Aims. Complex joint fractures of the lower extremity are often accompanied by soft-tissue swelling and are associated with prolonged hospitalization and soft-tissue complications. The aim of the study was to evaluate the effect of vascular impulse technology (VIT) on soft-tissue conditioning in comparison with conventional elevation. Methods. A total of 100 patients were included in this prospective, randomized, controlled monocentre study allocated to the three subgroups of dislocated ankle fracture (n = 40), pilon fracture (n = 20), and intra-articular calcaneal fracture (n = 40). Patients were randomized to the two study groups in a 1:1 ratio. The effectiveness of VIT (intervention) compared with elevation (control) was analyzed separately for the whole study population and for the three subgroups. The primary endpoint was the time from admission until operability (in days). Results. The mean length of time until operability was 8.2 days (SD 3.0) in the intervention group and 10.2 days (SD 3.7) in the control group across all three fractures groups combined (p = 0.004). An analysis of the subgroups revealed that a significant reduction in the time to operability was achieved in two of the three: with 8.6 days (SD 2.2) versus 10.6 days (SD 3.6) in ankle fractures (p = 0.043), 9.8 days (SD 4.1) versus 12.5 days (SD 5.1) in pilon fractures (p = 0.205), and 7.0 days (SD 2.6) versus 8.4 days (SD 1.5) in calcaneal fractures (p = 0.043). A lower length of stay (p = 0.007), a reduction in pain (p. preop. = 0.05; p. discharge. < 0.001) and need for narcotics (p. preop. = 0.064; p. postop. = 0.072), an increased reduction in swelling (p < 0.001), and a lower revision rate (p = 0.044) could also be seen, and a trend towards fewer complications (p = 0.216) became apparent. Conclusion. Compared with elevation, VIT results in a significant reduction in the time to achieve operability in complex joint fractures of the lower limb. Cite this article: Bone Joint J 2021;103-B(4):746–754

Aims. Both anatomical and reverse total shoulder arthroplasty (aTSA and rTSA) provide functional improvements. A reported benefit of aTSA is better range of motion (ROM). However, it is not clear which procedure provides better outcomes in patients with limited foward elevation (FE). The aim of this study was to compare the outcome of aTSA and rTSA in patients with glenohumeral osteoarthritis (OA), an intact rotator cuff, and limited FE. Methods. This was a retrospective review of a single institution’s prospectively collected shoulder arthroplasty database for TSAs undertaken between 2007 and 2020. A total of 344 aTSAs and 163 rTSAs, which were performed in patients with OA and an intact rotator cuff with a minimum follow-up of two years, were included. Using the definition of preoperative stiffness as passive FE ≤ 105°, three cohorts were matched 1:1 by age, sex, and follow-up: stiff aTSAs (85) to non-stiff aTSAs (85); stiff rTSAs (74) to non-stiff rTSAs (74); and stiff rTSAs (64) to stiff aTSAs (64). We the compared ROMs, outcome scores, and complication and revision rates. Results. Compared with non-stiff aTSAs, stiff aTSAs had poorer passive FE and active external rotation (ER), whereas there were no significant postoperative differences between stiff rTSAs and non-stiff rTSAs. There were no significant differences in preoperative function when comparing stiff aTSAs with stiff rTSAs. However, stiff rTSAs had significantly greater postoperative active and passive FE (p = 0.001 and 0.004, respectively), and active abduction (p = 0.001) compared with stiff aTSAs. The outcome scores were significantly more favourable in stiff rTSAs for the Shoulder Pain and Disability Index, Simple Shoulder Test, American Shoulder and Elbow Surgeons score, University of California, Los Angeles score, and the Constant score, compared with stiff aTSAs. When comparing the proportion of stiff aTSAs versus stiff rTSAs that exceeded the minimal clinically important difference and substantial clinical benefit, stiff rTSAs achieved both at greater rates for all measurements except active ER. The complication rate did not significantly differ between stiff aTSAs and stiff rTSAs, but there was a significantly higher rate of revision surgery in stiff aTSAs (p = 0.007). Conclusion. Postoperative overhead ROM, outcome scores, and rates of revision surgery favour the use of a rTSA rather than aTSA in patients with glenohumeral OA, an intact rotator cuff and limited FE, with similar rotational ROM in these two groups. Cite this article: Bone Joint J 2023;105-B(12):1303–1313

