To see what we've been doing.
Saturday, July 11, 2026
Tuesday, July 7, 2026
Two year outcomes for reverse total shoulder for osteoarthritis and the effect of survivorship bias.
How survivorship bias can affect reports of two-year outcomes
When we quote a two-year success rate for an operation, the figure typically describes the outcomes for the patients who came back to be assessed at two or more years after surgery.
Those who did not return are not a random sample of all the patients having the surgery. Patients lost to follow-up tend to have fared worse than those who complete follow-up, because the reasons they drop out are often themselves adverse — death, revision, or disappointment with an early result that leads them to transfer their care elsewhere [1, 2, 3].
Excluding patients who have done poorly prior to the two-year mark — considering only those who remain for the two-year analysis — makes the operation look better than it actually was, a phenomenon known as survivorship bias [4].
A model of survivorship bias
Here is an illustrative hypothetical example of a thousand patients having an RSA for cuff-intact osteoarthritis with two-year follow-up.
In the first six months patients were lost as follows: 20 shoulders revised, 5 patients dead, 15 transferred to another practice, and 40 who simply stopped responding — 80 in all. Of note, certain problems occur early in the postoperative period: acute infection, instability, acromial fracture, and life-limiting frailty.
During the second six months there were 7 more revisions and 8 deaths, 25 transfers, and 110 who did not respond for unknown reasons — 150 in the interval.
The second-year losses are almost entirely loss of contact rather than loss of the shoulder: 3 revised and 12 more dead, 35 gone to other practices, and 200 more who stopped answering — 250 in that interval.
By the time of the two-year analysis, 30 of the 1,000 have been revised, 25 have died, and 75 have transferred their care; 350 more have gone quiet without a recorded reason. That is 480 lost from view, none of whom are included in the two-year analysis of results; only 520 remain available for study.
Figure 1. A hypothetical cohort of 1,000 patients followed over two years. At each interval, patients drop out of view for four kinds of reasons, and the mix changes as time passes. Of the 520 analyzed at two years, 9 out of 10 report success; but out of all 1,000 operated on, a successful outcome is documented for only about 4.7 out of every 10.
The four reasons carry different weight, and their proportions change over the two years. Revisions come early: RSA’s dominant early failures — acute infection, instability, and acromial fracture — appear mostly in the first months, so revisions decline from 20 in the first half-year to 3 in the second year. Deaths run the other way, accumulating with time in an elderly group, from 5 to 8 to 12 across the intervals. Transfers of care, and above all patients who simply stop responding, grow steadily as contact is lost — from 40 silent patients in the first six months to 200 in the second year. By the time of the two-year analysis, loss of contact, not loss of the shoulder, accounts for most of the missing.
Applying the model of survivorship bias to actual data on two-year outcomes for RSA for osteoarthritis
The model becomes meaningful when anchored to what the literature reports. Two published figures set the scale.
The first is the rate of two-year follow-up. In a multicenter shoulder arthroplasty registry, only about half of patients — about 5 out of 10 — provided two-year patient-reported outcomes [5]; registries that send repeated reminders might improve the return to 8 out of 10.
The second is the success rate among those who do return: in high-volume single-surgeon series of RSA for this diagnosis, about 9 out of 10 report being better and satisfied [6, 7].
Consider the combined effect of these two rates. Of the 520 with a known two-year result, 9 out of 10 — 468 patients — report success; but across the full 1,000 who had the surgery, those 468 provide an overall documented success rate of only about 4.7 out of 10.
Had the follow-up rate reached 8 out of 10 rather than 5, about 800 would have a known result, and the same 9 out of 10 would give about 720 documented successes — roughly 7 out of 10.
The span from 4.7 to 7 out of 10 is set entirely by the follow-up rate.
The 4.7 out of 10 is the floor. It counts every patient without a documented success as a non-success, so that the true whole-cohort success rate (if it could be known) might be higher than 4.7 out of 10.
The missing patients are not a random subset of the cohort. Those lost to revision, death, or a transfer of care each had reason to fare worse than the returners. The larger unresponsive group has, on average, done less well than those who answer — though that association is weaker and less certain than for the other types of losses [1, 2, 3].
Reporting the returners’ 9 out of 10 as the rate for the overall cohort makes the operation look better than the data support [4].
However, a successful outcome is only documented for about 5 of every 10 patients considering all those having the surgery, not 9 out of 10.
A fuller accounting would mean following the patients who leave — the revised, the transferred, and above all the ones who quietly stop answering — well enough to know how they actually did. Until a series does that, the appropriate two-year estimate for successful outcome for RSA in osteoarthritis is somewhere above the documented 4.7 out of 10 rate for the entire cohort and below the rate of 9 out of 10 considering only the patients returning for two year analysis.
And two years is the favorable case. The effect of survivorship bias grows as follow-up lengthens: over five, ten, and fifteen years, follow-up falls further and the number of missing patients grows, so the distance between the returners’ success rate and the whole-cohort success rate only widens. The longer the follow-up a reported success rate claims, the greater the effect of survivorship bias.
It's about survivorship
Bald eagles
Montlake Cut
Follow on twitter/X: https://x.com/RickMatsen
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133//
References
[1] Solberg TK, Sørlie A, Sjaavik K, Nygaard ØP, Ingebrigtsen T. Would loss to follow-up bias the outcome evaluation of patients operated for degenerative disorders of the lumbar spine? A study of responding and non-responding cohort participants from a clinical spine surgery registry. Acta Orthop. 2011;82(1):56–63.
