Tuesday, November 17, 2026

Weightlifting after shoulder arthroplasty: when can I start, how much can I lift, and which type of replacement is best for a lifter?

Summary

About 45 percent of Americans do some form of resistance training.1 When considering a shoulder arthroplasty, these individuals want to know three things: when they can start lifting, how much they can lift, and which type of replacement will let them lift the most with the least risk.

We lack robust evidence with which to answer these questions. What we think we know is based on opinion and uncontrolled clinical series.

Most people who lifted weights before a shoulder replacement will lift again. In recent studies, more than 90 percent of weightlifters returned to lifting, most within the first year. Light strengthening usually begins about 3 months after surgery, and heavier lifting usually waits until 4 to 6 months or later. In the one study that measured maximum lifts before and after surgery, lifters regained a little more than half of their previous bench press and military press. No study tells us how much weight is safe. Therapists’ guidelines advise lighter weights with more repetitions, with the hands kept in front of the body and out of the overhead position.

In the first two years after surgery, patients with anatomic and reverse total shoulders returned to lifting at similar rates. The two types differ in their risks. The main long-term concern with an anatomic total shoulder is loosening of the plastic glenoid component. A reverse carries a risk of stress fracture of the acromion, which is highest in older women with weak bone and in patients with inflammatory arthritis. The ream-and-run and the pyrocarbon hemiarthroplasty avoid a plastic glenoid component, but there are no published data on lifting after either. They have their own risks.

How many lifters return?

Most weightlifters do get back to lifting. In an ASES multicenter study of 175 weightlifters followed for about two years, 93 percent returned to lifting and 79 percent said their performance was the same as or better than before surgery.1 In a single-surgeon series of 184 weightlifters followed for an average of five and a half years, 70 percent reported no difficulty lifting, while another 22 percent reported some difficulty.2

When can I start lifting?

·      Most protocols protect the shoulder for the first 6 weeks and begin strengthening at about 12 weeks.

·      After an anatomic total shoulder, the early weeks protect the healing subscapularis tendon repair. The therapists’ consensus statement begins strengthening at 12 weeks and advanced return to sport after 4 months.3

·      After a reverse total shoulder, one multicenter protocol began light resistance exercises such as rows and curls at 7 to 12 weeks and progressive weight training at 12 weeks.4 A recent review suggests waiting at least 6 months before heavy work.5

·      In practice, about 1 in 5 weightlifters had returned to lifting by 3 months after surgery. By 6 months about 6 in 10 had returned, and by 1 year about 9 in 10 had.1 In two small series of lifters with an anatomic total shoulder using an inlay glenoid component, lifters returned at an average of about 16 weeks in one6 and at a median of 8.5 months in the other.7

·      Recovery after a ream-and-run is gradual. Comfort and function keep improving until about 20 months after surgery.8

·      Patients with osteoporosis who have a reverse may need a slower start. In one database study, starting physical therapy within 6 weeks of surgery was associated with more acromial stress fractures in these patients.9 A recent review points out that this study cannot prove cause and that the best timing and intensity of rehabilitation have not been defined.10

How much can I lift?

No study has measured how much weight is safe after a shoulder replacement.

·      Only one study has measured how much patients could lift. Eighteen male weightlifters with an inlay anatomic total shoulder were followed for an average of about 7 years. On average, they regained 57 percent of their previous one-repetition maximum on the bench press, 56 percent on the military press and 73 percent on the single-arm overhead press. Seventy-eight percent returned to the same or a higher level of activity. Two glenoid components were removed after injuries.7

·      In an earlier series of 16 weightlifters with the same type of implant, 44 percent said they were lifting heavier weights than before surgery and 50 percent the same weights. The weights themselves were not reported. At an average of about 3 years, there was no loosening on x-ray and no revision.6

·      The therapists’ consensus statement for anatomic total shoulders recommends lighter weights with higher repetitions, keeping the weights in front of the body and below shoulder level. The therapists did not agree on a maximum weight. Some suggested a limit of 15 pounds, and others suggested 25 pounds. The statement advises against heavy impact loading.3

·      After a reverse, the multicenter protocol told patients to keep their hands within eyesight and their elbows slightly bent and to minimize overhead movements. Military press, behind-the-head pulldowns and wide-grip bench press were not allowed.4

·      In the single-surgeon series, patients had no lifting limit after 12 weeks. They were told only that heavier weights put greater loads on the joint and might affect loosening and wear.2 At an average of five and a half years, most were lifting with little or no difficulty.

·      We found no study of lifting load as a risk factor for acromial stress fracture after a reverse. The strongest known predictors are weak bone and rotator cuff deficiency.11

·      Consider machines, cable rows and neutral-grip or close-grip pressing. Avoid behind-the-neck presses and pulldowns, deep dips, and Olympic or kipping lifts.

Which type of replacement gives me the most lifting with the least risk?

The studies do not show a clear winner for lifting.

·      The multicenter study compared 54 patients with a reverse to 54 with an anatomic total shoulder, all with osteoarthritis and matched for age and sex. Return to lifting was nearly the same (98 and 96 percent). Patients with a reverse more often kept up how often they lifted (91 vs 72 percent) and more often reported improved deltoid raises (83 vs 50 percent). Bench press, overhead press, curls and push-ups did not differ.1

·      A companion paper from the same group found similar return to sport after the two types. The reverse group more often reported stable or improved performance (88 vs 79 percent).12 Its weightlifters are the same patients as in the first study.

