Session 16: Shoulder and Elbow

Zaid Issa, Stig Brroson, Jeppe Vejlgaard Rasmussen
Zaid Issa and Stig Brorson: CEBO (Centre of Evidence-Based Orthopaedics), Shoulder and Elbow Unit, Department of Orthopaedic Surgery, Zealand University Hospital, University of Copenhagen, Denmark; Jeppe Vejlgaard Rasmussen: Shoulder and elbow unit, Department of Orthopaedic Surgery, Herlev, Gentofte Hospital, University of Copenhagen, Denmark.

Background: The results after stemless total shoulder arthroplasty (TSA) are promising, but only few and small studies have compared the results with that of stemmed TSA.
Aim: The aim was to compare the patient-reported outcome after stemless and stemmed TSA for primary osteoarthritis and to compare different stemless arthroplasty systems.
Materials and Methods: We included all anatomical TSAs reported to the Danish shoulder arthroplasty registry from 2014 to 2018. The Western Ontario Osteoarthritis of the Shoulder (WOOS) index at one year was used as a patient-reported outcome. The raw score was converted to percentages of the maximum score. General linear models were used to analyse the difference in mean WOOS index between stemless and stemmed TSA and between stemless arthroplasty systems.
Results: 1197 stemmed and 253 stemless TSAs were included. The mean WOOS index was 82 (SD=21) for stemmed TSA and 86 (SD=19) for stemless TSA. The stemless TSA had a statistically significant better score compared with stemmed TSA in the univariate model (4, CI: 0-7). The difference remained statistically significant in the multivariate model (6, CI 1-10) but did not exceed the minimal clinically important difference (MCID) which is estimated to ten. There was no difference in the mean WOOS index between the Nano and the Eclipse system (difference 3, CI -10 to 5).
Interpretation / Conclusion: We found a statistically significant better WOOS index score of stemless TSA compared with stemmed TSA, but the difference did not exceed the MCID. The stemmed and stemless TSA can be used equivalently in the treatment of patients with primary glenohumeral osteoarthritis.

Adam Witten, Karen Mikkelsen, Thomas Wagenblast Mayntzhusen, Mikkel Bek Clausen, Kristian Thorborg, Per Hölmich, Kristoffer Weisskirchner Barfod
Sports Orthopedic Research Center – Copenhagen, Department of Orthopedic Surgery, Copenhagen University Hospital Hvidovre, Denmark; Department of Midwifery, Physiotherapy, Occupational Therapy and Psychomotor Therapy, Faculty of Health, University College Copenhagen

Background: There is no universally appraised definition of the entity known as subacromial impingement syndrome (SIS). This makes it difficult for clinicians to interpret scientific results, and to communicate with each other and patients.
Aim: To map the literature concerning the terminology and the diagnostic criteria used in clinical studies investigating SIS.
Materials and Methods: Pubmed, Embase, CINAHL and SPORTDiscus were searched from inception to June 2020 using known terms for SIS. New terms, that were discovered, were fed back into the search string. Peer-reviewed clinical studies investigating SIS were eligible for inclusion. Studies containing secondary analyses of a previously published study, reviews, pilot studies and studies with less than ten participants were excluded. Two reviewers independently screened titles and abstracts, three reviewers independently applied in- and exclusion criteria to full-text versions of the articles, and one reviewer extracted data. Disagreement between the reviewers was resolved by dialogue.
Results: 11.056 records were identified. 911 were retrieved for full-text screening. 535 were included. In total, 20 different terms for SIS were identified. The diagnostic criteria were generally based on pain provocative shoulder tests. Over 100 different test combinations were identified. The most commonly used tests were Hawkin’s, Neer’s and Painful Arc (53%, 51% and 31% of the studies, respectively). 58% of the studies reported use of imaging. 30% of the studies used a combination of clinical tests and imaging. 9% of the studies specified that they included patients with full-thickness supraspinatus tears and 46% specified that they did not.
Interpretation / Conclusion: There is a worrying lack of consensus regarding terminology and diagnostic criteria for SIS. The variation in diagnostic criteria is so extensive that many studies are hardly comparable to each other with overlapping inclusion criteria and exclusion criteria. This calls for careful consideration when interpreting the results of studies investigating SIS and when comparing studies. It also highlights the need for a consensus regarding terminology and diagnostic criteria for SIS.

