019 (Census Bureau) to $34.02 billion in 2020 (Census Bureau). Half the 2022 total, $21.26 billion, came from patients whose main diagnosis was musculoskeletal (Census Bureau). The wider group of offices of other health practitioners (chiropractors, optometrists, podiatrists and therapists) spent $1.55 billion on advertising in 2024 (Census AIES); Census publishes no advertising figure for therapy offices alone.
Is the industry consolidating?
Slowly, and further than the top four suggest. In the 2022 Economic Census the four largest firms took 12.9% of receipts (Census Bureau), up from 10.2% in 2017 (Census Bureau), and the 50 largest took 33.1% (Census Bureau). But 1,948 multi-location firms ran 20,758 of the 50,492 offices (Census Bureau) and took 54.9% of receipts, more than the 29,734 single-location firms (Census Bureau; our calculation). Just 14,348 offices, 28.4% of the total, belonged to the 77 firms with 25 or more locations (Census Bureau; our calculation).
Largest U.S. outpatient therapy operators, as they report it
Clinics
Source
Select Medical, end of 2025 (taken private July 2026)
The three that file with the SEC ran 3,579 clinics between them, equal to about 6.6% of the establishments BLS counts (our calculation; company counts include managed clinics). Two have since left the public markets: ATI was delisted on December 4, 2024 (ATI 10-K) and filed to stop reporting on March 17, 2025 (ATI Form 15), and Select Medical was taken private at $16.50 a share effective July 1, 2026 (Select Medical 8-K). U.S. Physical Therapy calls the business highly fragmented, with no company holding a significant national share (USPH 10-K).
The chains are still buying and opening clinics. U.S. Physical Therapy grew from 591 clinics at the end of 2021 (USPH 10-K), adding 47 owned clinics in 2025 (USPH 10-K) and closing 23 (USPH 10-K), and bought fourteen single clinics besides (USPH 10-K). Select Medical started up 27 clinics in 2025 and closed or sold 32 (Select Medical 10-K). ATI closed or sold 35 in 2024, citing losses and liquidity (ATI 10-K). FYZICAL's franchise page, read November 1, 2026, says it averages 324 days from signing to opening (FYZICAL).
Mostly to referring doctors, and with small advertising budgets. U.S. Physical Therapy says it focuses its marketing primarily on referring physicians, from orthopedic surgeons to primary care doctors (USPH 10-K). ATI lists advertising to raise patient awareness alongside sales calls on referral sources (ATI 10-K). ATI recorded $2.6 million of advertising in clinic costs (ATI 10-K) and $5.2 million in corporate costs in 2024 (ATI 10-K), against $6.9 million of corporate advertising in 2019 (ATI 10-K). Together that is 1.1% of its $690.0 million net patient revenue (ATI 10-K; our calculation). ATI reports an average Google rating of 4.9 stars (ATI 10-K) and a Net Promoter Score of 75 (ATI 10-K), both its own measures; Athletico's homepage said on November 1, 2026 that 98% of patients would recommend it, without a source (Athletico).
Who uses physical therapy, and how many finish their care?
Far fewer people use physical therapy than have the problems it treats. The CDC's National Health Interview Survey found 24.3% of U.S. adults had chronic pain in 2023, 8.5% of them pain that often limited life or work (CDC NCHS), up from 20.4% in 2019 (CDC NCHS); women reported it more often, 25.4% against 23.2% of men (CDC NCHS). In 2019, 39.0% of adults had back pain in the past three months (CDC NCHS) and 58.9% had pain of some kind (CDC NCHS). Among adults with high-impact chronic pain that year, physical, rehabilitative or occupational therapy was the most-used non-drug treatment, at 25.9% (Pain Research and Management, from NHIS 2019).
The last national counts of visits are old. From federal MEPS survey data, U.S. adults with musculoskeletal conditions made about 60.00 million ambulatory visits a year to physical therapists in 2009-2012 (Physiotherapy Theory and Practice), costing $9.37 billion a year (Physiotherapy Theory and Practice). An average episode of care in 2008-2012 was 9.9 visits with mean spending of $1,708 (Physical Therapy journal), and 54% of episodes carried an out-of-pocket cost (Physical Therapy journal). Hinge Health, which sells a digital alternative to employers, says its claims analysis shows only about 9% of U.S. adults pursued in-person physical therapy in 2023 (Hinge Health 10-K); that is a company estimate.
How many people with back pain get physical therapy?