·      In the single-surgeon series, 74 percent of patients with an anatomic replacement returned to lifting without difficulty, compared with 57 percent of those with a reverse. This difference was not statistically significant. The reverse patients were older, and most had diagnoses other than osteoarthritis.2

The risks differ by type.

·      Anatomic total shoulder. The main long-term concern is loosening or wear of the plastic glenoid component. Symptomatic loosening can appear 5 to 10 years after surgery.13 When a loose glenoid component is revised, about one in five patients needs yet another operation.14 Heavy loading may hasten loosening, but follow-up in the lifting studies is too short to tell.

·      Reverse total shoulder. Stress fractures of the acromion or scapular spine occur after about 1 in 26 reverse shoulders, most often in the first year.15 The risk is higher with inflammatory arthritis, a massive rotator cuff tear, osteoporosis, prior shoulder surgery, cuff tear arthropathy, female sex and older age.11Patients with a stress fracture, and those with a stress reaction without a visible fracture, had poorer results than patients without either. Some of these fractures do not show on plain x-rays, so a CT scan may be needed.16 A review also suggests long-term limits on high-impact and contact sports because they may speed wear of the components.5

·      Ream-and-run and pyrocarbon hemiarthroplasty. These avoid a plastic glenoid component. We found no published data on return to lifting after either procedure. After a ream-and-run, recovery takes about 20 months.8 In our minimum 10-year follow-up, 4 of 34 ream-and-run patients (12 percent) had the humeral component exchanged, while none of 29 patients with an anatomic total shoulder needed another operation. Only half of the eligible patients responded, so these numbers are uncertain.13 In a series of 103 pyrocarbon hemiarthroplasties in patients with an average age of 56, 93 percent of those who played sports returned to them. Erosion of the glenoid was the main concern. When the pyrocarbon head was oversized, the revision rate rose from 1.3 percent to 25 percent.17

The authors of the multicenter study advise that their results be used to guide counseling after surgery rather than to choose an implant.1 For most patients, the condition of the shoulder should determine the type of replacement. A lifter with an intact rotator cuff and good bone can expect to return to lifting after either an anatomic or a reverse total shoulder. A lifter with osteoporosis or inflammatory arthritis who needs a reverse should be especially careful with early and heavy loading.1

Examples from our practice

Individual cases prove nothing about safety, but they show what is possible. Each of these weightlifters chose a ream-and-run to avoid the risks and limits of a plastic glenoid component.

·      A 70-year-old weightlifter and former competitive powerlifter had a ream-and-run. Ten years later his x-rays showed a stable arthroplasty, and he sent us a workout video to celebrate his 80th birthday. Read more

·      A 38-year-old bodybuilder was pressing 50-pound dumbbells 10 months after his ream-and-run. He reported only a little discomfort during and after workouts. Read more

·      A 40-year-old weightlifter and firefighter returned to firefighting 10 months after his ream-and-run. Read more

·      A weightlifter who had ream-and-runs on both shoulders reported that about a year after the second side he was pain-free during and after lifting about 90 percent of the time. Read more

Cautions

·      Most of these studies relied on what patients reported. Only one small series measured how much weight patients lifted, and none compared outcomes for light and heavy lifting.

·      Most follow-up was about two years. Glenoid loosening and component wear develop over many more years.

·      Patients were not randomized to the type of replacement.

These studies tell us what patients chose to do and how they felt about it. They do not show that unrestricted heavy lifting is safe for the long-term survival of the implant.

What does the field need to do?

Current studies tell us whether patients returned to lifting. They do not tell us how much loading is too much, whether heavy lifting shortens the life of the implant, or whether the answers differ by type of implant. Several kinds of studies would help.

·      Measure what patients actually lift. Only one small series has recorded patients’ maximum lifts before and after surgery.7 No study has related the weights, exercises, frequency and years of lifting to the survival of the implant. “Returned to weightlifting” counts a patient doing 10-pound curls the same as one bench-pressing 225 pounds. Future studies need a standardized lifting questionnaire, ideally supported by training logs or app data, so that outcomes can be related to the amount of loading.

·      Follow lifters for at least 10 years. Glenoid loosening, component wear and glenoid erosion develop over many years, but the lifting studies stop at about 2 years. A short activity questionnaire added to registries and large multicenter databases, linked to revision surgery and serial x-rays, would show whether heavy lifters fail sooner and with which implants.

·      Measure the loads produced by common lifts. Instrumented implants and validated computer models could show the forces on the joint and the strain on the acromion during the bench press, overhead press, row and pulldown. This would put the lists of lifts to favor or avoid on an evidence base and give a rational basis for any weight limit.

·      Randomize what can be randomized. Randomizing lifters to a type of implant is not realistic, but rehabilitation can be randomized. A trial of early versus delayed progressive loading would answer the question raised by the database study of acromial stress fractures, especially for patients with osteoporosis who have a reverse.9 So would a trial comparing restrictive with liberal advice about lifting.

·      Study the glenoid-sparing options. There are no lifting data for the ream-and-run or the pyrocarbon hemiarthroplasty, even though these procedures are often offered to young, active patients. Glenoid wear and erosion should be measured against activity level.

·      Reduce bias in who gets counted. In our 10-year study, only half of the eligible patients responded.13 In the single-surgeon series, only 5 percent of patients named weight training as their usual sport.2 Complete follow-up and analyses that account for which patients receive which implant are needed before comparisons between implants can be trusted.