[2] Murnaghan ML, Buckley RE. Lost but not forgotten: patients lost to follow-up in a trauma database. Can J Surg. 2002;45(3):191–195.
[3] Torrens C, Martínez R, Santana F. Patients lost to follow-up in shoulder arthroplasty: descriptive characteristics and reasons. Clin Orthop Surg. 2022;14(1):112–118.
[4] Elston DM. Survivorship bias. J Am Acad Dermatol. Published online June 18, 2021.
[5] Patel M, Sekar MG, McDaniel L, Kisana HM, Sykes JB, Amini MH. Changes from baseline in patient-reported outcomes and patient satisfaction do not vary significantly between 1 and 2 years postoperatively after shoulder arthroplasty: a multicenter analysis of 2580 patients. Semin Arthroplasty JSES. 2025;35(2):235–245.
[6] Puzzitiello RN, Moverman MA, Glass EA, Swanson DP, Bowler AR, Le K, Kirsch JM, Lohre R, Jawa A. Clinically significant outcome thresholds and rates of achievement by shoulder arthroplasty type and preoperative diagnosis. J Shoulder Elbow Surg. 2024;33(7):1448–1456.
[7] Ahmed AF, Glass EA, Swanson DP, et al. Predictors of poor and excellent outcomes following reverse shoulder arthroplasty for glenohumeral osteoarthritis with an intact rotator cuff. J Shoulder Elbow Surg. 2024;33(6S):S55–S63.
Thursday, July 2, 2026
Robotics in Shoulder Arthroplasty: ASES Podcast 156.
Robotics in Shoulder Arthroplasty is a timely and informative ASES podcast hosted by Peter Chalmers and Brian Waterman with two guests: Michael Freehill and Vahid Entezari.
Here is a summary of the points the participants made. Where the guests referred to a specific study, a link to the primary source has been added so that readers can consult it directly and form their own view.
The episode presented enthusiasm tempered by candor about the current state of evidence. Both guests described themselves as early adopters, and both acknowledged that outcome data specific to robotic shoulder arthroplasty is not yet available. Their case for the technology rested primarily on precision in executing a preoperative plan, the ability to generate intraoperative data, enhanced capability in complex and deformed anatomy, and anticipated — as yet undemonstrated — long-term benefit. They also included a discussion of the market and workflow factors that shape adoption.
Here are some details.
Dr. Entezari suggested that shoulder arthroplasty has two distinct steps — planning and execution — and that while preoperative planning has advanced considerably, precision execution has historically lacked dedicated tools. He referenced previous work on the accuracy of planning and patient-specific instrumentation (PSI) showing that preoperative planning and PSI tightened the spread of deviation from the preoperative plan in comparison to manual techniques. See Accuracy of 3-Dimensional Planning, Implant Templating, and Patient-Specific Instrumentation in Anatomic Total Shoulder Arthroplasty.
From this he drew two conclusions: that surgeons doing manual-only surgery may underestimate their own deviation (“you don’t know what you don’t know”), and that enabling technologies — navigation, PSI, or robotics — become especially relevant in cases with significant deformity or difficult exposure. These technologies have the potential to reduce substantial deviation from the preoperative plan.
Dr. Freehill built on this from a training perspective, recalling high-volume leaders he trained under questioning their own accuracy: whether their humeral head cuts and glenoid version were as accurate as they desired. He described preoperative planning as a major advance and PSI as a further step for deformed anatomy, while noting that guides still have limitations related to soft tissue and exposure. He also emphasized that most shoulder arthroplasties are performed by lower-volume surgeons — those doing fewer than roughly ten a year — and suggested these surgeons might benefit from tools that help with execution.
Both guests placed the clearest value of robotics in complex cases: significant glenoid deformity (such as B2 and B3 morphology and higher degrees of retroversion), revision situations, and cases requiring bone graft or augment preparation.
Dr. Entezari described a reverse arthroplasty for arthropathy that had developed after a dislocation and fracture, with anterior glenoid bone loss, in which the robot was used to shape a humeral head autograft to fit the glenoid face, offering it as an example of capabilities that could extend to graft and augment work.
The guests pointed to potential new capabilities: a robotically controlled burr that can sculpt bone at angles a saw or reamer cannot achieve, the potential for subscapularis-sparing and minimally invasive approaches. They pointed out that robotics could capture potentially useful data such as how precisely the preoperative plan was carried out.
The hosts pointed to the mixed hip and knee literature, noting meta-analyses that have not shown a long-term benefit in clinical outcomes even though the rate of use of robotics was rapidly increasing.
The guests did not claim superiority in patient reported outcomes for robotics in shoulder arthroplasty. Dr. Entezari suggested that malpositioning may not affect these outcomes but that adherence to the preoperative plan may matter for long-term implant survival.
Dr. Freehill cited NICE’s assessment of robot-assisted orthopaedic surgery. The relevant NICE document is its early value assessment of robot-assisted surgery for hip and knee replacement (HealthTech guidance HTG743) in which patient-reported outcomes and complications were similar between robotic and conventional surgery and implant alignment was consistently more precise with robotics. The revision data were limited, and the committee was uncertain whether more precise alignment yields better clinical outcomes. He pointed to the need for head-to-head studies, while acknowledging such studies are difficult and slow to complete.
