Getting the Most From Your Ortho PAs

Disclaimer: This guide is educational and is not legal or billing advice. Verify anything you act on with your healthcare counsel and certified coders.

Most ortho practices that employ PAs could get more from them without anyone working harder. The gains usually come from billing each service the right way, putting the PA in the OR on cases Medicare actually pays for, building clinic templates around what patients want from each clinician, and keeping good PAs long enough to recover the cost of training them. This guide covers each of those, and ends with a one-page self-audit. Every number and rule is footnoted to a primary source listed at the end.

---

1. Medicare billing for PA services

The basics

• 85% of the fee schedule. Medicare pays PA services at 80% of the lesser of the actual charge or 85% of what a physician would be paid under the Physician Fee Schedule.¹ In practice, the allowed amount is 85% of the physician rate, and the usual coinsurance split applies.
• Bill at full charge. PA claims go out at the full physician charge. The PA's NPI tells the Medicare contractor to pay at 85%.²
• Direct payment since 2022. Since January 1, 2022, Medicare may pay PAs directly. Assignment for PA services is mandatory.³
• Medicare defers to state law. PA services are covered when performed under the general supervision of an MD or DO and in accordance with state law.³ Medicare's own examples of covered PA services include physical exams, minor surgery, setting casts for simple fractures and interpreting x-rays.³
• Co-signing doesn't change the billing provider. A physician reviewing or co-signing a PA's chart, or discussing the patient with the PA, does not allow the PA's service to be billed under the physician's name.²

Incident-to: 100%, with strict conditions

Incident-to billing lets a PA's service be billed under the physician's NPI at the full physician rate. The requirements:

• The physician furnished a direct, personal, professional service to initiate the course of treatment, and the PA's service is an incidental part of it.⁴
• The physician personally furnished a professional service, established the diagnosis and initiated treatment.²
• The service is in the office setting, Place of Service 11 or 50.²
• The service is furnished under the physician's direct supervision.⁵
• The PA is a direct financial expense to the billing physician (W-2, leased employee or independent contractor) or has the same employer tax ID.²

The discrepancy between the manual and the regulation. The Medicare Benefit Policy Manual's incident-to section still describes the physician as "physically present in the same office suite."⁴ The current regulation says direct supervision may include virtual presence through real-time audio and video (not audio-only), except for services with a 010 or 090 global surgery indicator.⁶ The manual section dates from 2003; the regulation is current. Procedures with a 010 or 090 global are excluded from the virtual option. Ask your coder or Medicare contractor how they read this for your service mix before relying on virtual supervision.

How common ortho visits usually sort out:

Split/shared visits (facility only)

A split or shared visit is an E/M visit performed in the facility setting by a physician and a PA in the same group.⁷ Office visits and nursing facility visits can't be billed as split/shared.⁷

Since January 1, 2024, CMS defines the "substantive portion" of a split/shared visit as more than half the total time, or a substantive part of the medical decision making.⁷ Modifier FS must be on the claim.⁷ AAPA also lists a requirement that the physician sign and date the medical record.²

Watch for this discrepancy: AAPA's March 2026 reimbursement summary lists only the time-based definition for 2024 onward.² The CMS manual includes both time and MDM.⁷ Follow the CMS manual.

The global period hides PA work

Codes with a 090 global are major surgeries, and their payment includes related follow-up visits during the post-op period.⁸ When a PA sees those post-op patients, that work is part of the surgeon's package. The Medicare contractor decides how much PA involvement in the global package it recognizes, consistent with its current practice.⁹ Physicians in the same group and specialty must bill and be paid as if they were a single physician.¹⁰

AAPA notes that a portion of a PA's productivity "may be 'hidden' in the global surgical package."¹¹ If you judge PAs by wRVUs alone, you will undercount the ones who carry your post-op clinic. Track global visits separately, for example with CPT 99024.

One study of total hip follow-up found that visits jumped right after the 90-day global period ended, and surgeons took a larger share of visits after it (73.8% of patients seen by surgeons in weeks 1 to 13, 86.8% after).¹² Look at your own pattern and ask whether it reflects clinical need or billing incentives.