A minority, and it depends on the insurer. In national surveys of primary care visits for low back pain, about 10.1% led to a physical therapy referral, a share that did not change from 1997 to 2010 (Spine). Medicaid patients' visits were about half as likely to end in a referral, an odds ratio of 0.48 (Spine). Studies of who actually went:
Low back pain patients who received physical therapy
Share
Source
Commercially insured adults with a new primary care visit, within 90 days (n=32,070)
Early physical therapy is linked to lower costs in claims studies, and to smaller gains in trials. In a 2012 claims study, patients who started within 14 days had back-pain costs $2,736 lower over 18 months (Spine) and lower odds of advanced imaging, an odds ratio of 0.34 (Spine); in the military system in 2007-2009, early guideline-based therapy went with 60% lower two-year costs (BMC Health Services Research). These are observational. In a Utah randomized trial of 220 adults in 2011-2013, four early sessions improved disability by 3.2 points on the Oswestry scale, short of the 6 points usually treated as meaningful (JAMA). A military trial of 119 service members, published in 2018, found no disability difference, with costs of $2,016 in the early-therapy group (Spine).
How many patients drop out?
Many, but the best measurement says "miss a visit", not "never finish". In records of 444,995 ATI Physical Therapy patients at 828 clinics in 2016-2017, 73% missed at least one appointment (PLOS ONE; three authors worked for ATI), and 85% of Medicaid patients did (PLOS ONE). Before that 2021 study, the authors note, published no-show rates came from two small studies outside the U.S., at 11-31% (PLOS ONE). Smaller U.S. studies:
Of 100 Stony Brook orthopedic clinic patients prescribed therapy in late 2022, 40 never attended (Cureus).
Of 104 shoulder surgery patients at a Chicago academic center (published 2025), 20 stopped postoperative therapy early (Journal of Shoulder and Elbow Surgery).
Among 26 Wisconsin patients who came only to their first evaluation (published 2025), 26.9% gave reasons tied to the therapy itself (Musculoskeletal Science and Practice).
In a 2024 survey, Alabama physical therapists (n=251) recommended an average of 23.8 visits for full recovery across 11 common orthopedic diagnoses (Southern Medical Journal).
At one Ohio practice in 2018-2019, patients who chose to use a gamified clinic app attended 13.21 visits on average against 10.72 for those who did not (JMIR Rehabilitation and Assistive Technologies); patients picked themselves, so the gap is not proof the app caused it.
Home exercise is where adherence slips. In a probability survey of 1,006 U.S. adults for Ohio State Wexner Medical Center in September 2025, 76% of those given home exercises did not complete all of them (Ohio State), though 24% completed all and 28% completed 75-99% (Ohio State).
How much physical therapy happens by video?
Little, after the pandemic peak. In commercial claims for 2,007,524 outpatient physical and occupational therapy patients in 2020-2021, only 2.5% had a telerehabilitation visit (Archives of PM&R); telehealth's share of visits peaked at 10.9% in April 2020 (Archives of PM&R) and ranged by state from 10.4% of patients in California to 0.3% in Wyoming (Archives of PM&R). The 2022 Economic Census found 21.3% of therapy offices with employees had some telemedicine revenue (10,755 of 50,492, (Census Bureau); our calculation).
Trials say video care works for knee and other joint problems. In Australia's PEAK trial (2019-2022) of 394 adults with chronic knee pain, physiotherapy by video call was non-inferior to in-person care (The Lancet). A 2021 trial of 415 Australians with knee osteoarthritis found six video consultations plus exercise improved pain and function (Annals of Internal Medicine). Meta-analyses agree. A 2017 review of 13 studies of 1,520 people found real-time telerehabilitation equivalent to face-to-face care (Clinical Rehabilitation); a 2023 review found internet-based telerehabilitation cheaper by $89.55 per person (Annals of Physical and Rehabilitation Medicine); and a 2025 network meta-analysis found self-managed apps reduced pain with an effect size of -0.89 (JOSPT).
What does Medicare pay for physical therapy?
Less per unit than five years ago. CMS's 2024 Medicare data list 77,901 physical therapists in private practice billing Original Medicare Part B (CMS), up from 74,061 in 2023 (CMS) and 58,806 in 2019 (CMS), a rise of 32.5% (our calculation). The most-billed service, therapeutic exercise (CPT 97110), went to 3,038,010 beneficiaries in 2024 (CMS) at an average payment of about $17.54 per 15-minute unit (CMS), against about $20.09 in 2019 (CMS): 12.7% less before inflation (our calculation). These code-level counts cover office claims from every provider type that bills the code, not only physical therapists.