Dr. Entezari cited a Monte Carlo simulation whose authors concluded that, when value is judged by tail-risk rather than by mean accuracy, the principal contribution of robotic assistance in shoulder arthroplasty is risk containment — with clinical and economic implications that remain unproven to date — and that further research linking its use to patient outcomes is warranted and may ultimately redefine the technology’s value proposition. See Mako Robotic-Assisted Glenoid Preparation in Reverse Shoulder Arthroplasty: A Tail-Risk Reduction Perspective Compared with Manual and Patient-Specific Guide Techniques.
With respect to the learning curve for robotics, Dr. Freehill cited experience suggesting an inflection point around ten cases, with time efficiency approaching break-even, and drew an analogy to hip and knee arthroplasty, where operative time tends to normalize with experience.
[Parenthetically, we can observe that the inflection point of ten cases is close to the annual case volume of a “low-volume surgeon.”]
Both guests suggested the real barrier to adoption is neither time nor cost but workflow change.
Dr. Freehill discussed how the technology is presented to hospital administrators, whom he characterized as focused on time-zero costs rather than on anticipated ten- to fifteen-year outcomes or on accuracy data. He noted that an administrator is unlikely to be swayed by work comparing robotic glenoid preparation favorably with patient-specific instrumentation and, still more, with manual technique, because their attention stays on the bottom line. He acknowledged the numbers work against robotics: citing the existing literature, he said that breaking even by minimizing revisions would require avoiding roughly one revision in fifteen (about 7%), whereas revision rates in reverse arthroplasty are currently under 2%, so on that basis one is, in his words, already arguing against the cost-benefit case. See Is Robotic-Assisted Reverse Shoulder Arthroplasty Economically Justified? A Break-Even Analysis. He suggested the more persuasive pitch to an administrator combines several elements: that a robot from certain companies is often already in the hospital for hip and knee procedures; that adopting cutting-edge technology is anticipated to improve market share; and that greater accuracy can be framed as serving the patient’s long-term interest.
Dr. Entezari added that hospitals in a region compete for the same patients, and that device companies invest heavily in direct-to-consumer advertising, so patient demand — already seen in hip and knee — becomes part of the rationale. He noted that costs should fall as more surgeons use the technology. He reiterated that outcome data are not yet available and that early adopters have the opportunity to generate the needed data.
The guests contrasted navigation and robotics. Navigation has a smaller footprint, lower cost, and a lower barrier to entry. Robotics offers greater precision, but with a larger and more expensive early-generation factor, and with current systems limited in the procedures they can perform and the range of implants they can accommodate.
Dr. Entezari suggested neither overselling nor underselling the technology — telling patients that it makes the surgery more precise while being candid that its effect on their individual outcome is not known. Dr. Freehill described a mix of curiosity about cutting-edge technology and hesitation about a non-human element in the operation, and said he uses clinic time to explain the surgeon’s central role.
Both guests urged caution about how the technology is learned. They argued that younger surgeons should first become proficient in manual arthroplasty, comparing the situation to skills that have faded elsewhere in orthopaedics, such as freehand pedicle screw placement, manual knee cuts, and open shoulder stabilization.
Dr. Entezari described losing registration during a case after contact with a tracker and having to rely on his manual skills to complete the operation. Dr. Freehill echoed the concern that trainees might become proficient with the technology without ever learning the underlying skills of standard shoulder arthroplasty.
Conclusion: this was a balanced presentation that emphasized the potential role of robotics in transferring a preoperative plan to the patient in the operating room, without claiming that robotic shoulder arthroplasty has been associated with improved patient outcomes.
Which is better?
Tundra Swans
Lake Washington
Follow on twitter/X: https://x.com/RickMatsen
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133//
Saturday, June 27, 2026
How to choose a humeral component in shoulder arthroplasty: ASES Podcast 155
Here are some takeaways from that presentation coupled with some additional thoughts from the U.W..
(1) The evolution in humeral components is attributed to (a) mitigation of complications (periprosthetic fracture, stress shielding from high canal-filling ratios, early loosening from certain ingrowth surfaces), (b) matching competitor features, and (c) responding to market forces.
The Standard Stem
(2) A standard length stem uses the canal as an alignment guide. It can be stabilized in the canal by (a) a bony ingrowth surface, (b) cement, or (c) reaming to a tight diaphyseal fit, reaming that, in any case, removes cortical bone asymmetrically and unpredictably. [7]. As discussed later in this post, the standard stem can also be stabilized in the canal by impaction grafting. Much of the stress-shielding literature conflates two different things — stem length and canal fill — and they are not the same. Proximal stress shielding tracks with the filling ratio — how much of the canal the implant occupies and how much load it diverts away from proximal bone — far more than with the absolute length of the stem. [9,10]. A long stem that fills the canal shields bone; a long stem that sits loosely in the canal, transferring load through impacted cancellous bone, need not.
(3) Osteoporosis or altered humeral anatomy may drive the use of a longer stem for better fixation and durability.
The Short Stem
(4) The short stem depends on loading at the metaphyseal level — in bone of variable shape and quality — making both fixation and orientation challenging.
CT of the proximal humerus showing the thin metaphyseal cortex
The short stem may not have enough leverage to reduce loosening.