---

2. First assist economics

How assistant-at-surgery pays

• A physician assistant-at-surgery is paid 16% of the surgical payment.¹³
• A PA is paid 85% of that, which is 13.6% of the surgeon's fee schedule amount.¹⁴
• The claim must carry modifier AS.¹⁴
• Medicare won't pay any assistant at surgery for procedures where a physician assists in fewer than 5% of cases nationally.¹³ Billing the patient for a non-payable assistant service can trigger penalties under SSA §1842(j)(2).¹³
• In a teaching hospital with a training program in the related specialty, Medicare generally doesn't pay for an assistant at surgery when a qualified resident is available. There are exceptions, including surgeons who never involve residents, and emergencies.¹⁵

The fee schedule's Assistant at Surgery indicator controls whether a code pays:¹⁶

• 2: payment restriction does not apply. Assistant may be paid.
• 0: payment restriction applies unless supporting documentation establishes medical necessity.
• 1: statutory payment restriction applies. Assistant may not be paid.
• 9: concept does not apply.

Common ortho procedures (2026 October release)¹⁶

Indicators can change with each quarterly release. Check the current RVU file before you rely on this table. A PA can still add value on indicator-1 cases (turnover, closing, a second room); just don't count on Medicare assistant revenue for them.

Worked example: a total knee in 2026

The formula is (Work RVU × Work GPCI + PE RVU × PE GPCI + MP RVU × MP GPCI) × conversion factor.¹⁷ For 27447 in the facility setting, the 2026 RVUs are Work 19.11, PE 11.58 and MP 4.02, and the 2026 conversion factor is $33.4009 for non-qualifying APM participants ($33.5675 for qualifying participants).¹⁶ A PA assistant gets that amount × 0.16 × 0.85.

Figures are our calculation from the CMS files, before sequestration, the coinsurance split and any multiple-procedure adjustment.¹⁷ Check them against the CMS Physician Fee Schedule Look-Up Tool for your locality.

For comparison, clinic visits (non-facility):¹⁷

To estimate your own first-assist revenue: take last year's Medicare cases on indicator-2 codes where a PA assisted or could have, and multiply each by the surgeon's locality amount × 0.136. Add indicator-0 cases only where you documented necessity.

The assistant fee on a knee is modest. The bigger value of a first-assist PA usually shows up in how many cases a surgeon can do and how the day runs, which is what the next section covers. It's also worth knowing that, across Medicare from 2010 to 2020, APPs were used as surgical assistants less often over time in every specialty studied, including orthopaedics.¹⁸

---

3. Deploying PAs: what the evidence says

The studies, including the ones that didn't show a benefit

The pattern: a PA can make the OR faster, but whether they do depends on the team and the surgeon (compare Hazzard with Cahan). The more reliable gains come from moving the surgeon's time to where only the surgeon can do the work.

APP involvement in joint replacement is growing fast. In Medicare data from 2014 to 2023, APPs providing arthroplasty-related services increased 87%, against 17% for surgeons.²⁷

Clinic templates

Patients have clear views on who should do what. In one survey, patients preferred the surgeon for follow-up of abnormal tests (82%), the first post-op visit (81%) and new-patient visits (81%). They were comfortable with the PA for pre-op teaching (73%).²⁸ 68% said they considered the midlevel provider's training background when choosing a surgeon.²⁸

A template built on that:

• Surgeon: new patients, surgical decision visits, first post-op visit, abnormal results.
• PA: pre-op education and optimization, later post-op visits, established follow-ups, injections and simple fracture care.
• Paired days: PA sees established patients in parallel rooms while the surgeon sees new consults.

Remember the billing: new patients and new problems seen by a PA bill under the PA's NPI at 85%, not incident-to.²

OR first assist

• Assign PAs first to indicator-2 cases, where the assistant fee is payable.¹⁶
• Confirm the practice agreement authorizes first and second assist.²⁹
• Confirm the supervising surgeon holds privileges at that hospital.³⁰
• If you're considering two rooms per surgeon, the Bohm PA-enabled "double OR" model is the published example.²²
• Measure your own OR times before and after. Given Cahan, don't assume the gain.²¹

Fracture clinic

Medicare lists setting casts for simple fractures and interpreting x-rays among covered PA services.³ For workers' comp patients, the supervising physician is the treating physician and must make and sign temporary disability determinations; under a protocol, the PA may authorize up to three days off work.³¹ Build your fracture clinic workflow so the physician's signature step isn't a bottleneck.

Post-op

Post-op visits inside a 090 global are part of the surgeon's package and don't generate separate payment.⁸ A PA carrying these visits frees surgeon clinic time for new patients. Note that patients prefer the surgeon for the first post-op visit;²⁸ the PA can take the later ones.

Call and trauma

• The Althausen data are the best case for PAs in hospital-based ortho: the value showed up in speed to evaluation and to surgery, not in the PA's own collections.²³
• In a general acute care hospital, the PA's supervising physician must hold privileges there.³⁰
• The physician must be reachable by phone or electronic communication when the PA examines the patient.³²

Scaling up

Since January 1, 2026, one physician may supervise up to eight PAs at a time, up from four.³³ If your model was capped by the old ratio, look again.