·      Test bone-health treatment before a reverse. Weak bone predicts acromial stress fracture.11 Whether screening for and treating osteoporosis before surgery prevents these fractures has not been tested in a randomized trial.10

Until such studies are done, advice about lifting after shoulder arthroplasty will remain informed opinion.

Related posts

·      Arthritis in a 37 year old body builder/weight lifter with B2 pathoanatomy (March 2024)

·      Shoulder arthritis in a young weightlifter and firefighter (March 2024)

·      Shoulder arthritis in a 38 year old body builder (December 2023)

·      Bilateral ream and runs in a weight lifter: one year after second side (August 2022)

·      Arthritis surgery for a 70 year old weight lifter/powerlifter: a ten year followup (March 2022)

·      The inlay glenoid arthroplasty in weight lifters (August 2021)

·      Ream and run for shoulder arthritis in a weight lifter (June 2021)

·      Canal sparing ream and run arthroplasty for a weight lifter (September 2020)

·      Ream and run and weight training (December 2019)

·      Shoulder arthritis in a body builder managed with a ream and run (November 2019)

·      Shoulder arthritis in a weight lifter: patient-reported outcome (July 2019)


How much lifting?

Osprey


Follow on twitter: https://twitter.com/shoulderarth

Follow on facebook: https://www.facebook.com/frederick.matsen

Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133/




References

1.     Diestel DR, Sudah S, Veale N, et al. Return to weightlifting after shoulder arthroplasty: an ASES multicenter study predicting performance after reverse shoulder arthroplasty and anatomic shoulder arthroplasty. J Shoulder Elbow Surg. 2026;35(7):1667-1676. doi:10.1016/j.jse.2026.02.015. PMID 41763290.

2.     Abdelshaheed J, Chatterji R, Levy J, et al. Return to weightlifting following anatomic and reverse shoulder arthroplasty. J Shoulder Elbow Surg. 2026;35(7):1660-1666. doi:10.1016/j.jse.2026.02.002. PMID 41720256.

3.     Kennedy JS, Garrigues GE, Pozzi F, et al. The American Society of Shoulder and Elbow Therapists’ consensus statement on rehabilitation for anatomic total shoulder arthroplasty. J Shoulder Elbow Surg. 2020;29(10):2149-2162. doi:10.1016/j.jse.2020.05.019. PMID 32534209.

4.     Krupp R, Ma CB, Nyland J, Getz C, Srinivasan M. Reverse total shoulder arthroplasty pain and function: new perspectives from a 10-year multicenter study at the 7-year follow-up. Arch Orthop Trauma Surg. 2023;143(7):4049-4063. doi:10.1007/s00402-022-04702-z. PMID 36436065.

5.     Lamb CJ, Ahmad A, Biedermann BM, et al. Rehabilitation and long term outcomes including return to work or sport following reverse total shoulder arthroplasty. Curr Rev Musculoskelet Med. 2025;18(4):173-181. doi:10.1007/s12178-025-09948-9. PMID 39920554.

6.     Yalcin S, Scarcella M, Everhart J, Samuel L, Miniaci A. Clinical and radiographic outcomes of total shoulder arthroplasty with a nonspherical humeral head and inlay glenoid in elite weight lifters: a prospective case series. Orthop J Sports Med. 2021;9(7):23259671211021055. doi:10.1177/23259671211021055. PMID 34377719.

7.     Barrett AA, Moutafis AD, Blashinsky ZA, et al. Return to sport and long-term performance levels in weight lifters treated for advanced glenohumeral arthritis with inlay total shoulder arthroplasty. Orthop J Sports Med. 2026;14(4):23259671261436433. doi:10.1177/23259671261436433. PMID 42052232.

8.     Gilmer BB, Comstock BA, Jette JL, Warme WJ, Jackins SE, Matsen FA. The prognosis for improvement in comfort and function after the ream-and-run arthroplasty for glenohumeral arthritis: an analysis of 176 consecutive cases. J Bone Joint Surg Am. 2012;94(14):e102. doi:10.2106/JBJS.K.00486. PMID 22810409.

9.     Su F, Kucirek N, Goldberg D, Feeley BT, Ma CB, Lansdown DA. Incidence, risk factors, and complications of acromial stress fractures after reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2024;33(1):65-72. doi:10.1016/j.jse.2023.06.008. PMID 37454923.

10.   Hsue LJ, Sridharan M, Tewari K, Newman-Hung NJ, Truong NM, Cheung EC. Acromial stress fractures in reverse total shoulder arthroplasty: diagnosis, prevention, and treatment. Curr Rev Musculoskelet Med. 2026;19:61. doi:10.1007/s12178-026-10054-7. PMID 42608488.

11.   Moverman MA, Puzzitiello RN, Glass EA, et al. Implant-positioning and patient factors associated with acromial and scapular spine fractures after reverse shoulder arthroplasty: a study by the ASES Complications of RSA Multicenter Research Group. J Bone Joint Surg Am. 2024;106(15):1384-1394. doi:10.2106/JBJS.23.01203. PMID 40305832.

12.   Corban J, Kirsch JM, Bowler A, et al. Return to sport after shoulder arthroplasty: an ASES multicenter analysis of sport-specific predictors of performance in reverse shoulder arthroplasty and total shoulder arthroplasty. J Shoulder Elbow Surg. 2026;35(9):2110-2119. doi:10.1016/j.jse.2026.03.014. PMID 41932675.