Canal filling can lead to stress shielding
And periprosthetic fracture
The absence of the control provided by a standard length stem
can lead to malpositioning of a short stem
(5) Stemless implants depend on fixation at the anatomic neck, where a thin cortical shell surrounds cancellous bone. Bone density at this level is of variable quality, especially in patients in the arthroplasty age range. While some surgeons claim “100% stemless,” that claim is not realistic for every patient having shoulder arthroplasty.
AP radiograph with the anatomic-neck fixation zone annotated
The big question about stemless is how a surgeon decides, in a given patient, whether good fixation can be achieved. Preoperative imaging may give some clues, but it comes down to the intraoperative findings. Some have advocated the “thumb” test;
a more reliable approach is to insert the trial nucleus and determine whether it is solid. When it is not solid because of “mushy” cancellous bone, the options are limited. In our experience, adding cancellous graft at the anatomic neck does not rescue fixation: the site is already failing in cancellous bone, and loose graft packed against soft cancellous bone has nothing rigid to lock against.
The ability to improve fixation with longer fins is likewise limited; upsizing the trial to gain purchase can risk the lateral cortex.
It is also worth noting that even when stemless fixation is achieved, adaptive bone changes are common; at short-to-mid term, however, they have not reliably correlated with loosening or worse function.[6,8]
So in the real world, the stemless-loving surgeon needs a Plan “B” when trial fixation is unstable. Conversion to a stemmed component is the most attractive option. One caveat: if the problem is mushy cancellous bone, fixation of a short stem can itself be challenging — a fatter stem versus bone graft, and the same question of how to assure durability in the desired position.
Avoiding Trouble with the Stemless
Be aware of common technical mistakes - assure complete head resection and avoid excessive varus or valgus.
Insufficient head resection
Varus cut
The free-hand cut is key.
Revision
(6) Two concerns arise when the humeral component needs revision. First, removal of a well-fixed implant can risk fracture of the tuberosity, bone loss, and shaft fracture, and may require a humeral osteotomy or window. Second, although some humeral implants are “convertible” (the humeral fixation system stays in place), retaining the implant only makes sense if it is well fixed at an appropriate height and acceptable version. One podcast participant reported that 25–40% of nominally convertible stems cannot actually be converted in practice. Taken together, the routine use of “convertible” implants seems unattractive.
Malrotation
Too low or too high position of the stem
Another approach
(7) Our experience is that secure, safely revisable humeral component fixation can be achieved with a smooth, standard-length stem set at a small filling ratio and fixed with impaction autografting — using bone from the resected humeral head that one of our fellows named “God’s Own Glue” (see "Procrustean Method") [2,3]. The graft is impacted between a smooth stem and the endosteal cortex, where it stabilizes the implant at the time of surgery, rather than relying on a bony ingrowth surface, cement, or a tight diaphyseal press fit. [4]. A low filling ratio protects the humerus from stress shielding. Because the stem is deliberately undersized, load is transferred through grafted cancellous bone rather than bypassed down a canal-filling implant — addressing stress shielding at the variable that drives it. [9,10].
In 48 ream-and-run and 78 total shoulder arthroplasties using a smooth, standard-length impaction-grafted stem, two-year radiographs showed adaptive changes that were generally minor and not associated with component shift or subsidence. Inserted this way, a smooth standard-length stem offers secure, bone-preserving fixation providing results can serve as a basis for comparison for other component designs and fixation methods. [5].
In a consecutive single-surgeon series of 458 anatomic total shoulder arthroplasties using this construct (mean follow-up 9.2 years; 114 shoulders beyond ten years), Simple Shoulder Test scores improved from 3.3 to 9.2 and were sustained — never declining by more than the MCID — past ten years. The overall revision rate was 2.6% (12 of 458). None of the revisions were performed for a humeral-component cause: there was no humeral component loosening, no periprosthetic fracture, and no stem-related failure.[11]
A low filling ratio protects the humerus from stress shielding
Two year followup
Six year followup
Impaction grafting with a low filling ratio avoids incomplete seating of the humeral stem (humerus captivus)
Impaction grafting a standard stem in revising a failed short stem
Impaction grafting a thin long stem in humeral deformity
Impaction grafting after fracture fixation
Impaction grafting a thin standard stem enables easy, safe removal followed by new implant insertion should revision be necessary.
Thus "revisability" does not require "convertibility"
Returning to the observation stated at the beginning of this post: humeral component evolution is attributed to (a) mitigation of complications, (b) matching competitor features, and (c) responding to market forces.
Impaction grafting of a thin, smooth stem is an approach for addressing periprosthetic fracture, stress shielding from high canal-filling ratios, and early loosening. It is not driven by market forces. Impaction grafting can be used with any standard length stem (anatomic or reverse) combined with freely available autograft from the patient’s own humeral head that would otherwise be discarded.
Should we be thinking straight?
Beauty in Simplicity
Tundra Swan
Union Bay Natural Area
Disclosure. The author has no financial relationship with any manufacturer of the orthopaedic devices discussed in this post.
Follow on twitter/X: https://x.com/RickMatsen
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133//
1. The ASES Podcast (American Shoulder and Elbow Surgeons), Episode 155: standard, short, and stemless humeral components and convertible designs. Available at: https://www.youtube.com/watch?v=q_s_g8oQgPg
2. Razfar N, Reeves JM, Langohr GDG, Willing R, Athwal GS, Johnson JA. Comparison of proximal humeral bone stresses between stemless, short stem, and standard stem length: a finite element analysis. J Shoulder Elbow Surg. 2016;25(7):1076–1083. doi:10.1016/j.jse.2015.11.011. PMID 26810016.