Measuring PA productivity fairly

• Don't judge by collections alone. The trauma program above recovered only half of PA cost in collections while cutting hours off time to surgery.²³
• Count global visits. RVU reports miss post-op work inside the global package.¹¹
• Use a reference point for volume, not a quota. Nationally, ortho PAs working 40+ hours a week see a mean of 68 and median of 60 patients a week.³⁴
• Watch for rules of thumb. We found no primary source for the often-repeated claim that PAs generate two to three times their salary in revenue, so we don't use it.

---

4. Retention

How many leave, and why

In NCCPA's 2025 data, 8.6% of PAs intended to leave their principal clinical position within 12 months (up from 7.8% in 2021).³⁵ Among those planning to leave, these factors were rated "very important":³⁵

In an AAPA survey of 1,261 PAs who had left a job, the top reasons were better work/life balance (16.8%), moving (16.7%), and better management, leadership or environment (13.6%). Better compensation or benefits came in at 11.9%.³⁶

At one academic center, APPs rated intrapractice partnership and collegiality lowest of all satisfaction factors. Bonuses, how rewards are distributed, and pay for work outside normal duties were also named as dissatisfiers.³⁷

Among ortho PAs specifically, 26.9% report one or more burnout symptoms, up 2.2 points since 2021. That's lower than primary care (34.7%), but not low.³⁴³⁵

What retention is worth

In one organization, a structured mentorship program raised retention of new PAs and NPs from 85% to 96% in year one and from 65% to 83% in year two. First-year productivity didn't differ between mentees and non-mentees (38th vs 37th percentile), so mentoring didn't cost output.³⁸ The authors valued the 15 retained clinicians at $1.29 million to $1.72 million in potential savings,³⁸ which works out to roughly $86,000 to $115,000 per departure avoided (our calculation). That's one organization's estimate, not an ortho or California benchmark.

In a pediatric academic system, a postgraduate APP fellowship cut turnover from 8.2% to 3.8%, and fellows reached productivity 4.2 months sooner than other hires.³⁹

What to do about it

• Assign a mentor to every new PA for at least the first year, with scheduled meetings.³⁸
• Pay for call and extra duties explicitly. Unpaid "outside normal duties" work is a named dissatisfier.³⁷
• Check pay against workload, not just against market. "Insufficient wages given the workload" is a top reason for planned departures.³⁵ Compare your PAs' pay with the BLS California figures and your own patient and case volume.
• Treat collegiality as a management task. Include PAs in practice meetings, case conferences and decisions about their templates.³⁷
• Protect family time. Predictable call and schedules address the work/life and family-care reasons directly.³⁵³⁶
• Ask about burnout in regular reviews, before it shows up as a resignation.
• Hold a stay interview once a year: what would make you leave, and what would make you stay?

---

5. PA utilization self-audit

Print this page. Answer each item yes or no. Every "no" is a place to look.

Practice agreement

• Signed and dated by each PA and supervising physician⁴⁰⁴¹
• Authorizes first and second assist for PAs who do it²⁹
• Names the diagnoses for any Schedule II furnishing⁴²
• States a deliberate chart-review policy (co-signature is optional in California)³⁰
• Reviewed since the 1:8 ratio took effect on January 1, 2026³³

Billing

• PA claims go out at full charge under the PA's NPI when the PA provided the service²
• Incident-to is used only for established-patient, physician-initiated plans in the office, with direct supervision²⁵
• Your coder has a written position on virtual direct supervision and the 010/090 exclusion⁶
• Split/shared visits are facility-only, carry modifier FS, and use the time-or-MDM test⁷
• Nobody believes a co-signature moves a PA's service to the physician's NPI²

First assist

• Modifier AS is on every PA assistant claim¹⁴
• Your schedulers know which of your common codes are indicator 2, 0 and 1¹⁶
• Indicator-0 cases have documented medical necessity when billed¹⁶
• You checked the current quarterly RVU file this quarter
• Teaching-hospital cases are screened for the resident-availability rule¹⁵

Deployment

• Surgeon time goes mainly to new patients, surgical decisions and first post-ops²⁸
• PAs carry pre-op teaching and later post-op visits²⁸
• Each PA's hospital work is covered by a supervising surgeon with privileges there³⁰
• You have measured OR time with and without the PA, not assumed it²¹

Productivity measurement

• Global-period post-op visits are tracked (for example, CPT 99024)¹¹
• Hospital-based PAs are judged on throughput as well as collections²³

Retention

• Every PA hired in the last two years has a named mentor³⁸
• Call and extra duties are paid explicitly³⁷
• Each PA had a pay, workload and burnout conversation in the last 12 months³⁵
• You know which of your PAs is most likely to leave, and why

---

A note from FirstAssistPA

This guide is from FirstAssistPA, a small placement service focused on orthopaedic PAs in California, founded by Anthony Adams. If you're hiring, your first placement is free. If you aren't happy with a PA we place, for any reason, we'll replace them at no charge.