We studied the position and rotational changes associated with elevation of the glenohumeral joint, using a three-dimensional magnetic-field tracking system on nine fresh cadaveric shoulders. The plane of maximal arm elevation was shown to occur 23 degrees anterior to the plane of the scapula. Elevation in any plane anterior to the scapula required external humeral rotation, and maximal elevation was associated with approximately 35 degrees of external humeral rotation. Conversely, internal rotation was necessary for increased elevation posterior to the plane of the scapula. The observed effects of this rotation were to clear the humeral tuberosity from abutting beneath the acromion and to relax the inferior capsular ligamentous constraints. Measurement of the obligatory humeral rotation required for maximal elevation helps to explain the relationship of the limited elevation seen in adhesive capsulitis and after operations which limit external rotation

Aims. The aim of this study was to compare the pain caused by the application of a tourniquet after exsanguination of the upper limb with that occurring after simple elevation. . Patients and Methods. We used 26 healthy volunteers (52 arms), each of whom acted as their own matched control. The primary outcome measure was the total pain experienced by each volunteer while the tourniquet was inflated for 20 minutes. This was calculated as the area under the pain curve for each individual subject. Secondary outcomes were pain at each time point; the total pain experienced during the recovery phase; the ability to tolerate the tourniquet and the time for full recovery after deflation of the tourniquet. . Results. There was a significant difference in the area under the pain curves in favour of exsanguination (mean difference 8.4; 95% confidence interval (CI) 3.0 to 13.7, p = 0.004). There was no difference between the dominant and non-dominant arms (mean difference -0.2; 95% CI -23.2 to 22.8, p = 0.99). The area under both recovery curves were similar (mean difference 0.7; 95% CI -6.0 to 4.6, p = 0.78). There was no statistical difference in recovery time, the actual mean difference being 30 seconds longer in the elevation group (p = 0.06). Conclusion. Many orthopaedic and plastic surgery procedures are done under local anaesthetic or regional block where a bloodless field and a motionless patient are essential. Optimising patient comfort during surgery with the tourniquet inflated is thus a priority. This study is useful in that it compares two common methods of preparation of the upper limb prior to tourniquet inflation and which have not previously been compared in this context. Following on the results of this study, we can confidently conclude that exsanguinating the upper limb before inflating a tourniquet is more comfortable than simply elevating the arm for patients undergoing a procedure under local or regional block, both during the procedure and in the recovery phase. . Take home message: Exsanguination rather than elevation is recommended in order to minimise patient discomfort and optimise the surgical field. Cite this article: Bone Joint J 2016;98-B:519–25

1. One hundred cases of congenital elevation of the scapula are reviewed; thirty-four were treated by operation. 2. A method of grading of appearance is suggested to assist in selecting cases for operation. 3. In very mild cases operation does not confer worthwhile benefit. 4. Very severely affected patients cannot be made to look normal. 5. The appearance often deteriorates during a period of rapid growth of the neck, making continued observation and reassessment essential. 6. The primary aim of treatment is to improve appearance; improvement in function is uncertain. 7. Excision of the supero-medial part of the scapula, and of the omo-vertebral bone when present, is recommended in most cases requiring operation. It is safe and simple, with little risk of loss of function and with simple after-treatment

We have treated seven children with relapsed infantile Blount’s disease by elevation of the hemiplateau using the Ilizarov frame. Three boys and four girls with a mean age of 10.5 years were reviewed at a mean of 29 months after surgery. All had improved considerably and were pleased with the results. The improvements in radiological measurements were statistically significant (p < 0.001). Three-dimensional CT reconstruction was useful for planning surgery. There were no major complications. The advantages of this technique are that in addition to elevation of the hemiplateau, rotational deformities and limb-length discrepancies may be addressed

Of forty-six patients with congenital elevation of the scapula, diastematomyelia was confirmed in four and was probably present in five others. Possible reasons for the previously unsuspected association are presented

Ulnar nerve function, during and after open reduction and internal fixation of fractures of the distal humerus with subperiosteal elevation of the nerve, was assessed by intra-operative neurophysiological monitoring. Intermittent recording of the compound muscle action potentials was taken from the hypothenar muscles in 18 neurologically asymptomatic patients. The mean amplitude of the compound muscle action potential after surgery was 98.1% (. sd. 17.6; −37% to +25%). The amplitude improved in six patients following surgery. Despite unremarkable recordings one patient had progressive paresis. Motor impairment is unlikely if the compound muscle action potential is continuously preserved and not reduced by more than 40% at the end of surgery. Temporary decreases in amplitude by up to 70% were tolerated without clinical consequences. However, repeated clinical examination is obligatory to recognise and treat early post-operative palsy