3. Reeves JM, Langohr GDG, Athwal GS, Johnson JA. The effect of stemless humeral component fixation feature design on bone stress and strain response: a finite element analysis. J Shoulder Elbow Surg. 2018;27(12):2232–2241. doi:10.1016/j.jse.2018.06.002. PMID 30104100.
4. Synnott S, Langohr GDG, Reeves JM, Johnson JA, Athwal GS. The effect of humeral implant thickness and canal fill on interface contact and bone stresses in the proximal humerus. JSES Int. 2021;5(5):881–888. doi:10.1016/j.jseint.2021.05.006.
5. Aibinder WR, Uddin F, Bicknell RT, Krupp R, Scheibel M, Athwal GS. Stress shielding following stemless anatomic total shoulder arthroplasty. Shoulder Elbow. 2023. doi:10.1177/17585732211058804. PMID 36895609.
6. Raiss P, Edwards TB, Deutsch A, Shah A, Bruckner T, Loew M, Boileau P, Walch G. Radiographic changes around humeral components in shoulder arthroplasty. J Bone Joint Surg Am. 2014;96(7):e54. doi:10.2106/JBJS.M.00378. PMID 24695931.
7. Denard PJ, Raiss P, Gobezie R, Edwards TB, Lederman E. Stress shielding of the humerus in press-fit anatomic shoulder arthroplasty: review and recommendations for evaluation. J Shoulder Elbow Surg. 2018;27(6):1139–1147. doi:10.1016/j.jse.2017.12.020. PMID 29422391.
8. Sheth MM, Kahsai EA, Yang J, Whitson AJ, Matsen FA III, Hsu JE. What is the clinical importance of radiographic changes around the humeral component in anatomic shoulder arthroplasty? A minimum 4-year follow-up study. J Shoulder Elbow Surg.2025;34(8):1877–1885. doi:10.1016/j.jse.2024.11.024. PMID 39800107.
9. Denard PJ, Hsu JE, Whitson A, Neradilek MB, Matsen FA III. Radiographic outcomes of impaction-grafted standard-length humeral components in total shoulder and ream-and-run arthroplasty: is stress shielding an issue? J Shoulder Elbow Surg.2019;28(11):2181–2190. doi:10.1016/j.jse.2019.03.016. PMID 31272887.
10. Kim SC, Park JH, Bukhary H, Yoo JC. Humeral stem with low filling ratio reduces stress shielding in primary reverse shoulder arthroplasty. Int Orthop. 2022;46(6):1341–1349. doi:10.1007/s00264-022-05383-4.
11. Lee M, Chebli C, Mounce D, Bertelsen A, Richardson M, Matsen FA III. Intramedullary reaming for press-fit fixation of a humeral component removes cortical bone asymmetrically. J Shoulder Elbow Surg. 2008;17(1):150–155. doi:10.1016/j.jse.2007.03.032. PMID 18029200.
12. Boorman RS, Hacker SA, Lippitt SB, Matsen FA III. A conservative broaching and impaction grafting technique for humeral component placement and fixation in shoulder arthroplasty: the Procrustean method. Tech Shoulder Elbow Surg.2001;2(3):166–175. doi:10.1097/00132589-200109000-00004.
13. Hacker SA, Boorman RS, Lippitt SB, Matsen FA III. Impaction grafting improves the fit of uncemented humeral arthroplasty. J Shoulder Elbow Surg. 2003;12(5):431–435. doi:10.1016/s1058-2746(03)00053-3. PMID 14564262.
14. Lucas RM, Hsu JE, Gee AO, Neradilek MB, Matsen FA III. Impaction autografting: bone-preserving, secure fixation of a standard humeral component. J Shoulder Elbow Surg. 2016;25(11):1787–1794. doi:10.1016/j.jse.2016.03.008. PMID 27262410.
15. Lee et al. Stress shielding effects of short stem alignment and bone density in reverse shoulder arthroplasty. J Orthop Res. 2026. doi:10.1002/jor.70140.
16. Vasiliadis AV, Giovanoulis V, Lepidas N, Bampis I, Servien E, Lustig S, Gunst S. Stress shielding in stemmed reverse shoulder arthroplasty: an updated review. SICOT J. 2024;10:37. doi:10.1051/sicotj/2024029.
17. Ritter D, Raiss P, Denard PJ, Werner BC, Müller PE, Woiczinski M, Wijdicks CA, Bachmaier S. Volumetric humeral canal fill ratio effects primary stability and cortical bone loading in short and standard stem reverse shoulder arthroplasty: a biomechanical and computational study. J Imaging. 2024;10(12):334. doi:10.3390/jimaging10120334.
18. Kramer M, Olach M, Zdravkovic V, Manser M, Raiss P, Jost B, Spross C. The effects of length and width of the stem on proximal humerus stress shielding in uncemented primary reverse total shoulder arthroplasty. Arch Orthop Trauma Surg. 2024. doi:10.1007/s00402-023-05129-w.