---

Sources

1. CMS, Medicare Claims Processing Manual, Ch. 12, §110. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf Chapter Rev. 13316, issued 07-24-25; §110 Rev. 2656.
2. AAPA, "Payer Reimbursement Policies for PAs." https://www.aapa.org/download/48117/ PDF dated Mar 30, 2026. Accessed 2026-09-22.
3. CMS, Medicare Benefit Policy Manual, Ch. 15, §190. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/bp102c15.pdf Chapter Rev. 13889, issued 07-30-26; §190 Rev. 11288 (eff. 01-01-22).
4. CMS, Medicare Benefit Policy Manual, Ch. 15, §60.2. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/bp102c15.pdf §60.2 Rev. 1, 10-01-03; chapter Rev. 13889.
5. 42 CFR 410.26(b)(5). https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-B/part-410/subpart-B/section-410.26 eCFR point-in-time 2026-09-01.
6. 42 CFR 410.32(b)(3)(ii). https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-B/part-410/subpart-B/section-410.32 eCFR point-in-time 2026-09-01.
7. CMS, Medicare Claims Processing Manual, Ch. 12, §30.6.18. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf §30.6.18 Rev. 12604, issued 05-03-24, effective 01-01-24.
8. CMS, Medicare Claims Processing Manual, Ch. 12, §40.1. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf §40.1 Rev. 11287, eff. 01-01-22.
9. CMS, Medicare Claims Processing Manual, Ch. 12, §110.1. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf Chapter Rev. 13316, issued 07-24-25.
10. CMS, Medicare Claims Processing Manual, Ch. 12, §30.6.5. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf Chapter Rev. 13316, issued 07-24-25.
11. AAPA, "PA Productivity." https://www.aapa.org/advocacy-central/reimbursement/pa-productivity/ Accessed 2026-09-22.
12. Ratnasamy PP et al. PLoS One. 2024;19(4):e0300460. PMID 38635750. https://pubmed.ncbi.nlm.nih.gov/38635750/ (abstract).
13. CMS, Medicare Claims Processing Manual, Ch. 12, §20.4.3. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf Chapter Rev. 13316; Rev. 2656.
14. CMS, Medicare Claims Processing Manual, Ch. 12, §110.2. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf Chapter Rev. 13316; §110.2 Rev. 2656.
15. CMS, Medicare Claims Processing Manual, Ch. 12, §100.1.7. https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/clm104c12.pdf Rev. 811, eff. 01-01-06.
16. CMS, 2026 Physician Fee Schedule Relative Value File, RVU26D (October release, released 08/26/2026): RVU26D.pdf layout and PPRRVU2026_Oct_nonQPP.csv / PPRRVU2026_Oct_QPP.csv. https://www.cms.gov/files/zip/rvu26d-updated-08-26-2026.zip CPT descriptors © AMA.
17. DERIVED by the author from CMS RVU26D (Oct 2026 release): PPRRVU2026_Oct_nonQPP.csv and GPCI2026.csv, https://www.cms.gov/files/zip/rvu26d-updated-08-26-2026.zip Method: (Work RVU × Work GPCI + PE RVU × PE GPCI + MP RVU × MP GPCI) × $33.4009; PA = × 0.85; PA assistant at surgery = × 0.16 × 0.85. Before sequestration, coinsurance split and multiple-procedure adjustments.
18. Modi PK et al. J Am Coll Surg. 2025;241(5):853-863. PMID 40747891. https://pubmed.ncbi.nlm.nih.gov/40747891/ (abstract).
19. Hazzard SM et al. Arthrosc Sports Med Rehabil. 2023;5(1):e185-e192. PMID 36866287. https://pubmed.ncbi.nlm.nih.gov/36866287/ (abstract).
20. Quanbeck Z et al. J Surg Educ. 2025;82(1):103353. PMID 39616973. https://pubmed.ncbi.nlm.nih.gov/39616973/ (abstract).
21. Cahan EM et al. Bone Joint J. 2021;103-B(2):347-352. PMID 33517742. https://pubmed.ncbi.nlm.nih.gov/33517742/ (abstract).