13.   Sharareh B, Whitson AJ, Matsen FA 3rd, Hsu JE. Minimum 10-year follow-up of anatomic total shoulder arthroplasty and ream-and-run arthroplasty for primary glenohumeral osteoarthritis. J Shoulder Elbow Surg. 2024;33(6):1276-1284. doi:10.1016/j.jse.2023.08.028. PMID 37777045.

14.   Sharareh B, Yao JJ, Matsen FA 3rd, Hsu JE. What is the optimal management of a loose glenoid component after anatomic total shoulder arthroplasty: a systematic review. J Shoulder Elbow Surg. 2023;32(3):653-661. doi:10.1016/j.jse.2022.10.006. PMID 36343791.

15.   Mahendraraj KA, Abboud J, Armstrong A, et al. Predictors of acromial and scapular stress fracture after reverse shoulder arthroplasty: a study by the ASES Complications of RSA Multicenter Research Group. J Shoulder Elbow Surg. 2021;30(10):2296-2305. doi:10.1016/j.jse.2021.02.008. PMID 33677115.

16.   Kriechling P, Hodel S, Paszicsnyek A, Schwihla I, Borbas P, Wieser K. Incidence, radiographic predictors, and clinical outcome of acromial stress reaction and acromial fractures in reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2022;31(6):1143-1153. doi:10.1016/j.jse.2021.11.012. PMID 34968697.

17.   Boileau P, Cointat C, Raynier JL, Schippers P, Ranieri R. Pyrocarbon hemiarthroplasty for the treatment of shoulder osteoarthritis in young, active patients: survival and risk factors for revision. J Shoulder Elbow Surg. 2026;35(2):421-437. doi:10.1016/j.jse.2025.06.021. PMID 40782928.

Sunday, November 15, 2026

When might a patient consider surgical revision for a failed arthroplasty, and when might they consider living with it?

In brief

In Kenny Rogers’s song “The Gambler” (you can Google the video), an old card player’s advice is that winning depends less on the cards you are dealt than on knowing which hands to keep playing, which to give up on, when to leave the table, and when to get out fast.

Patients and surgeons faced with an unsatisfactory shoulder replacement have the same kinds of choices. Some patients choose to accept a failed arthroplasty rather than undergo another operation with its share of uncertain outcomes. Others accept this uncertainty and decide to proceed with revision. In discussing the pros and cons of nonoperative and operative care, surgeons have the opportunity to share with patient what they know from the literature and from their own experience. Here is some of the published evidence bearing on that discussion.

Revision can help, but typically less than the first operation did. In 1,030 revisions from a high volume revision practice, the ASES score (0 to 100) rose on average from 33 to 58 [1]. In 320 revisions to a reverse total shoulder from the same practice, only about half of the patients reached an acceptable symptom state at one year, and about one in five needed yet another revision [2]. Compared with patients after a primary reverse, patients after a revision have more pain, less motion, lower satisfaction and more complications [3, 4].

The type of failure matters. Gains in patient-reported outcome were largest, about 30 points, for a worn glenoid after a hemiarthroplasty, a loose glenoid or rotator cuff failure after an anatomic total shoulder, and a failed baseplate of a reverse. They were smaller for instability (20 points), and smallest for a loose humeral stem (10 points, with one in three revised again) and a fracture around the implant (8 points) [1]. Revising a failed reverse does worse than revising other types of shoulder arthroplasty: one in three needed another revision [2]. 

Revision deserves consideration when:

·       a patient’s loss of comfort and function can be related to a mechanical problem that surgery can effectively address. For example, conversion to a reverse total shoulder is often effective for an anatomic arthroplasty with symptomatic glenoid bone, glenoid component or rotator cuff failure; in these cases the results approach those of a primary reverse [5, 6].

Revision may also deserve consideration when:

·       a patient’s loss of comfort and function can be related to a mechanical problem that surgery may address, but with less certain chances of success, such as major glenoid or humeral bone loss or disabling instability [7, 8].

Nonoperative management is often the better choice when:

·       the patient’s pain has no identified cause;

·       the main complaint is limited rotation, which revision does not reliably restore [5, 6];

·       the problem is an acromial or scapular spine stress fracture: internal fixation may help these fractures heal, but it has not been shown to give better function than treatment without surgery, and fixation has its own complications [9-11];

·       the shoulder has already been revised, because each further revision is less likely to succeed [2, 12];

·       the patient’s medical risk is high.

Many patients have chosen to live with an unsatisfactory arthroplasty. In one review, about one in four reverse shoulders had a complication but only one in ten was revised [13]. In a larger review of modern designs, a third of postoperative humeral fractures and about a quarter of dislocations were managed without revising the components [14]. In the national registry for England and Wales, 27 percent of patients with a reverse had an unsatisfactory shoulder score, but fewer than 5 percent of them were revised [15].

Who should do the revision? Because the decision to revise is tricky and a good result is harder to achieve than after a primary, the choice of surgeon matters: as always, the surgeon is the method. But that choice is confounded by distance and economics. Traveling to a distant, high-volume, experienced revision surgeon costs the patient money and time and gives up the convenience of in-person follow-up; furthermore, referring the patient away costs the local surgeon revenue. The published results of one of the largest revision practices are similar to those of multi-surgeon health systems, national registries and other centers [1, 2, 12, 16], so there is no evidence that only a handful of surgeons can do revisions well, but logic suggests that a complex revision may fare best in the hands of a surgeon with substantial revision experience. The most useful information is a surgeon’s own record: the patient may ask how many revisions they do, how many of their patients are satisfied, and how many need yet another operation [17].