An innovative Kirschner (K-) wire point was developed and compared in fresh pig femora in terms of drilling efficiency and temperature elevation with the trochar and diamond points currently used in clinical practice. The tips of thermal couples were machined to the defined geometry and the temperature measured during drilling. Using the same drill speed (rev/min) and feed rate, the new K-wire point produced the lowest thrust force and torque as measured by a Kistler dynamometer. Drill point temperatures were highest with the trochar geometry (129 ± 6°C), followed by the diamond (98 ± 7°C). The lowest temperatures were recorded with the Medin K-wire (66 ± 2°C). On repeated drilling it could be used for up to 30 holes before reaching the less satisfactory drill performance of the diamond tip. The new K-wire provides a better alternative as it requires less effort for insertion, generates less heat and may be re-used

We report 12 consecutive cases of vertical scapular osteotomy to correct Sprengel’s deformity, performed during a 16-year period, with a mean follow-up of 10.4 years. The mean increase in abduction of the shoulder was 53°. The cosmetic appearance improved by a mean of 1.5 levels on the Cavendish scale. Neither function nor cosmesis deteriorated with time. We recommend the procedure for correction of moderate deformities with a functional deficit.

Aims. Surgeons commonly resect additional distal femur during primary total knee arthroplasty (TKA) to correct a flexion contracture, which leads to femoral joint line elevation. There is a paucity of data describing the effect of joint line elevation on mid-flexion stability and knee kinematics. Thus, the goal of this study was to quantify the effect of joint line elevation on mid-flexion laxity. Methods. Six computational knee models with cadaver-specific capsular and collateral ligament properties were implanted with a posterior-stabilized (PS) TKA. A 10° flexion contracture was created in each model to simulate a capsular contracture. Distal femoral resections of + 2 mm and + 4 mm were then simulated for each knee. The knee models were then extended under a standard moment. Subsequently, varus and valgus moments of 10 Nm were applied as the knee was flexed from 0° to 90° at baseline and repeated after each of the two distal resections. Coronal laxity (the sum of varus and valgus angulation with respective maximum moments) was measured throughout flexion. Results. With + 2 mm resection at 30° and 45° of flexion, mean coronal laxity increased by a mean of 3.1° (SD 0.18°) (p < 0.001) and 2.7° (SD 0.30°) (p < 0.001), respectively. With + 4 mm resection at 30° and 45° of flexion, mean coronal laxity increased by 6.5° (SD 0.56°) (p < 0.001) and 5.5° (SD 0.72°) (p < 0.001), respectively. Maximum increased coronal laxity for a + 4 mm resection occurred at a mean 15.7° (11° to 33°) of flexion with a mean increase of 7.8° (SD 0.2°) from baseline. Conclusion. With joint line elevation in primary PS TKA, coronal laxity peaks early (about 16°) with a maximum laxity of 8°. Surgeons should restore the joint line if possible; however, if joint line elevation is necessary, we recommend assessment of coronal laxity at 15° to 30° of knee flexion to assess for mid-flexion instability. Further in vivo studies are warranted to understand if this mid-flexion coronal laxity has negative clinical implications. Cite this article: Bone Joint J 2021;103-B(6 Supple A):87–93

Aims. The aim of this study was to establish consensus statements on medial patellofemoral ligament (MPFL) reconstruction, anteromedialization tibial tubercle osteotomy, trochleoplasty, and rehabilitation and return to sporting activity in patients with patellar instability, using the modified Delphi process. Methods. This was the second part of a study dealing with these aspects of management in these patients. As in part I, a total of 60 surgeons from 11 countries contributed to the development of consensus statements based on their expertise in this area. They were assigned to one of seven working groups defined by subtopics of interest. Consensus was defined as achieving between 80% and 89% agreement, strong consensus was defined as between 90% and 99% agreement, and 100% agreement was considered unanimous. Results. Of 41 questions and statements on patellar instability, none achieved unanimous consensus, 19 achieved strong consensus, 15 achieved consensus, and seven did not achieve consensus. Conclusion. Most statements reached some degree of consensus, without any achieving unanimous consensus. There was no consensus on the use of anchors in MPFL reconstruction, and the order of fixation of the graft (patella first versus femur first). There was also no consensus on the indications for trochleoplasty or its effect on the viability of the cartilage after elevation of the osteochondral flap. There was also no consensus on postoperative immobilization or weightbearing, or whether paediatric patients should avoid an early return to sport. Cite this article: Bone Joint J 2023;105-B(12):1265–1270