Kristine Bramsen Haugaard, Klaus Bak, Per Hölmich, Kristoffer Seem, Kristoffer Weisskirchner Barfod
Sports Orthopedic Research Center - Copenhagen (SORC-C), Department of Orthopedic Surgery, Copenhagen University Hospital Hvidovre, Denmark; Adeas Hospitals, Skodsborg and Parken, Denmark

Background: Acromioclavicular (AC) joint dislocations are common injuries accounting for 9-12% of all injuries to the shoulder girdle. The frequency is widely reported in the literature; however only limited research is available that describes the incidence and basic epidemiological features of the injury in a general urban population.
Aim: To investigate the incidence and epidemiology of acute acromioclavicular joint dislocations in the capital region of Denmark.
Materials and Methods: All patients with acute AC joint dislocation admitted to the Emergency Departments at three University Hospitals serving a population of 549,225 residents in the Capital region of Denmark were prospectively registered from 1 January to 31 December 2019. Patients with trauma to the shoulder, pain from the AC joint and increased coracoclavicular distance on radiographs were included and classified according to Rockwood’s classification. Rockwood type I injuries were excluded as they represent a sprain to the joint rather than a true dislocation. Data on age, sex, affected side and mechanism of injury were registered.
Results: 106 patients, male:female ratio 8.6:1, were included. The overall incidence of AC joint dislocations was 19.3 per 100,000 person years at risk (PYRS); 34.9 per 100,000 PYRS for males and 4.0 per 100,000 PYRS for females. The age distribution was bimodal peaking at the ages of 20-24 (39.8 per 100,000 PYRS) and 55-59 (43.6 per 100,000 PYRS). Rockwood type III was the most common type of AC joint dislocation accounting for 55.7% of the injuries. Type II and V accounted for 40.6% and 3.8% respectively. There were no type IV or VI dislocations. The most common mechanism of injury was sports accounting for 80/106 (75.5%) with cycling accounting for almost half of all injuries with 51/106 (48.1%).
Interpretation / Conclusion: The incidence of AC joint dislocations was 19.3 per 100,000 PYRS. Rockwood type III was the most common type of injury accounting for 55.7% of the injuries. Young and middle-aged males were at highest risk. 75.5% of the injuries occurred during sports, most frequently during cycling.

Nicolai Sandau, Peter Buxbom, Asbjørn Hróbjartsson, Ian A Harris, Stig Brorson
Centre for Evidence-Based Orthopedics, Department of Orthopedic Surgery, Zealand University Hospital Køge; Centre for Evidence-Based Medicine Odense (CEBMO) and Cochrane Denmark, Department of Clinical Research, University of Southern Denmark; Whitlam Orthopaedic Research Centre, Ingham Institute for Applied Medical Research, South Western Sydney Clinical School, University of New South Wales (UNSW Sydney), Liverpool, NSW, 2170, Australia