19. John PB, Nageswaran S. Mechanobiological evaluation of solid and multiple porous humeral stem architectures in reverse shoulder arthroplasty based on design and materials: a finite element study. Front Bioeng Biotechnol. 2026;13:1675726. doi:10.3389/fbioe.2025.1675726.
20. Takayama K, Ito H. Association between the canal filling ratio and bone resorption in trabecular metal stems in reverse total shoulder arthroplasty: a radiographic analysis using tomosynthesis. JSES Int. 2024;8(5):1077–1086. doi:10.1016/j.jseint.2024.05.010.
Wednesday, June 24, 2026
Are patients having shoulder arthroplasty doing better over the last decade, if so, why?
This post examines two questions of importance to the surgeons and patients interested in shoulder arthroplasty:
(1) Over the past decade, is there evidence that patient-reported outcomes (PROs) after shoulder arthroplasty have gotten better by a clinically significant amount — that is, by at least the minimal clinically important difference (MCID)?
(2) If they have, is that improvement associated with the surgeon, the technology, the shoulder, or the patient?
Based on the available evidence, the answers are: (1) yes, modestly; and (2) the gain is not attributable to new technology or to any change in the shoulders we treat. The largest lever on the outcome is the patient, while the decade’s trend itself most plausibly reflects accumulated surgical experience and care.
A word on the measure. Surrogates — such as radiographic component position, glenoid version, three-dimensional planning accuracy, and data on implant survivorship drawn from registries — are not the outcomes of primary interest. The important outcome is what the patients report. A technology that improves a surrogate but does not move a PRO beyond its MCID has not improved the outcome for the patient.
Now the detail.
1. Are patient-reported outcomes improving over time?
Anatomic TSA (aTSA). In a single-institution series of 1,899 arthroplasties (2008–2018), the 5-year ASES improvement rose by 1.65 points per year of index surgery on univariable analysis and by 2.20 points per year after adjustment for patient factors [1]. The per-year increment is small — well under one MCID — and reaches clinical significance only when accumulated across the decade. The authors could not attribute the gain to any single factor [1].
Reverse TSA (RSA). Some of the most relevant data come from the U.K. National Joint Registry (24,411 RSAs, 2013–2021): mean 6-month Oxford Shoulder Score (OSS) improvement rose from 15.84 to 20.29 — about half a point per year, roughly one MCID across the whole period [2].
In the Danish Shoulder Arthroplasty Registry (2,867 arthroplasties for osteoarthritis, 2006–2015), the adjusted WOOS score rose by about 10 points [3] — short of the WOOS MCID of 12.3 [16].
So.....both aTSA and RSA show measured, year-over-year PRO improvement. The increments are small per year and reach clinically meaningful magnitude only when accumulated across the decade — about one MCID for RSA across the period, and a larger accumulated difference for aTSA, whose adjusted per-year gain compounds over ten years. These are population- and cohort-level shifts: a patient operated late in the decade does modestly better, on average, than one operated early.
Survivorship bias. A patient-reported outcome plotted against time after surgery is available only for survivors: to contribute a 6-month OSS or a 5-year ASES, a patient must be alive, unrevised, and reachable. Those who die, are revised, or move their care to another surgeon before the minimum follow-up never enter the average — which is therefore selected for better outcomes.
Which way this biases the trend is not obvious. Perioperative mortality and early reoperation fell across this same decade [2], so fewer patients were removed by death in the later years; if anything the later cohorts are less selected, and across-year survivorship is unlikely to be what produced the rising trend. The selection that matters is not across calendar years but across follow-up length — the late-timepoint averages rest on progressively smaller, more selected subsets of the original cohorts.
Implant durability is vulnerable to the same problem in a more specific way. When revision is estimated with a standard Kaplan–Meier survival curve, a patient who dies is treated as still at risk of a revision that can no longer occur. This overestimates the cumulative revision rate, and it overestimates most where mortality is highest — the older patients who receive reverse replacements [15]. Comparisons of durability across groups with different mortality are confounded for the same reason: part of any apparent difference reflects who lived long enough to ever reach a revision, not how the implant performed. Revision rates in older, higher-mortality groups should accordingly be read as upper bounds rather than firm figures [15].
The usual adjustments for age, comorbidity, and surgeon volume operate only among the patients who actually reported a score; they cannot recover those missing because they died or never returned. So even where adjustment strengthens the trend [1], survivorship bias remains in the data. It matters most where follow-up completion is low and uneven, as it was in the National Joint Registry [2].
So two cautions in reading these data. The within-period findings — smoking [9], unemployment and education [10], resilience and mental health [11], and the absence of any effect of glenoid shape [5] or technology [4] — compare patients operated on in the same period, so the across-year attrition that inflates the temporal trend cannot account for them. The year-over-year trend is the vulnerable part: surgeon experience, technique, rehabilitation, and perioperative care all improved over the same years, so a dataset of this kind cannot apportion the gain among them; the measured improvement is best taken as the most that can be claimed, not a precise figure.
2. The surgeon, the technology, the shoulder, or the patient — what is the change associated with?
The surgeon. Surgeon volume and fellowship training have a well-documented effect on revision, reoperation, complications, and length of stay — and a weak-to-absent effect on the size of the PRO improvement patients realize. In the National Joint Registry, surgeons averaging at least 10.4 shoulder replacements a year had lower revision risk (the hazard of revision roughly doubled below that threshold) along with fewer reoperations, fewer serious adverse events, and shorter stays, but no patient-reported outcome was assessed [12]. Fellowship training shows the same pattern: shoulder-and-elbow and sports-medicine fellowship-trained surgeons have significantly lower complication rates at 90 days, 1 year, and 5 years, again with no PRO assessed [14]. The most direct test comes from the Australian registry: a surgeon’s two-year revision rate showed no clinically relevant correlation with PROMs, and all surgeons produced similar postoperative Oxford scores [13]. Revision rates and PROs measure two different facets of the result — surgeon factors move the first; patient factors move the second.