22. Bohm ER et al. Can J Surg. 2010;53(2):103-108. PMID 20334742. https://pubmed.ncbi.nlm.nih.gov/20334742/ (abstract).
23. Althausen PL et al. J Orthop Trauma. 2016;30 Suppl 5:S40-S44. PMID 27870674. https://pubmed.ncbi.nlm.nih.gov/27870674/ (abstract).
24. Hollenbeck BK et al. Ann Surg. 2023;277(1):e40-e45. PMID 33914476. https://pubmed.ncbi.nlm.nih.gov/33914476/ (abstract).
25. Sharabianlou Korth MJ et al. Orthopedics. 2022;45(5):e252-e256. PMID 35576483. https://pubmed.ncbi.nlm.nih.gov/35576483/ (abstract).
26. Olsen AS et al. J Arthroplasty. 2023;38(6S):S77-S80. PMID 37001621. https://pubmed.ncbi.nlm.nih.gov/37001621/ (abstract).
27. Silvestre J et al. J Arthroplasty. 2026 (online May 7, 2026). PMID 42106085. https://pubmed.ncbi.nlm.nih.gov/42106085/ (abstract).
28. Manning BT et al. Iowa Orthop J. 2019;39(1):211-216. PMID 31413696. https://pubmed.ncbi.nlm.nih.gov/31413696/ (abstract).
29. 16 Cal. Code Regs. §1399.541(i) (Cornell LII copy). https://www.law.cornell.edu/regulations/california/16-CCR-1399.541 Amendment filed 7-19-2024, operative 10/1/2024. Accessed 2026-09-22.
30. Cal. Bus. & Prof. Code §3502 (subds. (a), (c), (f)). https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=BPC&sectionNum=3502 Stats. 2019, Ch. 707 (SB 697). Accessed 2026-09-22.
31. Cal. Labor Code §3209.10. https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=LAB&sectionNum=3209.10 Amended by Stats. 2004, Ch. 100. Accessed 2026-09-22.
32. Cal. Bus. & Prof. Code §3501 (subds. (f)(1), (k)). https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=BPC&sectionNum=3501 Amended by Stats. 2019, Ch. 707 (SB 697), eff. Jan 1, 2020. Accessed 2026-09-22.
33. Cal. Bus. & Prof. Code §3516(b). https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=BPC&sectionNum=3516 Amended by Stats. 2025, Ch. 194, Sec. 15 (AB 1501), eff. Jan 1, 2026. Accessed 2026-09-22.
34. NCCPA, 2025 Statistical Profile of Board Certified PAs by Specialty. https://www.nccpa.net/wp-content/uploads/documents/Reports/Statistical-Profile-of-Board-Certified-PAs-by-Specialty.pdf © NCCPA 2026 (2025 data). Accessed 2026-09-22.
35. NCCPA, 2025 Statistical Profile of Board Certified PAs. https://www.nccpa.net/wp-content/uploads/documents/2025-Statistical-Profile-of-Board-Certified-PAs.pdf © NCCPA 2026 (data as of Dec 31, 2025). Accessed 2026-09-22.
36. Reed H, Bernard K, Smith N. JAAPA. 2021;34(8):43-47. PMID 34320540. https://pubmed.ncbi.nlm.nih.gov/34320540/ (abstract). AAPA 2019 survey, n = 1,261 PAs who left a job.
37. Venegas B et al. J Healthc Manag. 2023;68(1):15-24. PMID 36602452. https://pubmed.ncbi.nlm.nih.gov/36602452/ (abstract).
38. Yun B et al. J Am Assoc Nurse Pract. 2025;37(10):573-581. PMID 39774034. https://pubmed.ncbi.nlm.nih.gov/39774034/ (abstract).
39. Merck T et al. J Pediatr Health Care. 2026;40(2):210-218. PMID 41528294. https://pubmed.ncbi.nlm.nih.gov/41528294/ (abstract). Pediatric multi-specialty program.
40. Cal. Bus. & Prof. Code §3502.3 (subd. (a)). https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=BPC&sectionNum=3502.3 Amended by Stats. 2019, Ch. 707 (SB 697), eff. Jan 1, 2020. Accessed 2026-09-22.
41. 16 Cal. Code Regs. §1399.540 (Cornell LII copy). https://www.law.cornell.edu/regulations/california/16-CCR-1399.540 Current on LII as of 2026-09-22.
42. Cal. Bus. & Prof. Code §3502.1 (subds. (b)(2), (d), (e), (g)). https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=BPC&sectionNum=3502.1 Stats. 2019, Ch. 707 (SB 697). Accessed 2026-09-22.