The time a surgeon and patient spend weighing a revision should be proportional to how likely and how serious its adverse outcomes are. After revision, about half of patients do not reach an acceptable result, a quarter to a third have a complication, and one in eight to one in three need yet another operation [2, 5, 18]. The risks, which include persistent pain, instability, infection, loosening and fracture, deserve a longer, more specific discussion than before a primary arthroplasty, with the option of living with the shoulder included in the discussion as well.

Five questions for the patient and surgeon: Is the problem bad enough to justify another operation? Do we know why it failed? Can that cause be reliably fixed? Might giving more time and rehab make the shoulder better? Does the patient understand and accept the odds: that, on average, only half of revised shoulders reach an acceptable result and about one in five needs another operation?

The bottom line. A patient might reasonably choose revision when the shoulder problem significantly limits the quality of their life, its mechanical cause has been identified, and that cause is one that revision has a good chance of fixing. A comfortable shoulder with a worrying x-ray, pain without a cause, or stiffness alone usually tips the other way. The decision belongs to the patient, made with realistic expectations and, ideally, with a surgeon experienced in revision surgery. Like the gambler’s, the best play comes from reading the hand honestly.

The best way to avoid this decision is a primary arthroplasty that doesn’t fail. Each failure we study carefully teaches us how to do the next primary better [19, 20].

In detail

Results by the reason for revision

Reason for revision

ASES before

ASES after

Gain

Needed another revision

Glenoid loosening with cuff failure (anatomic total)

29

64

35

11%

Worn glenoid (hemiarthroplasty)

29

63

34

8%

Baseplate failure (reverse)

32

63

31

13%

Glenoid loosening (anatomic total)

31

59

29

13%

Rotator cuff failure (anatomic total or hemiarthroplasty)

34

62

28

12%

Instability (any implant)

32

52

20

17%

Infection (any implant)

35

52

17

26%

Humeral loosening (any implant)

37

47

10

33%

Periprosthetic fracture (any implant)

30

38

8

0%

From Kolakowski et al. [1]: 1,030 revisions in 859 patients by one surgeon, 2002 to 2022. Scores are from the 439 patients (51 percent) with at least one year of follow-up; patients who needed another revision were not included in the scores, so the gains are probably overstated.

Results by the type of implant that failed

 

Failed hemiarthroplasty (111)

Failed anatomic total (120)

Failed reverse (89)

Main reasons for revision

Rotator cuff failure 68%, glenoid wear 23%

Rotator cuff failure 50%, glenoid loosening 44%

Instability 44%, baseplate failure 34%, humeral loosening 19%

Needed another revision

13.5%

11.7%

32.6%

Implant survival at 10 years

74%

82%

60%

ASES score at 1 year

63

64

53

Reached an acceptable symptom state

53%

53%

37%

Satisfaction at 1 year (of 10)

7.4

7.9

6.8

From Patel et al. [2]: 320 first-time revisions to a reverse for mechanical failure, from the same practice. The acceptable symptom state was defined as an ASES score of at least 63.5 at one year.

Of the 58 shoulders that needed a second revision, 17 (29 percent) needed a third. Other reports agree that revising a failed reverse is the hardest case: a review of 355 such revisions found complications in 32 percent and re-revision in 27 percent [18], and the Australian registry found a 24 percent rate of second revision at 8 years [16]. Not every series agrees. A German study of 101 revision reverses found 85 percent survival at 10 years, with no difference by the type of implant that had failed [21].

When is revision likely to help?

No study has randomly assigned patients with a failed shoulder to revision or to living with it, but studies that scored patients before and after revision show where the gains are most likely.

Rotator cuff failure or a worn or loose glenoid. Converting a failed hemiarthroplasty or anatomic total shoulder to a reverse can restore elevation. In a review of 1,041 such revisions, the ASES score improved from 33 to 62, pain from 6.7 to 2.7 (of 10) and forward elevation from 59 to 108 degrees, with reoperation in 12.5 percent [5]. Compared with a primary reverse, results are somewhat worse: in one registry study, 72 percent of revised patients said they were better, compared with 95 percent after a primary [3]. A matched study found no improvement in rotation, baseplate loosening in 21 percent, and a 19-point lower ASES score when the old humeral stem was retained [6]. For a loose glenoid component, 72 percent of patients reached a minimal clinically important improvement in one series [22]; every revision strategy carries a reoperation rate of about 20 percent, but satisfaction is highest after conversion to a reverse (86 percent) [23]. Results are worse when the shoulder is also unstable [24].

A failed reverse. Revision for a failed baseplate gained 31 points [1]. Otherwise, gains are real but smaller than after revising other types of shoulder implants. In 33 failed reverses revised to a new reverse, pain fell from 6.5 to 2.0 and the ASES score rose from 31 to 68, but patients rated their shoulders at only 51 percent of normal, and 30 percent needed another operation [25]. In 52 others, complications occurred in 31 percent [4]. Results after revision for a loose humeral stem are the least predictable, with re-revision in 29 percent and a mean ASES score of 52 [26], although one smaller study found better results [27].