Background: Proximal humeral fractures are the third most common non-vertebral osteoporotic fractures among the elderly. Both operative and non-operative treatment options exist. Many systematic reviews have conducted meta-analyses to summarize the current evidence, but the conclusions have been conflicting. Conflicting conclusions may be caused by methodological flaws that introduce bias.
Aim: We aim to study the relation between methodological quality and conclusions of meta-analyses comparing operative with non-operative treatments for PHFs.
Materials and Methods: We conducted a systematic search of EMBASE, PubMed, The Cochrane Library and Web of Science for meta- analyses comparing non-operative with operative treatment for PHFs. The methodological quality of the included meta-analyses was assessed using AMSTAR2. The conclusions were scored for three outcome domains (functional outcome, quality of life, and harm) on a validated scale ranging from 1 to 6, with 1 defined as conclusions highly favouring non-operative treatment and 6 highly favouring operative treatment. Lastly, we analysed the association between methodological quality and conflicting conclusions.
Results: We included 21 systematic reviews: 19 meta-analyses and 2 network meta- analyses. Most (n = 18, 95%) of the meta-analyses were rated as having ‘critically low’ quality, while the remaining one (5%) was rated as having ‘low’ quality. The most under-reported AMSTAR2 items were related to protocol, source of funding for included studies, list of excluded studies and appropriate statistical methods when combining results from non-randomized studies of intervention. The conclusions were conflicting for all three outcome domains, even for meta-analyses reporting similar PICO question. Due to the consistently low quality, it was not possible to determine an association between methodological quality and conflicting conclusions.
Interpretation / Conclusion: It was not possible to determine an association between methodological quality and conflicting conclusions. Efforts are needed to improve the quality of future meta-analyses comparing operative with non-operative treatments for PHFs.

Clausen Mikkel Bek, Merrild Mikas, Pedersen Mads, Holm Kika, Andersen Lars, Zebis Mette, Jakobsen Thomas, Thorborg Kristian
Department of Midwifery, Physiotherapy, Occupational Therapy and Psychomotor Therapy, Faculty of Health, University College Copenhagen, Copenhagen, Denmark; National Research Centre for the Working Environment, Copenhagen, Denmark; Section for Orthopaedic and Sports Rehabilitation (SOS-R), Health Centre Nørrebro, City of Copenhagen, Copenhagen, Denmark; Sports Orthopedic Research Center – Copenhagen (SORC-C), Department of Orthopedic Surgery, Copenhagen University Hospital, Amager-Hvidovre, Denmark

Background: According to evidence-based guidelines for treatment of subacromial impingement (SIS), non-operative care with three months of exercise therapy is first line of treatment, but guideline adherence is unknown.
Aim: We investigated to what degree current care complies with clinical guidelines and to what extent successful outcomes are achieved.
Materials and Methods: We invited all 4521 patients diagnosed with SIS at any Danish hospital during a 3-months period to participate in this nation-wide retrospective population-based cohort study. The questionnaire used to obtain patient-reported information on content of care was based on the Danish National Clinical Guidelines for treatment of SIS and referral guidelines. Nine members of the working group responsible for the National Clinical Guidelines, including three orthopedic surgeons, commented on the questionnaire. We developed a revised version based on systematic condensation of all comments. Participants also reported patient acceptable symptom-state. Invitations were sent to eligible patients 3.5 months after diagnosis at the hospital.
Results: In total, 3306 eligible patients completed the questionnaire at median 16.7 weeks after diagnosis at a hospital. In total, 45% had completed 12 weeks of exercise therapy, while 13% had not engaged with exercise therapy at all. The remaining patients had completed less than 12 weeks and were either still engaged with exercise therapy (29%) or had stopped (13%). From the full cohort, 21 % had underwent surgery for their shoulder condition at four months follow-up, with 40% of these reporting to have engaged with exercise therapy for 12 or more weeks before surgery. Exercise therapy most commonly included mobility (93%), strengthening (87%), stretching (77%), and posture correction/scapula setting (57%). Only 43% of patients undergoing non-operative care had reached acceptable symptom-state.
Interpretation / Conclusion: Less than half of patients diagnosed with subacromial impingement follow the clinical guidelines recommending three months of exercise therapy. Furthermore, less than half of the patients reaches an acceptable symptom-state. Future investigations should focus on the link between guideline adherence and treatment result.