Aims. The scapula is a rare site for a primary bone tumour. Only a small number of series have studied patient outcomes after treatment. Previous studies have shown a high rate of recurrence, with functional outcomes determined by the preservation of the glenohumeral joint and deltoid. The purpose of the current study was to report the outcome of patients who had undergone tumour resection that included the scapula. Methods. We reviewed 61 patients (37 male, 24 female; mean age 42 years (SD 19)) who had undergone resection of the scapula. The most common resection was type 2 (n = 34) according to the Tikhoff-Linberg classification, or type S1A (n = 35) on the Enneking classification. Results. The ten-year disease-specific survival was 76%. High tumour grade (hazard ratio (HR) 4.27; p = 0.016) and a total resection of the scapula (HR 3.84; p = 0.015) were associated with worse survival. The ten-year metastasis-free and local recurrence-free survivals were 82% and 86%, respectively. Total scapular resection (HR 6.29; p = 0.004) was associated with metastatic disease and positive margins were associated with local recurrence (HR 12.86; p = 0.001). At final follow-up, the mean shoulder forward elevation and external rotation were 79° (SD 62°) and 27° (SD 25°), respectively. The most recent functional outcomes evaluated included the mean Musculoskeletal Tumor Society Score (76% (SD 17%)), the American Shoulder and Elbow Score (73% (SD 20%)), and the Simple Shoulder Test (7 (SD 3)). Preservation of the glenoid (p = 0.001) and scapular spine (p < 0.001) improved clinical outcomes; interestingly, preservation of the scapular spine without the glenoid improved outcomes (p < 0.001) compared to preservation of the glenoid alone (p = 0.05). Conclusion. Resection of the scapula is a major undertaking with an oncological outcome related to tumour grade, and a functional outcome associated with the status of the scapular spine and glenoid. Positive resection margins are associated with local recurrence. Cite this article: Bone Joint J 2023;105-B(12):1314–1320

Aims. Head-taper corrosion is a cause of failure in total hip arthroplasty (THA). Recent reports have described an increasing number of V40 taper failures with adverse local tissue reaction (ALTR). However, the real incidence of V40 taper damage and its cause remain unknown. The aim of this study was to evaluate the long-term incidence of ALTR in a consecutive series of THAs using a V40 taper and identify potentially related factors. Methods. Between January 2006 and June 2007, a total of 121 patients underwent THA using either an uncemented (Accolade I, made of Ti. 12. Mo. 6. Zr. 2. Fe; Stryker, USA) or a cemented (ABG II, made of cobalt-chrome-molybdenum (CoCrMo); Stryker) femoral component, both with a V40 taper (Stryker). Uncemented acetabular components (Trident; Stryker) with crosslinked polyethylene liners and CoCr femoral heads of 36 mm diameter were used in all patients. At a mean folllow-up of 10.8 years (SD 1.1), 94 patients (79%) were eligible for follow-up (six patients had already undergone a revision, 15 had died, and six were lost to follow-up). A total of 85 THAs in 80 patients (mean age 61 years (24 to 75); 47 (56%) were female) underwent clinical and radiological evaluation, including the measurement of whole blood levels of cobalt and chrome. Metal artifact reduction sequence MRI scans of the hip were performed in 71 patients. Results. A total of 20 ALTRs were identified on MRI, with an incidence of 26%. Patients with ALTR had significantly higher median Co levels compared with those without ALTR (2.96 μg/l (interquartile range (IQR) 1.35 to 4.98) vs 1.44 μg/l (IQR 0.79 to 2.5); p = 0.019). Radiological evidence of osteolysis was also significantly associated with ALTR (p = 0.014). Median Cr levels were not significantly higher in those with ALTR compared with those without one (0.97 μg/l (IQR 0.72 to 1.9) v 0.67 μg/l (IQR 0.5 to 1.19; p = 0.080). BMI, sex, age, type of femoral component, head length, the inclination of the acetabular component, and heterotopic ossification formation showed no significant relationship with ALTR. Conclusion. Due to the high incidence of local ALTR in our cohort after more than ten years postoperatively, we recommend regular follow-up investigation even in asymptomatic patients with V40 taper and metal heads. As cobalt levels correlate with ALTR occurrence, routine metal ion screening and consecutive MRI investigation upon elevation could be discussed. Cite this article: Bone Joint J 2022;104-B(7):852–858