An unstable reverse. Here living with it rarely works: closed reduction avoids further surgery in only 29 percent of cases, and after a first revision 78 percent of shoulders are stable [7]. Stability can usually be achieved in the end (89 percent in one series), but it took an average of 3 operations after a failed primary reverse and 4.5 after a failed revision, and revising both the humeral and glenoid sides left fewer shoulders unstable (10 vs. 33 percent) than revising one side [8].

When might living with it be the better choice?

Many complications are never revised. In a review of 782 reverse total shoulders, 24 percent of patients had a complication, but only 10 percent had a revision [13]. In modern designs, each structural complication affects a few percent of patients [14, 28], yet a third of postoperative humeral fractures and about a quarter of dislocations were managed without revising the components [14]. The clearest evidence comes from the National Joint Registry for England and Wales: 27 percent of patients with a reverse had an unsatisfactory Oxford Shoulder Score, yet only 4.9 percent of these were revised, compared with 10.6 percent of unsatisfactory anatomic total shoulders and 13.9 percent of unsatisfactory hemiarthroplasties [15]. The authors suggest that older patients may accept a poorly functioning shoulder and that surgeons may be reluctant to undertake difficult revisions in high-risk patients. A low revision rate does not mean the shoulder succeeded.


Unsatisfactory patient-reported results compared with structural complications that are potential indications for revision after reverse total shoulder arthroplasty. The figures come from different cohorts with different denominators [14, 15, 28].

Acromial and scapular spine stress fractures. Most of these fractures, 88 to 94 percent, are managed without surgery [29, 30]. Fixation heals more of them, about 80 to 90 percent compared with about 45 percent without surgery [10, 11, 29], although one series reported healing in only 40 percent of acromial and 33 percent of scapular spine fractures after fixation [29]. Healing does not reliably bring a better result. In one series of 30 fractures, the 23 treated without surgery improved more (Constant score 35 to 61) than the 7 treated with fixation (32 to 45) [9]. Patients with these fractures do worse than patients without them, whether or not the fracture heals or is fixed: in a matched comparison, their ASES scores were 64 vs. 88 [10, 31, 32]. Fixation is difficult because the deltoid pulls the fragments apart and the bone is thin, and in a review of 18 fixed fractures, 5 had complications [11]. With nonoperative care, angulation of about 36 degrees or more or displacement of about 10 mm or more predicted unacceptable pain and function, which may help identify patients for whom fixation is worth its risks [33]. 

Pain without a clear cause, and limited rotation. Once the possibility of infection has been evaluated, unexplained pain is a weak reason for revision: if we don’t know why the shoulder hurts, we can’t be confident that another operation will fix it. Revision to a reverse improves elevation, not rotation; external rotation barely changed (20 to 26 degrees) in pooled data [5].

Who should do the revision?

Published results describe what particular surgeons achieved with their particular patients; they do not transfer to another surgeon using the same implant. The two Tampa studies come from one of the largest revision practices in the country, averaging 52 revisions a year [1]. Their re-revision rates, 15 percent overall [1] and 18 percent at about 4 years after revision to a reverse [2], are similar to those of Kaiser Permanente’s multi-surgeon system (13 percent at 3 years and 19 percent at 7) [12] and the Australian registry (20 to 24 percent at 8 years) [16]. After revision of a failed reverse, the Tampa rates (21 and 33 percent) bracket those of a pooled review (27 percent) [18] and another academic center (30 percent) [25].

These comparisons cannot settle the question. A referral practice probably treats the hardest cases, so matching others’ results may reflect better surgery; a surgeon who revises readily will record more re-revisions than one who accepts poor results [15]; and patients lost to follow-up or revised elsewhere are missed. The only direct evidence on experience comes from Kaiser Permanente: after a failed reverse, surgeons who had done 4 to 11 previous revisions had a much lower risk of re-revision than those who had done fewer than 4, but surgeons who had done more than 11 did not differ significantly [12].

Distance and economics. Consider a patient in rural Oregon whose reverse has failed. Having the revision done by a high-volume surgeon in Minnesota costs the patient money and time, may fall outside their insurance network, and gives up convenient in-person follow-up; if a complication arises after the patient returns home, a local surgeon who did not do the operation must manage it or send the patient back to Minnesota. Referring the patient away costs the Oregon surgeon revenue. The surgeon who did the original operation knows the shoulder’s history but may find it hard to be objective about why it failed.

Adverse outcomes to discuss before revision

Informed consent should be proportional to risk. The patient should hear, in numbers rather than words like “rare”, how often each adverse outcome occurs after a revision like theirs, and what it would mean for them. More than one visit, the presence of family members, and a written summary may help.

Adverse outcome after revision

How often

What it may mean

Not reaching an acceptable result (persistent pain or poor function)

About half of patients [2]

A shoulder that is better than before but still limiting

Any complication

23 to 35 percent [4-6, 18]

Further treatment, often surgery

Another revision

12 to 33 percent, highest after a failed reverse [1, 2, 18]

Each further revision is less likely to succeed

Instability or dislocation

The leading cause of re-revision (28 percent of re-revisions) [18]

Closed reduction or more surgery, sometimes several operations

Baseplate loosening

10 percent after a revision reverse vs. 1.2 percent after a primary [34]; 21 percent after converting a failed anatomic total [6]

Pain, loss of function, revision with bone grafting

Infection

4 percent needed revision for infection [2]; 15 percent of revisions for a loose glenoid had positive cultures [23]

Antibiotics, often more surgery, sometimes a staged revision

Humeral loosening or periprosthetic fracture

About 4 and 2 percent needed revision [2]

Complex reconstruction, with poorer results

Acromial or scapular spine fracture

About 2 to 3 percent after reverse arthroplasty [14, 31]

Usually treated without surgery; fewer than half heal without surgery, and function is often reduced

Approximate rates from the studies discussed in this post. The risk for an individual patient depends on what failed, the bone available, prior surgery and the patient’s health.