Jørgen Friis, Tim Houbo Petersen
Section for Sportstraumatology, Copenhagen University Hospital, Bispebjerg and Frederiksberg, Copenhagen NV, Denmark

Background: Irreparable Supraspinatus tendon tear is a challenging condition to treat with modalities ranging from physiotherapy to reverse arthroplasty (RSA). Superior Capsular Reconstruction (SCR) is a method to restore stability and to reduce pain by inserting a graft between the glenoid and the greater tuberosity. Using a fascia lata autograft was originally described by Mihata in 2012, but other types of grafts in order to reduce donor site morbidity are available.
Aim: The objective of this prospective study was to determine the clinical result one year post-surgery after arthroscopic SCR with a human decellularised dermal matrix graft.
Materials and Methods: Ten patients (7 men, 3 women) aged 59-75 year (average 66) with large irreparable rotator cuff tear underwent Superior Capsular Reconstruction with a human decellularized dermal graft (Lifenet Health). Seven patients had prior ipsilateral shoulder surgery. DASH and WORC score were calculated before surgery and 1 year post- operatively as well as ROM. MRI was done before operation, but only present in 5 patients after 1 year. Measurement of the acromiohumeral interval (AHI) and scoring according to fatty infiltration (Goutallier) and the Hamada classification was done.
Results: At one year follow-up two patients had a RSA due to unacceptable outcome. In the remaining 8 patients DASH score improved from 52 to 38 and WORC score from 60 to 31. Active flexion and abduction increased from 80 to 135 degrees and from 65 to 170 degrees (median values). After one year MRI showed all grafts present but defects in 4/5 grafts and no improvement in the AHI, fatty infiltration or Hamada classification.
Interpretation / Conclusion: Superior Capsular Reconstruction (SCR) with a dermal allograft in patients with large irreparable supraspinatus tendon tear gives good clinical results by improving pain and functional score in 8/10 patients. However the graft seems to have a high risk of failure, and no improvement in the radiological parameters for degenerative conditions in the cuff and joint was observed.

Sabine Simonsen, Mette Friberg Hitz, Søren Ohrt-Nissen, Stig Brorson
Centre for Evidence-Based Orthopaedics, Zealand University Hospital, Køge; National Research Center for Bone Health, Medical Department, Zealand University Hospital, Køge and Inst of Clinical Medicine, Copenhagen University; Centre for Evidence-Based Orthopaedics, Zealand University Hospital, Køge; Centre for Evidence-Based Orthopaedics, Zealand University Hospital, Køge and Inst of Clinical Medicine, Copenhagen University

Background: Proximal humeral fractures (PHF) are closely associated with osteoporosis and are the third most common non-vertebral fractures in elderly. Falls from standing height account for about 94% of proximal humeral fractures in patients older than 65. The most commonly used classification systems are the Neer and the AO classification. Both systems describe morphological aspects of the fracture anatomy aiming to support diagnostics, treatment, prognostics, research and communication. Few studies have studied the association between BMD (hip, neck and lumbar scores), BMI and fracture morphology of PHF. We hypothesize that there will be an association, and a potential for predicting patients need special need for osteoporosis management.
Aim: To study the association between the fracture morphology classified as 5 ordinal Neer-categories and 9 AO-groups, respectively, and bone mass density, age and body mass index.
Materials and Methods: A consecutive series of patients referred to Fracture Liason Service (FLS) within a 12- month period were classified based on plain radiographs according to Neer into 5 categories: 1-part, 2-part, 3-part, 4-part and + (fracture dislocation or articular surface) and 9 AO groups (A1 to C3) by the senior author. FLS included DXA scan and subtraction of BMD, BMI and evaluation of vertebral fractures and previous peripheral osteoporotic fractures.
Results: We included 56 patients, 15 one-part, 25 two-part, 15 three-part and 1 fracture dislocation. 35 fractures were AO type A, 19 were type B and 2 were type C. We found that 28 had osteopenia and 20 had osteoporosis. 28% had a previous osteoporotic fracture. There was no difference in BMI, age, or sex distribution between the groups. We found no significant correlation between Neer category and BMI, age or BMD hip, neck or lumbar scores. We found no association between Neer category and BMD hip, neck or lumbar scores on the ordinal regression analysis (p = 0.285). The same correlation and regression analysis was done for AO categories and no significant association was found (p = 0.100).
Interpretation / Conclusion: We found no association between BMD and BMI and the fracture morphology. Though, for more safe estimates more patients need to be included.