The fee schedule's conversion factor, the dollar value that turns relative value units into payment, was $32.3465 in 2025 (Federal Register), after $32.7442 in early 2024 (Federal Register). For 2026 Medicare uses two factors for the first time: $33.4009 for most private-practice therapists (Federal Register) and $33.5675 for qualifying alternative payment model participants (Federal Register). The 2027 proposed rule would lower the first to $32.8409 (Federal Register, proposed). U.S. Physical Therapy estimated fee-schedule cuts of about 3.5% in 2021, 0.75% in 2022, 2.0% in 2023 and 1.8% in 2024 (USPH 10-K), and a rise of about 1.75% for 2026 (USPH 10-K).
Medicare outpatient therapy rules (PT and speech therapy combined)
From a small base. Medicare paid for 21,465 30-day periods of musculoskeletal monitoring device supply (CPT 98977) on office claims in 2022, the first year therapists could bill it (CMS), 92,410 in 2023 (CMS) and 152,184 in 2024 (CMS): 7.1 times as many in two years (our calculation), still a small number next to the three million beneficiaries who received therapeutic exercise. 53,465 beneficiaries had monitoring set-up and education in 2024 (CMS), and a month of treatment management (98980) paid about $38.73 on average (CMS). For 2026 the CPT panel added codes for fewer than 16 days of monitoring (Federal Register). U.S. Physical Therapy began monitoring in 2024 but its CEO called the first version "clunky" (USPH earnings call, SEC 8-K), and it plans to restart in 2026 after the rule change (USPH, SEC 8-K). Mentions of "remote therapeutic monitoring" in SEC filings fell from 21 documents in 2022 (SEC EDGAR full-text search) to 8 in 2025 (SEC EDGAR full-text search).
How often are physical therapy claims paid in error?
About one claim dollar in five. CMS's 2025 improper payment review estimated a 19.9% improper payment rate for physical therapists in private practice (CMS CERT), about $731.8 million (CMS CERT), roughly three times Medicare's overall rate of 6.55 percent (CMS CERT; our calculation). The rate was 15.8% in the 2024 report (CMS CERT) and 18.6% in 2023 (CMS CERT). Missing paperwork, not fraud, drove 88.3% of the 2025 errors (CMS CERT). By code, the overpayment rate was 25.2% for therapeutic exercise (CMS CERT), 30.5% for manual therapy (CMS CERT) and 18.1% for therapeutic activities (CMS CERT). The last nationwide HHS Inspector General audit, of claims from July to December 2013, found 61 percent of sampled outpatient therapy claims did not comply (HHS OIG), an estimated $367 million (HHS OIG).
Do exercise videos help patients do their home program?
Somewhat, and mostly in the short term. A 2022 JOSPT meta-analysis of 11 trials (n=1,144) found digital rehabilitation improved exercise adherence with a standardized effect of 0.53, at intermediate follow-up only (JOSPT). A 2019 meta-analysis rated the evidence that multimedia instructions beat written ones as very low quality, an effect of 0.60 whose confidence interval crossed zero (Clinical Rehabilitation). A 2022 review found Seven of the ten trials favored adding a digital tool (Archives of Physiotherapy). The trials point in both directions:
Trial or study
Finding
Source
Shoulder exercise instruction, 40 people
Handout-only learners made more than twice as many errors as video or live learners
Patients tend to prefer video. In a 2019 U.S. study, when upper-limb rehabilitation patients were offered both a paper handout and a video of themselves doing the exercises, 69% preferred the video (Journal of Hand Therapy).
Digital exercise companies now compete with clinics for the same patients. Hinge Health, which sells app-based exercise therapy with video to employers and health plans, had 2,830 clients (Hinge Health 10-K) and about 25 million eligible people at the end of 2025 (Hinge Health 10-K), and over 532,000 members at the end of 2024 (Hinge Health prospectus). It estimates its platform cut the care-team hours of traditional physical therapy by about 97% (Hinge Health 10-K), and 64% of 2024 members were onboarded the day they filled in their intake form (Hinge Health prospectus). These are the company's own figures.
What do physical therapy videos on YouTube and TikTok show?
Physical therapists make many of them, and most score low on quality scales. Researchers have rated the most-viewed videos on common conditions with standard tools such as DISCERN and the Global Quality Score (GQS, 1 to 5):
In a 2024 study, physiotherapists and osteopaths posted 86.0% of the first 100 TikTok videos on frozen shoulder, and their videos scored lower than doctors' (Journal of Orthopaedic Surgery and Research).