Questions about the evidence

How much did patients improve? Patel et al. did not report preoperative scores. The companion study did, but only half of its patients had a year of follow-up.  We need data on how much revision improves the patients' comfort and function.

Implant survival versus success? Survivorship counts only subsequent operations, so a patient with a poor result who declines another surgery counts as a success. Asking each patient whether they are satisfied and whether the shoulder is better, the same or worse would tell us more [17].

How to consider infections? Patel et al. excluded 182 of 646 first-time revisions (28 percent) because cultures or pathology showed infection. Infection and mechanical failure can go together.



Considering revision


Red-breasted Nuthatch


Follow on twitter: https://twitter.com/shoulderarth

Follow on facebook: https://www.facebook.com/frederick.matsen

Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133/

References

1.         Kolakowski L, Stadecker M, Kucharik M, et al. Trends in 1030 revision shoulder arthroplasty cases: changing rates, indications, and treatments. J Shoulder Elbow Surg. 2026;35:1434-43. doi:10.1016/j.jse.2025.11.023

2.         Patel R, Kucharik MP, Schmidt CM II, et al. Revision reverse shoulder arthroplasty after failed shoulder arthroplasty: long-term survivorship and risk factors for failure. J Shoulder Elbow Surg. 2026;35:2346-53. doi:10.1016/j.jse.2026.04.028

3.         Tobin JG, Thomas SK, Elwell JA, et al. Anatomic total shoulder arthroplasty revised to reverse total shoulder arthroplasty: clinical and radiographic outcomes compared to primary reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2025;34:1525-31. doi:10.1016/j.jse.2024.09.019

4.         Tobin JG, Guareschi ME, Elwell JA, et al. Revision reverse total shoulder arthroplasty: clinical and radiographic outcomes compared to primary reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2026;35:1890-5. doi:10.1016/j.jse.2026.01.020

5.         Welch JM, Bethell MA, Meyer AM, et al. Outcomes and complications of revision reverse shoulder arthroplasty after failed primary anatomic shoulder arthroplasty or hemiarthroplasty: a systematic review. J Shoulder Elbow Surg. 2024;33:2306-13. doi:10.1016/j.jse.2024.03.053

6.         Hao KA, Simcox TG, Hones KM, et al. Conversion of failed anatomic total shoulder arthroplasty to reverse shoulder arthroplasty yields similar pain and functional outcomes to primary reverse shoulder arthroplasty but has a higher baseplate failure rate: a matched cohort study. J Am Acad Orthop Surg. 2025;33:e1215-25. doi:10.5435/JAAOS-D-24-01296

7.         Howard WH, Robinson LT, Vasimalla M, Faulkner H, Yeoh T, Low AK. Instability after reverse shoulder arthroplasty: a systematic review of interventions and clinical outcomes. J Shoulder Elbow Surg. 2025;34:2735-45. doi:10.1016/j.jse.2025.02.059

8.        Melbourne C, Munassi SD, Ayala G, et al. Revision for instability following reverse total shoulder arthroplasty: outcomes and risk factors for failure. J Shoulder Elbow Surg. 2023;32:S46-52. doi:10.1016/j.jse.2023.01.023

9.         Schenk P, Aichmair A, Beeler S, Jentzsch T, Gerber C. Clinical results of conservative versus operative treatment of acromial and scapular spine fractures following reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2022;31:2076-81. doi:10.1016/j.jse.2022.03.005

10.      Kriechling P, Hodel S, Paszicsnyek A, et al. Incidence, radiographic predictors, and clinical outcome of acromial stress reaction and acromial fractures in reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2022;31:1143-53. doi:10.1016/j.jse.2021.11.012

11.       Cassidy JT, Paszicsnyek A, Ernstbrunner L, Ek ET. Acromial and scapular spine fractures following reverse total shoulder arthroplasty: a systematic review of fixation constructs and techniques. J Clin Med. 2022;11:7025. doi:10.3390/jcm11237025

12.      Dillon MT, Prentice HA, Burfeind WE, Singh A. Risk factors for re-revision surgery in shoulder arthroplasty. J Am Acad Orthop Surg. 2020;28:e1049-58. doi:10.5435/JAAOS-D-19-00635

13.      Zumstein MA, Pinedo M, Old J, Boileau P. Problems, complications, reoperations, and revisions in reverse total shoulder arthroplasty: a systematic review. J Shoulder Elbow Surg. 2011;20:146-57. doi:10.1016/j.jse.2010.08.001

14.      Shah SS, Roche AM, Sullivan SW, et al. The modern reverse shoulder arthroplasty and an updated systematic review for each complication: part II. JSES Int. 2021;5:121-37. doi:10.1016/j.jseint.2020.07.018