Aims. The use and variety of stemless humeral components in anatomical total shoulder arthroplasty (TSA) have proliferated since their advent in 2004. Early outcomes are reassuring but independent mid-term results are scarce. This independent study reports a consecutive series of 143 Eclipse stemless shoulder prostheses with a minimum five-year (5 to 10) follow-up. Methods. Outcomes of 143 procedures undertaken for all indications in 131 patients were reviewed, with subset analysis of those for osteoarthritis (OA) (n = 99). The primary outcome was the Oxford Shoulder Score (OSS) at a minimum of five years. Secondary outcomes were ranges of motion and radiological analysis of humeral radiolucency, rotator cuff failure, and glenoid loosening. Results. Mean OSS at mean follow-up of 6.67 years (5.0 to 10.74) was 40.12 (9 to 48), with no statistically significant difference between those implanted for a non-OA indication and those for OA (p = 0.056) or time-dependent deterioration between two years and five years (p = 0.206). Ranges of motion significantly improved compared with preoperative findings and were maintained between two and five years with a mean external rotation of 38° (SD 18.1, 0 to 100) and forward elevation of 152° (SD 29.9, 90 to 180). Of those components with radiographs suitable for analysis (n = 83), 23 (28%) were found to have a least one humeral radiolucent line, which were predominantly incomplete, less than 2 mm, and in a single anatomical zone. No humeral components were loose. A radiolucent line was present around 22 (15%) of glenoid components, and 15 (10%) of components had failed. Rotator cuff failure was found in 21 (15%) components. The mean time to either glenoid or rotator cuff failure was greater than three years following implantation. Survivorship was 96.4% (95% CI 91.6 to 98.5, number at risk 128) at five years, and 94.3% (95% CI 88.2 to 97.3, number at risk 76) at seven years, both of which compare favourably with best results taken from available registries. Conclusion. Functional and radiological outcomes of the Eclipse stemless TSA are excellent, with no loose humeral components at minimum five-year follow-up. The presence of radiolucent lines is of interest and requires long-term observation but does not impact on the clinical results. Of the eight revisions required, this was predominantly for glenoid and rotator cuff failure. Cite this article: Bone Joint J 2022;104-B(1):83–90

Aims. This study compared patients who underwent arthroscopic repair of large to massive rotator cuff tears (LMRCTs) with isolated incomplete repair of the tear and patients with incomplete repair with biceps tendon augmentation. We aimed to evaluate the additional benefit on clinical outcomes and the capacity to lower the re-tear rate. Methods. We retrospectively reviewed 1,115 patients who underwent arthroscopic rotator cuff repair for full-thickness tears between October 2011 and May 2019. From this series, we identified 77 patients (28 male, 49 female) with a mean age of 64.1 years (50 to 80). Patients were classified into groups A (n = 47 incomplete) and B (n = 30 with biceps augmentation) according to the nature of their reconstruction. Clinical scores were checked preoperatively and at six months, one year, and two years postoperatively. In preoperative MRI, we measured the tear size, the degree of fatty infiltration, and muscle volume ratio of the supraspinatus. In postoperative MRI, the integrity of the repaired rotator cuff tendon was assessed using the Sugaya classification. Tendon thickness at the footprint was evaluated on T2-weighted oblique coronal view. Results. There were no significant differences in the initial preoperative demographic characteristics. In both groups, there were significant improvements in postoperative clinical scores (p < 0.001). However, most clinical outcomes, including range of motion measurements (forward elevation, external rotation, internal rotation, and abduction), showed no differences between the pre- and postoperative values. Comparing the postoperative outcomes of both groups, no further improvement from biceps augmentation was found. Group B, although not reaching statistical significance, had more re-tears than group A (30% vs 15%; p = 0.117). Conclusion. In LMRCTs, biceps augmentation provided no significant improvement of an incomplete repair. Therefore, biceps augmentation is not recommended in the treatment of LMRCTs. Cite this article: Bone Joint J 2022;104-B(11):1234–1241