Tara Padtoft, Stig Brorson
Department of Orthopaedic Surgery, Zealand University Hospital

Background: Recently, it has been proposed that stable fractures of the olecranon (Mayo Type II) in elderly with low functional demand can be managed non-surgically. When non-surgical management is considered, functional aspects of bandaging as well as biomechanics and pathoanatomy are taken into account. We hypothesized that a thorough understanding of these aspects can be found in the rich late 18th and early 19th century medical literature.
Aim: To provide a review of historical approaches to the biomechanics, pathoanatomy, functional bandaging and complications of olecranon fractures in the pre-surgery period (1750-1850) and to discuss whether the historical sources can inform current non-surgical management.
Materials and Methods: We searched in bibliographical databases, national libraries and historical medical encyclopedias. References from potentially eligible monographs and articles were hand searched. Drawings and engravings were analyzed qualitatively by the authors.
Results: We found a comprehensive knowledge of diagnostics, biomechanics and pathoanatomy in the period 1750-1850. The deforming force of the triceps muscle on the proximal fragment was well understood. Reduction of the fragment was attempted, but retention was difficult. Several ingenious devices and functional bandages in different degrees of extension and with direct pressure were proposed for keeping the fragments together. Anchylosis was a known complication. A debate on osseous versus fibrous healing of olecranon fractures and the functional consequences of fibrous healing can be found in the early 19th century sources. A collection of previous unknown illustrations will be presented.
Interpretation / Conclusion: A rich literature on the biomechanics, pathoanatomy, functional bandaging and complications after olecranon fractures can be found in the late 18th and early 19th century. The discussion on fibrous versus osseous healing as well as principles of functional bandaging may have interest for a modern reader.

Marc Randall Kristensen Nyring, Bo Sanderhoff Olsen, Alexander Amundsen, Jeppe Vejlgaard Rasmussen
Department of Orthopaedics, Herlev and Gentofte University Hospital

Background: The minimal clinically important difference (MCID) is an important instrument in the interpretation of changes in PROM scores. To our knowledge, no MCID of the Western Ontario Osteoarthritis of the Shoulder Index (WOOS) score has ever been reported and no studies have reported a MCID for the Oxford Shoulder Score (OSS) based on patients with glenohumeral osteoarthritis, treated with an anatomical total shoulder arthroplasty (aTSA).
Aim: The aim of this study was to determine MCID for WOOS and OSS in a cohort of patients with glenohumeral osteoarthritis treated with an aTSA.
Materials and Methods: All patients treated with an aTSA for glenohumeral osteoarthritis at our institution between March 2017 and February 2019 were included. Each patient completed the WOOS and the OSS preoperatively and one year postoperatively. At one year the patients were asked to rate their overall improvement on a 7-point scale. We used an anchor-based method as our primary method to calculate the MCID, supported by two different distribution-based methods.
Results: The MCID of WOOS was 12.3 according to the anchor-based method and 14.2 and 10.3 according to the two distribution-based methods. The MCID of OSS was 4.3 according to the anchor-based method and 5.8 and 4.3 according to the two distribution- based methods.
Interpretation / Conclusion: For patients with glenohumeral osteoarthritis treated with an aTSA the MCID values were 12.3 points for WOOS and 4.3 points for OSS. To our knowledge, this is the first study to report a MCID value for WOOS and the first study to report a MCID value for OSS in this subgroup of patients.