The technology. Advanced technology does not appear to be related to patient-reported outcomes. A systematic assessment of three-dimensional planning, patient-specific instrumentation, navigation, stemless humeral components, and augmented glenoid components found that no individual technology was associated with a statistically or clinically significant improvement in PROs [4]. A single-surgeon consecutive series of 389 anatomic arthroplasties reaches the same conclusion from the other side: implant choice and surgical technology do not predict outcome [5]. The durable result there came from fundamentals — subscapularis management, a conservative neck cut, a securely seated all-polyethylene glenoid, soft-tissue balancing — not new technology.
Cross-linked polyethylene is the one material change with an apparent benefit: in the Australian registry it had a lower anatomic revision rate than conventional polyethylene [6]. But the benefit is on durability — reduced wear-particle osteolysis and loosening — not on comfort and function; no study has shown that it improves PROs. And the advantage appears smaller in contemporary practice, where glenoid component design and fixation, not polyethylene type, emerge as the dominant predictor of revision [7].
The shoulder. Glenoid morphology — the usual proxy for a “worse” arthritic shoulder — does not predict outcome: in a 389-case series, B2 and B3 glenoids met or exceeded the outcomes of other glenoid types, and symptomatic glenoid loosening occurred in 1 of 389 shoulders [5]. Nor has the case mix drifted toward milder disease. In the Australian registry, the reverse diagnosis mix barely changed across the decade — osteoarthritis fell only slightly as a share of reverse procedures (48.2% to 44.7%) while cuff arthropathy rose by almost the same margin (36.2% to 39.7%) — if anything a drift toward classic cuff-deficient disease, the opposite of an “easier shoulders” explanation [8]. A shift in shoulder characteristics does not appear to be driving the PRO gain.
The patient. The data point to the patient — and to the judgment applied in selecting and preparing the patient. Three patient-side domains stand out. Modifiable risk: in an analysis of 14,465 arthroplasties, smoking was independently associated with increased surgical complications [9]; cessation, along with glycemic, nutritional, and opioid optimization, appears to improve the result. Socioeconomic context: in a nationwide cohort of 2,292 arthroplasties, after adjustment for age, sex, diagnosis, implant, and comorbidity, unemployment was associated with a WOOS deficit of roughly 14 to 19 points at one year and lower educational attainment with a further 5 to 8 points [10] — differences that exceed the MCID. Resilience and mental health: in 399 anatomic arthroplasties, greater resilience and better mental health were associated with better outcomes [11].
So among the candidate factors, the patient is the strongest predictor of the outcome a given arthroplasty achieves — selection, optimization, socioeconomic context, and resilience each move the PRO, several beyond the MCID. Whether the patient mix shifted across the decade in a way that would explain the time trend is a separate question, and these data do not show that it did.
Conclusion
The evidence answers both opening questions.
First, patient-reported outcomes after shoulder arthroplasty have improved over the past decade — modestly, and by an amount that reaches the MCID only when accumulated across the period, not within any single year.
Second, two different questions hide inside "what is the change associated with," and they have different answers.
(A) The decade-long trend is real but cannot be apportioned by the available data; it is most plausibly the product of accumulated surgeon experience, refinements in technique, rehabilitation, and perioperative care. The case mix did not drift toward easier shoulders [8], and no individual technology was associated with the improved outcomes over the decade [4].
(B) The other question — what determines the result for an individual patient — has a clearer answer. The available evidence has not shown a patient-reported benefit from any technology assessed, nor from any change in the shoulders we treat. Surgeon volume and training affect revision and complication rates, but not, as far as the data show, the patient-reported result. That leaves the patient: selection, optimization of modifiable risk, socioeconomic context, and resilience.
The greatest demonstrated effect on the outcome therefore lies in choosing and preparing the patient; the experience the surgeon brings to the procedure most likely matters too, though these data cannot show by how much.
Market forces are clearly driving orthopaedic companies to develop new implants and technologies that strengthen their competitive position, even though evidence that these have a major impact on patient outcomes is lacking.
What this analysis does offer, though, is actionable intelligence: the surest routes to improving outcomes for our patients are (1) thoughtful patient selection and optimization and (2) education that enables surgeons to improve their craft of shoulder arthroplasty.
Will the future be better?
Broad-billed Hummingbird
Tucson
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133//
References
[1] Mathew JI, Nicholson AD, Finocchiaro A, Okeke L, Dines DM, Dines JS, Taylor SA, Warren RF, Gulotta LV. Outcomes of shoulder arthroplasty by year of index procedure: are we getting better? J Shoulder Elbow Surg. 2022;31(2):245-251. doi:10.1016/j.jse.2021.08.024. PMID 34592407.