On TikTok elbow tendinopathy videos, self-identified health professionals scored a median DISCERN of 25 out of 75, no better than other users (Journal of Hand Therapy).
Of 91 YouTube balance exercise videos searched in August 2025, 24% contained inaccurate information (The Physician and Sportsmedicine); of 136 core stability videos, 11% were misleading (BMC Public Health).
In a 2026 study, German-language knee osteoarthritis videos scored 15.4% on an information-quality checklist, with no better score for YouTube's health certification label (ZEFQ).
Across 157 of the most-viewed #backpain TikTok videos in September 2021, chiropractors' videos drew the most views and comments (Cureus).
In a 2023 study, only 14% of 632 Instagram and Twitter posts about physical therapy treatments cited a reference, and 57% showed a possible conflict of interest (Brazilian Journal of Physical Therapy).
Most of these studies sample videos from search results on one day, in one country, so they describe what a searcher saw, not everything on the platform.
Can a physical therapist get YouTube's health label?
Not as an individual. YouTube's health source program, which adds an information panel and ranks qualifying channels in health searches, lists five eligible licensed professions, and physical therapists are not one of them (YouTube Help); clinics can apply only as organizations, and channels need more than 1,500 public watch hours in the past year (YouTube Help).
The audience is there. In the National Cancer Institute's 2022 HINTS survey, 59% of Americans had watched a health-related video on YouTube in the past year (NCI HINTS). Pew found 36% of U.S. adults got health information from social media at least sometimes in October 2025 (Pew Research Center) and Four-in-ten ever got health and wellness information there (Pew Research Center). Of 6,828 prominent U.S. health and wellness influencers, 62% had a TikTok account (Pew Research Center) and 41% called themselves some kind of health care professional (Pew Research Center).
Can patients see a physical therapist without a referral?
In every state, with limits in most. The American Physical Therapy Association, which represents the profession, says that as of July 1, 2025, all 50 states, the District of Columbia and the U.S. Virgin Islands allow some form of direct access (APTA). Its 2025 report counts 21 states with unrestricted access (APTA) and 29 states plus DC and the Virgin Islands with provisional access, which adds time, visit or referral conditions (APTA). Alabama, in 2024, was the last of three holdout states to move past limited access (APTA). Nebraska had direct access by omission in 1957 (APTA), 21 states adopted it in the 1980s (APTA), and Michigan became the 50th state in 2015 (APTA). Aetna dropped its referral requirement for physical therapy in June 2025, affecting more than 26 million members (APTA).
Florida allows direct access with a clock: if therapy for a condition not assessed by a practitioner of record is needed beyond 30 days, the therapist needs a plan of care approved by a physician or other listed practitioner (Florida Statutes 486.021). Medicare needs no referral, but a physician or nurse practitioner must certify the plan of care, recertified at least every 90 days, and a plan sent within 30 days when an order is on file needs no signature (eCFR, 42 CFR 424.24).
Can a therapist treat patients in other states by video?
Through the Physical Therapy Licensure Compact, in most states. The compact commission lists 38 member jurisdictions issuing and accepting privileges (PT Compact Commission), with 3 more enacted but not yet live and 10 non-members (PT Compact Commission). More than 11,000 therapists and assistants use it (PT Compact Commission). A privilege costs a $45 commission fee plus a state fee (PT Compact Commission), requires two years free of discipline (PT Compact Commission) and is usually issued within minutes (PT Compact Commission). Florida's statutes enact the compact (s. 486.112, ch. 2024-15) (Florida Statutes), yet the commission's map still lists Florida as a non-member; the two sources disagree, and we could not find the commission's reason.
What rules apply to physical therapy ads, testimonials and reviews?
State health-practitioner law, HIPAA and the FTC all apply, and the penalties are real money. A clinic that posts a patient testimonial video needs the patient's written HIPAA authorization: 45 CFR 164.508(a)(3) says a covered entity must obtain an authorization before using health information for marketing (eCFR). In 2016 Complete P.T., Pool & Land Physical Therapy, a Los Angeles-area practice, paid HHS $25,000 (HHS OCR, archived) over patient testimonials with full names and photos posted without authorization; the complaint dated from August 8, 2012 (HHS OCR, archived) and the practice had 10 days to take the information down (HHS resolution agreement, archived). Its resolution agreement set a three (3) years compliance term, while OCR's case summary says one year; we treat the signed agreement as the record (HHS resolution agreement, archived). A New Jersey practice paid $30,000 in 2023 for answering online reviews with patients' health details (HHS OCR, archived).