15.       O’Malley O, Davies A, Rangan A, Sabharwal S, Reilly P. Is there a difference in thresholds for revision between shoulder arthroplasty types? A National Joint Registry study. PLoS One. 2025;20:e0330975. doi:10.1371/journal.pone.0330975

16.      Gill DRJ, Page RS, Graves SE, Rainbird S, Hatton A. The rate of 2nd revision for shoulder arthroplasty as analyzed by the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). Acta Orthop. 2021;92:258-63. doi:10.1080/17453674.2020.1871559

17.       Matsen FA 3rd. How do our surgical patients do in the long term? A critical but difficult question. Shoulder Arthritis Blog. November 11, 2026. https://shoulderarthritis.blogspot.com/2026/11/long-term-followup-of-our-patients.html

18.      Wu L, Al-Asadi M, Abdel Khalik H, et al. Revision of reverse total shoulder arthroplasty: a scoping review of indications for revision, and revision outcomes, complications, and rerevisions. J Shoulder Elbow Surg. 2025;34:2645-54. doi:10.1016/j.jse.2025.02.039

19.      Menendez ME, Matsen FA 3rd. Learning from surgical failures. J Bone Joint Surg Am. 2026;108:547-8. doi:10.2106/JBJS.25.01110

20.      Hackett DJ, Hsu JE, Matsen FA 3rd. Primary shoulder hemiarthroplasty: what can be learned from 359 cases that were surgically revised? Clin Orthop Relat Res. 2018;476:1031-40. doi:10.1007/s11999.0000000000000167

21.      Trefzer R, Leisner M, Weishorn J, et al. Long-term results and implant survival of revision reverse shoulder arthroplasty after a mean follow-up of ten years. Bone Joint J. 2025;107-B:1212-9.

22.      Hao KA, Hones KM, O’Keefe DS, et al. Quantifying success after first revision reverse total shoulder arthroplasty: the minimal clinically important difference, substantial clinical benefit, and patient acceptable symptomatic state. J Shoulder Elbow Surg. 2023;32:e516-27. doi:10.1016/j.jse.2023.03.032

23.      Sharareh B, Yao JJ, Matsen FA 3rd, Hsu JE. What is the optimal management of a loose glenoid component after anatomic total shoulder arthroplasty: a systematic review. J Shoulder Elbow Surg. 2023;32:653-61. doi:10.1016/j.jse.2022.10.006

24.      Aibinder WR, Schoch B, Schleck C, Sperling JW, Cofield RH. Revisions for aseptic glenoid component loosening after anatomic shoulder arthroplasty. J Shoulder Elbow Surg. 2017;26:443-9. doi:10.1016/j.jse.2016.08.009

25.      Olson JJ, Granger CJ, Hill JR, Aleem AW, Keener JD, Zmistowski BM. Revision of failed reverse total shoulder arthroplasty with reverse: short-term clinical outcomes. J Shoulder Elbow Surg. 2024;33:908-15. doi:10.1016/j.jse.2023.07.030

26.      Layuno-Matos JG, Hutchinson A, Karadimas T, Frankle MA. Revision for humeral stem loosening: a systematic review. J Shoulder Elbow Surg. 2025;34:1313-21. doi:10.1016/j.jse.2024.08.053

27.      Hao KA, Boschert EN, O’Keefe DS, et al. Comparison of clinical outcomes of revision reverse total shoulder arthroplasty for failed primary anatomic vs. reverse shoulder arthroplasty. JSES Int. 2023;7:257-63. doi:10.1016/j.jseint.2022.11.003

28.      Shah SS, Gaal BT, Roche AM, et al. The modern reverse shoulder arthroplasty and an updated systematic review for each complication: part I. JSES Int. 2020;4:929-43. doi:10.1016/j.jseint.2020.07.017

29.      Brusalis CM, Taylor SA. Periprosthetic fractures in reverse total shoulder arthroplasty: current concepts and advances in management. Curr Rev Musculoskelet Med. 2020;13:509-19. doi:10.1007/s12178-020-09654-8

30.      Marigi EM, Eboh S, Marigi IM, et al. Acromial and scapular fractures after reverse shoulder arthroplasty: comparison of 3018 reverse total shoulders by inlay and onlay humeral component design. J Shoulder Elbow Surg. Published online 2025. doi:10.1016/j.jse.2024.09.029

31.      Routman HD, Simovitch RW, Wright TW, et al. Acromial and scapular fractures after reverse total shoulder arthroplasty with a medialized glenoid and lateralized humeral implant: an analysis of outcomes and risk factors. J Bone Joint Surg Am. 2020. doi:10.2106/JBJS.19.00724

32.      Moore JW, Reis R, Welsh ME, Rogalski BL, Friedman RJ, Eichinger JK. Acromion and scapular spine fractures following reverse total shoulder arthroplasty: operative versus nonoperative management. Semin Arthroplasty JSES. Published online 2025. doi:10.1053/j.sart.2025.03.001

33.      Rhee SM, Kantanavar R, Woo CW, et al. Acromial and scapular spine fractures following reverse total shoulder arthroplasty: treatment dilemma and radiographic prognosticators for enhanced clinical outcomes. A multicenter study. J Shoulder Elbow Surg. Published online 2025. doi:10.1016/j.jse.2025.05.005

34.      Bitzer A, Rojas J, Patten IS, Joseph J, McFarland EG. Incidence and risk factors for aseptic baseplate loosening of reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2018;27:2145-52. doi:10.1016/j.jse.2018.05.034