[2] O’Malley O, Davies A, Taghavi Azar Sharabiani M, Rangan A, Sabharwal S, Reilly P. Are we getting better over time? Clinical and patient-reported outcomes for reverse shoulder arthroplasty: a National Joint Registry cohort study. BMJ Open. 2025;15(9):e096084. doi:10.1136/bmjopen-2024-096084. PMID 40953878.
[3] Rasmussen JV, Amundsen A, Sørensen AKB, Klausen TW, Jakobsen J, Jensen SL, Olsen BS. Increased use of total shoulder arthroplasty for osteoarthritis and improved patient-reported outcome in Denmark, 2006-2015: a nationwide cohort study from the Danish Shoulder Arthroplasty Registry. Acta Orthop. 2019;90(5):489-494. doi:10.1080/17453674.2019.1633759. PMID 31240980.
[4] Schiffman CJ, Prabhakar P, Hsu JE, Shaffer ML, Miljacic L, Matsen FA 3rd. Assessing the value to the patient of new technologies in anatomic total shoulder arthroplasty. J Bone Joint Surg Am. 2021;103(9):761-770. doi:10.2106/JBJS.20.01853. PMID 33587515.
[5] Schiffman CJ, Chin, Whitson AJ, Hsu JE, Matsen FA 3rd. Anatomic shoulder arthroplasty: a single surgeon’s consecutive series of 458 patients (389 anatomic TSAs, mean 9.2-year follow-up). University of Washington SHORE Registry. [Manuscript under review.]
[6] Page RS, Alder-Price AC, Rainbird S, Graves SE, de Steiger RN, Peng Y, Holder C, Lorimer MF, Gill SD. Reduced revision rates in total shoulder arthroplasty with crosslinked polyethylene: results from the Australian Orthopaedic Association National Joint Replacement Registry. Clin Orthop Relat Res. 2022;480(10):1940-1949. doi:10.1097/CORR.0000000000002293. PMID 35901440.
[7] Gill DRJ, Corfield S, Harries D, Page RS. Modeling highly crosslinked polyethylene vs. non-highly crosslinked polyethylene glenoid revision rates for anatomic shoulder arthroplasty in osteoarthritis including differing polyethylene glenoid fixation designs. Semin Arthroplasty JSES. 2024;34(4):843-853. doi:10.1053/j.sart.2024.06.003.
[8] Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). Hip, Knee & Shoulder Arthroplasty: 2025 Annual Report. Adelaide: Australian Orthopaedic Association; 2025.
[9] Althoff AD, Reeves RA, Traven SA, Wilson JM, Woolf SK, Slone HS. Smoking is associated with increased surgical complications following total shoulder arthroplasty: an analysis of 14,465 patients. J Shoulder Elbow Surg. 2020;29(3):491-496. doi:10.1016/j.jse.2019.07.012. PMID 31519425.
[10] Jensen ML, Valsamis EM, Madrid AS, Olsen BS, Rasmussen JV. Association between socioeconomic status and patient-reported outcome at 1 year after shoulder arthroplasty for osteoarthritis or cuff-tear arthropathy: a nationwide cohort study of 2,292 arthroplasties. Acta Orthop. 2025;96:45-51. doi:10.2340/17453674.2024.42700.
[11] Levins JG, Dasari SP, Quinlan NJ, Whitson AJ, Matsen FA 3rd, Hsu JE. Anatomic shoulder arthroplasty: the correlation between patient resilience, mental health, and outcome. J Shoulder Elbow Surg. 2024;33(6S):S9-S15. doi:10.1016/j.jse.2024.03.008. PMID 38548096.
[12] Valsamis EM, Collins GS, Pinedo-Villanueva R, Whitehouse MR, Rangan A, Sayers A, Rees JL. Association between surgeon volume and patient outcomes after elective shoulder replacement surgery using data from the National Joint Registry and Hospital Episode Statistics for England: population based cohort study. BMJ. 2023;381:e075355. doi:10.1136/bmj-2023-075355. PMID 37343999.
[13] Hoskins W, Bingham R, Corfield S, Harries D, Harris IA, Vince KG. Do the revision rates of arthroplasty surgeons correlate with postoperative patient-reported outcome measure scores? A study from the Australian Orthopaedic Association National Joint Replacement Registry. Clin Orthop Relat Res. 2024;482(1):98-112. doi:10.1097/CORR.0000000000002737. PMID 37339166.
[14] Harkin WE, Saad Berreta R, Turkmani A, Williams T, Scanaliato JP, McCormick JR, Nicholson GP, Garrigues GE. How fellowship training affects complication rate after shoulder arthroplasty: a nationwide assessment. J Shoulder Elbow Surg. 2025;34(2):499-506. doi:10.1016/j.jse.2024.05.014. PMID 38944372.
[15] Lacny S, Wilson T, Clement F, Roberts DJ, Faris PD, Ghali WA, Marshall DA. Kaplan-Meier survival analysis overestimates the risk of revision arthroplasty: a meta-analysis. Clin Orthop Relat Res. 2015;473(11):3431-3442. doi:10.1007/s11999-015-4235-8. PMID 25804881.
[16] Nyring MRK, Olsen BS, Amundsen A, Rasmussen JV. Minimal clinically important differences (MCID) for the Western Ontario Osteoarthritis of the Shoulder Index (WOOS) and the Oxford Shoulder Score (OSS). Patient Relat Outcome Meas. 2021;12:299-306. doi:10.2147/PROM.S316920. PMID 34588833.
Subscribe to:
Posts (Atom)