Rule
Figure
Source
Florida: maximum administrative fine per count for health practitioners
Florida law also requires any ad that names a practitioner to state the type of license held (Florida Statutes 456.072(1)(t)) and bans misleading, deceptive, or fraudulent representations (Florida Statutes 456.072(1)(a)). Florida's physical therapy rules contain no separate testimonial rule.
The FTC's revised Endorsement Guides took effect on July 26, 2023 (Federal Register): if a clinic cannot show a patient's results are typical, the ad should say what results people can generally expect (eCFR, 16 CFR 255.2), and free or discounted treatment given to a testimonialist is a material connection to disclose (eCFR, 16 CFR 255.5). The FTC's rule on consumer reviews and testimonials took effect on October 21, 2024 (Federal Register). It bans a fake consumer testimonial (eCFR, 16 CFR 465.2), rewards conditioned on positive reviews under § 465.4 (eCFR) and review suppression under § 465.7 (eCFR). FTC staff sent the first warning letters under it to 10 companies on December 22, 2025 (FTC, archived). We found no FTC action against a physical therapy practice.
What the popular physical therapy numbers really say
Physical therapy marketing agencies and software vendors repeat a handful of numbers about patients. We traced each to the organization that produced it. Pages that repeat a claim are linked only as dated evidence of how it spread, never as a source for the number.
"70% of patients don't complete their plan of care"
No published study is behind it. WebPT, which sells practice software to therapy clinics, quotes its co-founder saying about 70% of patients fail to complete their course of care (WebPT); a 2018 WebPT webinar page says 70% too (WebPT), and the figure travels on through an at-home therapy company, 70% (Luna), a home exercise software firm that attributes it to WebPT's co-founder, 70% (Kemtai), and a 2023 trade article by a gait-analysis vendor's executive saying only 30% complete their care (Healthcare IT Today). The oldest form we found is the homepage of Strive Labs, a patient-retention software company WebPT later bought, captured in August 2018 with "70%" of patients who fail to complete care and no source (Strive Labs, archived). WebPT's 2018 "facts" post states a narrower claim, that only 30% of patients attend all the visits their insurer authorizes (WebPT), and says its unlinked facts came from Strive Labs' analysis of about 30,000 insurance claims from one unnamed region in 2012 (WebPT). That analysis was never published. Attending every authorized visit is also not the same as finishing a plan of care.
Use instead: the ATI records study above: 73% of 444,995 patients missed at least one visit in 2016-2017.
"90% of people with back pain never see a physical therapist"
It is vendor arithmetic. WebPT divided 11.7 million adult outpatient therapy users in 2011 by "a hundred million" people with musculoskeletal problems to get 9.58% (WebPT); Strive Labs' 2018 homepage said 8% (Strive Labs, archived); a gait-analysis vendor's executive wrote in 2023 that as many as 91% who could benefit do not get therapy (Healthcare IT Today). A Colorado clinic blog cites a Spine study as showing only 10% of back-pain patients were referred (MEND Physical Therapy). That study counted visits, not people: about 10.1% of U.S. primary care visits for low back pain in 1997-2010 led to a referral (Spine), 17.1 million referrals from 170 million visits (Spine). Across all outpatient visits for back or neck pain, the referral share was about 20% (JAMA Internal Medicine), and in one Utah health plan 6.4% of new back-pain episodes started with a physical therapist (Journal of Evaluation in Clinical Practice).
Use instead: the share of back-pain patients who received therapy in each study in the table above, from 7.0% to 20.0%, with its population.
"Video home exercise programs raise adherence by X%"
The number usually quoted comes from one small stroke trial. Physitrack, which sells home-exercise software, says one trial reported 76% adherence at three months for video against 55% for paper (Physitrack), and markets an NHS falls-prevention cohort at 84% adherence (Physitrack). The trial is a Hong Kong pilot of 56 stroke patients, 75.6% against 55.2% (Hong Kong Physiotherapy Journal), not outpatient orthopedic therapy. An Australian trial of 80 people found an app with calls and texts raised adherence by 1.3/11 points (Journal of Physiotherapy), and the meta-analyses rate the evidence low.
Use instead: the 2022 JOSPT meta-analysis effect of 0.53 at intermediate follow-up, and say the evidence for video alone is weak.