Aims. Existing literature indicates that inferiorly inclined glenoid baseplates following reverse total shoulder arthroplasty (RSA) produce better outcomes compared to superiorly inclined baseplates. We aim to compare clinical outcomes for RSAs with superiorly and neutrally/inferiorly inclined lateralized glenospheres. Methods. We retrospectively reviewed 154 consecutive patients undergoing RSA between July 2015 and July 2017 by one single-fellowship trained surgeon (AJ). Two raters (KAM and MVS) independently measured glenoid inclination in preoperative and minimum two year follow-up radiographs (anteroposterior/Grashey) using the RSA angle. Inclination was then compared to patient-reported outcomes, range of motion (ROM), and independently assessed degree of scapular notching and staging of heterotopic ossification at two year follow-up. Results. Median postoperative inclination for each group was found to be -3.6° (interquartile range (IQR) -2.1 to -6.9) and 6.0° (3.2° to 10.1°) for the neutrally/inferiorly and superiorly inclined cohorts, respectively. Preoperative inclination was highly associated with postoperative inclination (p = 0.004). When comparing superiorly and neutrally/inferiorly inclined glenospheres, there were no differences in heterotopic ossification (p = 0.606), scapular notching (p = 0.367), American Shoulder and Elbow Surgeons score (p = 0.419), Single Assessment Numeric Evaluation (p = 0.417), Visual Analogue Scale (VAS) pain score (p = 0.290), forward elevation (p = 0.161), external rotation (p = 0.537), or internal rotation (p = 0.656). Conclusion. Compared to neutral and inferior inclination, up to 6° ± 3° of superior glenoid baseplate inclination on a lateralized RSA design produces no differences in postoperative ROM or patient-reported outcomes, and produces similar levels of scapular notching and heterotopic ossification. Additionally, the degree of preoperative inclination represents an important factor in surgical decision-making as it is strongly associated with postoperative inclination. It is important to note that the findings of this study are only reflective of lateralized RSA prostheses. Cite this article: Bone Joint J 2021;103-B(2):360–365

We reviewed the outcome of 28 patients who had been treated using the Aequalis fracture prosthesis for an acute fracture of the proximal humerus at a mean follow-up of 39.3 months (24 to 63). The mean age of the patients at the time of the fracture was 66.3 years (38 to 80). The mean Constant score was 68.2 (37 to 84) for the operated shoulder, which represented 89.5% of the mean score for the uninjured side (p < 0.001). The quality of the reconstruction as shown on the immediate post-operative radiographs was categorised into three types, anatomical, acceptable, and unacceptable, depending on the position of the tuberosities relative to the prosthetic head and the humeral shaft. Anatomical reconstruction was associated with a higher mean Constant score as well as higher mean values of anterior forward elevation, abduction and external rotation than the other types, but the differences were not statistically significant (p > 0.231). A total of 18 patients had active anterior elevation ≥150°. Their mean active abduction and external rotation were 163.6° and 31.3°, respectively. In seven of the 28 patients, the mean active anterior elevation, abduction and external rotation were 130.7°, 129.2° and 22.8°, respectively. In all, 12 patients were very satisfied with the results, 12 were satisfied, two were dissatisfied and two were disappointed; 26 reported no or only mild pain while only two had moderate pain. In five patients proximal migration of the humeral head was shown on the anteroposterior radiographs of the shoulder. No evidence of loosening was found in any prosthesis

Aims. The aim of this study was to report the outcomes of different treatment options for glenoid loosening following reverse shoulder arthroplasty (RSA) at a minimum follow-up of two years. Patients and Methods. We retrospectively studied the records of 79 patients (19 men, 60 women; 84 shoulders) aged 70.4 years (21 to 87) treated for aseptic loosening of the glenosphere following RSA. Clinical evaluation included pre- and post-treatment active anterior elevation (AAE), external rotation, and Constant score. Results. From the original cohort, 29 shoulders (35%) were treated conservatively, 27 shoulders (32%) were revised by revision of the glenosphere, and 28 shoulders (33%) were converted to hemiarthroplasty. At last follow-up, conservative treatment and glenoid revision significantly improved AAE, total Constant score, and pain, while hemiarthroplasty did not improve range of movement or clinical scores. Multivariable analysis confirmed that conservative treatment and glenoid revision achieved similar improvements in pain (glenoid revision vs conservative, beta 0.44; p = 0.834) but that outcomes were significantly worse following hemiarthroplasty (beta -5.00; p = 0.029). Conclusion. When possible, glenoid loosening after RSA should first be treated conservatively, then by glenosphere revision if necessary, and last by salvage hemiarthroplasty. Cite this article: Bone Joint J 2019;101-B:461–469