"Seeing a physical therapist first cuts costs by 72%"
It compares therapy-first with imaging-first, in one Utah health system. A workers' compensation network manager wrote in 2015 that starting with therapy instead of an MRI reduced costs 72% (MedRisk). The origin is APTA's March 2015 release on a Health Services Research study, which said costs were nearly 72% lower, $1,871 against $6,664 in charges (APTA, via archived release); the study's abstract reports a $4,793 difference for patients sent to imaging first (Health Services Research). Other studies measure direct access, not imaging: in Northwest commercial claims, therapy-first patients had an 89.4 percent lower probability of an opioid prescription but a higher probability of hospitalization (Health Services Research); a South Carolina registry found $1543 lower paid claims with direct access (JOSPT), a figure a clinic blog repeats as $1543 saved (MEND Physical Therapy). In national OptumLabs claims, therapy-first patients had 13% lower costs at 30 days in unrestricted-access states (Physical Therapy journal) but 25% higher in provisional-access states (Physical Therapy journal).
Use instead: the OptumLabs results, both of them, with the state access rule beside each.
"20% of patients drop out within the first three visits"
Same source, same gap. Strive Labs' 2018 homepage said 20% drop out after the third visit, with no source (Strive Labs, archived), and WebPT repeats it. We found it in no study.
Use instead: the ATI no-show figures, 73% of all patients and 85% of Medicaid patients missing at least one visit.
"Up to 70% of patients don't do their home exercises" and "only 35% fully adhere"
These point to a 1993 Dutch study whose abstract gives no such rate. Medical Economics wrote in 2026 that home-exercise nonadherence "can reach as high as 70%" (Medical Economics), and a U.S. neck-pain study repeats "as high as 70%" as background (Musculoskeletal Science and Practice). WebPT's 2018 post says only 35% fully adhere and links to Sluijs et al. (WebPT). That study drew on 222 Dutch private-practice therapists (Physical Therapy journal) and 1,681 patient questionnaires (Physical Therapy journal), and its abstract reports the factors linked to non-compliance, not a rate. We could not read the full text, so whether a percentage appears inside it is unconfirmed.
Use instead: the September 2025 Ohio State survey of U.S. adults: 76% did not complete every home session, but most completed at least half.
We started from the questions people search for about physical therapy video statistics and looked for the organization that actually produced each answer. Every figure on this page sits in an evidence ledger with its source, the exact table or paragraph it came from, the measurement period, the population measured, and the date we checked it. This edition holds 366 figures from 71 publishers.
On November 1, 2026, each sourced figure on this page was re-fetched with the fetch cache cleared and matched against the recorded wording: BLS Occupational Employment and Wage Statistics (through the BLS data API), Quarterly Census of Employment and Wages open data and the Occupational Outlook Handbook (through Internet Archive copies, because bls.gov refuses automated requests); Census County Business Patterns, Nonemployer Statistics, the 2017 and 2022 Economic Census and the Annual Integrated Economic Survey (through the Census data API); CMS's Medicare Physician & Other Practitioners data (through the data.cms.gov API), the CY 2025, 2026 and 2027 Physician Fee Schedule rules in the Federal Register and CMS's CERT improper payment reports; an HHS Inspector General audit; the 10-K and 8-K filings of U.S. Physical Therapy, ATI Physical Therapy, Select Medical, Concentra and Hinge Health on SEC EDGAR; APTA and the Physical Therapy Compact Commission; Florida statutes and Board of Physical Therapy rules; HIPAA, HHS civil money penalty and FTC rules in the eCFR and Federal Register and HHS settlements (through Internet Archive copies); CDC NCHS data briefs, Pew Research Center and the National Cancer Institute's HINTS; peer-reviewed studies through PubMed; and the pages behind each popular claim. The homepage figures are our own measurement on the same day, described in the first section. We left out a payer-mix table Census publishes only for the wider practitioner group, BLS monthly employment for the therapy industry (the same ground is covered by QCEW), clinic counts on chain websites that print them only in page code, and a 1993 study's full text we could not open.
We only cite the producer: government agencies, regulators, company filings and official posts, and the survey owner's own report. We do not cite statistics roundups, Statista-style aggregators, or salary sites without a public method. We did not bypass paywalls or sign-up forms; where a full report was gated, we used only the figures its publisher states publicly.
Figures marked "our calculation" are arithmetic on the cited numbers, and anyone can reproduce them:
1527 Overpass elements -> 1524 with a usable URL -> minus 4 on directory/social hosts {'facebook.com': 3, 'yellowpages.com': 1} -> one row per registrable host = 956
93 / 298 readable homepages with video_player = true in homepages.csv; rule: video player: a <video> element or player iframe/placeholder (YouTube, Vimeo, Wistia, Vidyard, Brightcove, Loom, Facebook video ...) at least 150x80 px, or a labelled click-to-play control or hidden player embed
56 / 298 readable homepages with video_self_hosted = true in homepages.csv; rule: video player that is a native <video> element (file served by the site or its site builder/CDN, incl. background videos)
38 / 298 readable homepages with video_embedded = true in homepages.csv; rule: video player from a third-party host (YouTube, Vimeo, Wistia, Facebook, TikTok ...)
27 / 298 readable homepages with youtube_player = true in homepages.csv; rule: a YouTube player (iframe, lite-youtube placeholder, lightbox embed link or hidden embed)
97 / 298 readable homepages with youtube_channel_link = true in homepages.csv; rule: link to a YouTube channel or handle (not a single video; template placeholders excluded)
207 / 298 readable homepages with online_booking_or_request = true in homepages.csv; rule: online booking or appointment request: a link/button labelled book/schedule (online, now, today, [a/your] [word] appointment/visit/evaluation/assessment/screening/session), request an appointment/evaluation, appointment request, make an appointment or online scheduling/booking (not "call to schedule", not tel:/mailto: links even when labelled "Appointment Request"), an appointment URL path, or a third-party scheduling tool
66 / 298 readable homepages with online_booking_tool = true in homepages.csv; rule: third-party scheduling tool: a link, button or iframe to a scheduling vendor (WebPT, Prompt, Raintree, SPRY, Net Health, Clinicient, Strata, PtEverywhere, Jane, Calendly, Acuity, Setmore, Square, Zocdoc, NexHealth, Luna, Tebra, Healthgrades booking, MyChart/Epic open scheduling, Kyruus, DocASAP ...; EMR hosts only on a scheduling/booking/patient URL), a widget request to a scheduling subdomain, or a booking-labelled link to any other third-party host
37 / 298 readable homepages with free_screening_offer = true in homepages.csv; rule: free or complimentary screening, consultation, assessment or evaluation offered in the page text
45 / 298 readable homepages with telehealth = true in homepages.csv; rule: telehealth, telemedicine, telerehab, virtual visit/care/therapy, online PT sessions, video visits or remote (therapeutic) monitoring/RTM mentioned in the page text or a link labelled so (negated mentions excluded)
68 / 298 readable homepages with hep_or_portal = true in homepages.csv; rule: home exercise program, patient portal or app: page text naming a home exercise program/HEP, home exercises, a patient/client portal, "our app"/"download our app", or MedBridge, Physitrack, PtEverywhere, PTWired, HEP2go; or a link to those vendors, an app store, MyChart or a patient portal/login
47 / 298 readable homepages with direct_access = true in homepages.csv; rule: direct access mentioned: "no referral/prescription needed/required", "without a (physician's) referral/prescription", "direct access" or "you don't need a referral"
98 / 298 readable homepages with insurance_mention = true in homepages.csv; rule: insurance named in the page text: Medicare, Medicaid, TRICARE, workers' comp, in-/out-of-network, cash-based, a named insurer, or "insurance accepted/plans/providers/verification" or "we accept/take/bill (most) insurance"
203 / 298 readable homepages with testimonials_reviews = true in homepages.csv; rule: testimonials or reviews shown: page text with testimonials, patient reviews/stories/success stories, "what our patients say", Google reviews, star ratings or review counts; a link labelled testimonials/reviews or to a /testimonials or /reviews path; or a review-widget request (Birdeye, Podium, Elfsight, Trustindex ...)
159 / 298 readable homepages with careers_link = true in homepages.csv; rule: link labelled careers, jobs, join our team, now hiring, employment, open positions or work with us, or a link to a /careers, /jobs, /employment or /join-our-team path
18 / 298 readable homepages with corporate_chain = true in homepages.csv; rule: corporate chain shown: the host is a multi-state chain, PE-backed platform or franchise (ATI, Athletico, Select/NovaCare/Kessler, Upstream, Ivy Rehab, BenchMark, PT Solutions, Results, Bay State, Excel, Professional PT, FYZICAL, Pivot, CORA, Therapeutic Associates, Phoenix ...), or the page names one of them with an ownership cue (copyright, part/member/affiliate of, owned/operated by, franchise) and no referral/biography cue, or a careers link to a chain job board, or "member/part of the X family/network" naming a PT/health group
This edition contains no Mark Studios customer data.
Limitations
Federal statistics do not separate physical therapy. Census and BLS count physical, occupational and speech therapists and audiologists as one industry (NAICS 621340), so clinic counts, receipts and concentration include all four.
Our homepage sample comes from a volunteer map. OpenStreetMap covers clinics someone mapped with a website, so the shares describe that sample, not every U.S. clinic; one chain website counts once however many clinics it serves.
Nobody measures what marketing video does for a clinic. The video studies here measure exercise adherence and content quality, not new patients or bookings.
Medicare code counts cover every provider type. The CPT-level figures include occupational therapists, chiropractors and physicians billing the same codes, and only office claims in Original Medicare.
Several figures come from companies with an interest. WebPT, Strive Labs, Physitrack, Prompt and Hinge Health sell to clinics, employers or patients, and the chains' Google ratings and Net Promoter Scores are their own measures; each is named beside its figure.
Many studies are small or from outside the U.S. Trials from Hong Kong, Australia, Norway and Korea and video studies searched from Brazil, India and Germany are labelled with their country; the national visit counts from MEPS date from 2009-2012.
How to cite this page
For a single figure, cite the original publisher named next to it, and link here if the verification trail is useful to your readers. Each figure has its own anchor, so you can link straight to it, for example https://www.markstudios.com/research/physical-therapy-video-statistics#MCR-049. To cite the page as a whole:
Mark Studios. "Physical Therapy Video Statistics." Last verified Nov 1, 2026. https://www.markstudios.com/research/physical-therapy-video-statistics
APA
Mark Studios. (2026, November 1). Physical therapy video statistics. https://www.markstudios.com/research/physical-therapy-video-statistics
MLA
“Physical Therapy Video Statistics.” Mark Studios, 1 Nov. 2026, www.markstudios.com/research/physical-therapy-video-statistics.
Chicago
Mark Studios. “Physical Therapy Video Statistics.” Last modified November 1, 2026. https://www.markstudios.com/research/physical-therapy-video-statistics.
BibTeX
@misc{markstudios_physical_therapy_video_statistics_2026,
author = {{Mark Studios}},
title = {{Physical Therapy Video Statistics}},
year = {2026},
month = nov,
url = {https://www.markstudios.com/research/physical-therapy-video-statistics},
note = {Last verified 2026-11-01}
}
Our calculations and the evidence table are free to reuse with attribution and a link to this page. Each underlying figure remains the claim of its original publisher, and some publishers ask for a link when their figures are reused.
We re-verify every figure quarterly and when a major source publishes a new edition. The next scheduled review is February 2027. Corrections go to contact@markstudios.com.
Nov 1, 2026: First published with 366 figures from 71 publishers, each checked against its original source.
Frequently asked questions
How many physical therapy websites have video?
In our November 1, 2026 read of 298 U.S. physical therapy clinic homepages drawn at random from OpenStreetMap, 31.2% showed a video player. 16.1% of homepages played a muted background loop and 9.1% embedded a YouTube player.
How many physical therapists are there in the U.S.?
BLS counted 267,330 wage and salary physical therapist jobs in May 2025, at a median annual wage of $102,760, and about 283,700 jobs counting the self-employed. It projects 12 percent growth from 2025 to 2035.
How many physical therapy clinics are there?
BLS counted 54,264 private offices of physical, occupational and speech therapists and audiologists in 2025; federal data do not separate physical therapy. Another 116,639 such businesses had no employees in 2024.
What percentage of physical therapy patients don't finish their care?
No published study supports the popular 70% figure; it traces to a software vendor's unpublished analysis of about 30,000 insurance claims. In records of 444,995 ATI Physical Therapy patients, 73% missed at least one appointment.
Do exercise videos improve home exercise adherence?
Modestly. A 2022 JOSPT meta-analysis of 11 trials found digital rehabilitation improved adherence with an effect size of 0.53 at intermediate follow-up. The often-quoted 76% versus 55% comes from a 56-patient stroke trial in Hong Kong, which measured 75.6% against 55.2%.
Can patients see a physical therapist without a referral?
Yes, in some form in all 50 states. APTA counts 21 states with unrestricted direct access; the rest add time or visit limits. Florida requires a practitioner-approved plan of care after 30 days.
Can a physical therapy clinic post patient testimonial videos?
Yes, with the patient's written HIPAA authorization and truthful claims. A California practice paid HHS $25,000 in 2016 for posting testimonials without authorization, and the FTC can seek up to $53,088 per violation of its reviews rule.
What does Medicare pay for physical therapy?
Payment follows the Physician Fee Schedule. The 2026 conversion factor is $33.4009 for most private-practice therapists, and claims above $2,480 per patient in 2026 need the KX modifier. Medicare paid about $17.54 per unit of therapeutic exercise in 2024.
