Free Papers

1Stig Brorson, 2Björn Salomonsson, 3Anne M Fenstad, 4Steen L. Jensen, 2Yilmaz Demir, 1Jeppe V. Rasmussen
1Department of Orthopaedic Surgery, Herlev Hospital, University of Copenhagen, Denmark
2Department of Orthopedics, Karolinska Institutet Danderyds Sjukhus AB, Stockholm, Sweden
3Norwegian Arthroplasty Register, Department of Orthopaedic Surgery, Haukeland University Hospital, Norway.
4Department of Orthopaedic Surgery, Aalborg University Hospital, Denmark.

Introduction
Shoulder hemiarthroplasty is a treatment option in severely comminuted and displaced fractures of the proximal humerus. Within the last decade reverse prostheses have been increasingly popular in cases where it has been considered impossible to obtain satisfactory fixation of the tuberosities. There is a lack of reliable reporting of implant survival and reasons for revision for stemmed hemiarthroplasty and particularly for reverse shoulder arthroplasty. Our primary aim was to report revision rates and reasons for revision of shoulder replacement after acute fractures of the proximal humerus. Our secondary aim was to compare revision rates and reasons for revision between hemiarthroplasty and reverse prostheses.
Methods
This register study is based on a common data set established through collaboration between the national shoulder arthroplasty registries in Denmark, Sweden, and Norway. It contains 6,756 shoulder replacements in acute fractures between 2004 and 2013.
Results
Stemmed hemiarthroplasty was used in 90.4% of acute fractures compared to 8.4% reverse prostheses. A total of 3.3% prostheses were revised. Relative risk for revision of reverse shoulder arthroplasty compared to hemiarthroplasty was 1.07 (p=0.24). In both prosthetic designs the most common reasons for revision were dislocation and instability.
Conclusion
Reoperations after shoulder replacement in acute fractures are rare, but the number of failures may be underestimated. We found no clinical or statistical significant difference in revision rate between hemiarthroplasty and reverse arthroplasty.

Magnus Ödquist1 MD, Kristofer Hallberg2 MD, Hans Rahme3 MD PhD, Björn Salomonsson2 MD PhD, Aldana Rosso4 PhD
1 Capio Artro Clinic, Stockholm 2 Karolinska Institutet, Danderyds sjukhus AB
3 Aleris Elisabethsjukhuset ,Uppsala 4 Skånes Universitetssjukhus, Lund

Introduction
Osteoarthritis of the shoulder leads to pain and loss of function. Shoulder arthroplasty is a good treatment option, resulting in a marked reduction of pain and increased range of motion. A large number of different implants available for the last 20 years and among these the Hemi Shoulder Arthroplasty (HSA), where only the humeral head is replaced has been in frequent use. Since revision surgery after resurfacing HSA is considerably less demanding than revision after a stemmed HSA, it has been argued that resurfacing is a good option in young and active patients where the risk for future revision might be higher.

Aim
The aim of this study is to compare the distribution and outcomes of stemmed and resurfacing Hemi shoulder arthroplasty within the Swedish Shoulder arthroplasty Registry, with regard to diagnoses and age-bands and 5 year follow up.

Material and Methods
We analysed all elective primary Hemi Shoulder Arthroplasties (HSA) within the Swedish Shoulder Arthroplasty Registry (SSAR) from 1999-2009 for the diagnoses primary and secondary Osteoarthritis (OA). There were a total of 1140 HSA shoulders, 950 with primary OA, 190 secondary OA. The mean age was 66.5 years (SD 11.5) with 63.8 years for Humeral Resurfacing (RH) and 67.2 years for Stemmed Humeral implants (SH).  Of these 92 were revised, 76 with primary OA, 16 secondary OA. PROM with WOOS and EQ-5D was collected at 5 year, up until 31 December 2014.

Results
We found a 80 % higher revision rate for resurfacing arthroplasty compared to stemmed arthroplasty, but the reason for revision are in both unspecified pain or glenoid erosion and do not differ. There were no differences in reason for revision or the outcome by PROM, between RH and SH implants.  The overall outcome by WOOS was 71% for Primary OA and 65% for secondary OA, and for EQ-5D it was 0.7 and 0.6 respectively. Satisfaction level was 63% satisfied Primary OA and 58% for secondary OA.

Conclusion
Between different concepts of hemi shoulder arthroplasty the patient reported outcomes are similar. However there is an 80 % higher revision rate for resurfacing arthroplasty compared to stemmed arthroplasty, without obvious reasons. There might be a lower threshold to revise a resurfacing arthroplasty, or it might have been used in patients with higher demands.

Jeppe V. Rasmussen; Bo S. Olsen; Ali Al-Hamdani; Stig Brorson.
Department of Orthopaedic Surgery, Herlev University Hospital, Denmark.

Introduction
Resurfacing hemiarthroplasty (RHA) has a bone preserving design facilitating revision to other arthroplasty designs. Especially younger patients are treated with a RHA in the expectation that it will not be a disadvantage should the need of a revision shoulder arthroplasty arise. The aim was to report outcome of primary RHA and of revision shoulder arthroplasty after RHA in patients under the age of 55 years.
Methods
We reviewed all patients with osteoarthritis reported to the Danish Shoulder Arthroplasty Registry from 2006-2013. There were 1,210 primary RHA. 241 (20%) were used in patients under the age of 55 years. 36 (14.9%) arthroplasties in young patients required revision. WOOS was used to evaluate outcome at one year.
Results
Mean WOOS of primary RHA was 55.3, range 1.0-99.0. This was inferior to the results of older patients with a mean difference of 17.5 (95% CI 13.4-21.7, P<0.001). 92 (45.8%) patients had a WOOS score below 50 and the relative risk of a WOOS score less than 50 after was 2.2 (95% CI 1.7; 2.9), P<0.001 with older patients as reference. The 8-year cumulative revision rate was 20%. The reasons for revision were mainly glenoid attrition (n=13) and technical failure including overstuffing and malposition of the implant (n=9). 15 arthroplasties were revised to stemmed hemiarthroplasty, 12 to anatomical total shoulder arthroplasty and 3 to reverse shoulder arthroplasty. The median WOOS score of revised RHA was 46.6, range 20.0-97.0, and 62% of the revised arthroplasties had a WOOS score below 50. 
Conclusion
The outcome of primary RHA in young patients is unpredictable and age should be regarded as a major risk factor for an inferior outcome. The outcome of revision shoulder arthroplasty after RHA was disappointing, as only 32% was revised with an acceptable outcome. Using RHA to delay total shoulder arthroplasty should be avoided. The role of RHA in the treatment osteoarthritis needs to be clarified, especially for younger patients.

Kim Schantz1, Ulrik Kragegaard Knudsen1, Tommy Henning Jensen1, John Kloth Petersen1, Signe Rosenlund2
1Department of Orthopedic Surgery and Traumatology, Koege Hospital, Denmark
2Department of Radiology and Diagnostic Imaging, Koege Hospital, Denmark
Keywords: Resurfacing hemiarthroplasty, Osteoarthritis, acromioclavicular joint resection, Patient-reported outcome, Revision rate
Disclosure: No conflicts of interest
Characters in total: 3223 (Background to Conclusion)

Background: The Copeland resurfacing hemiarthroplasty (RHA) has been used to treat patients with degenerative shoulder disease since the mid 1990’s. The Copeland prosthesis has been linked to high revision rates and inferior patient-reported outcome [Rasmussen et al., 2014 and Voorde et al., 2015].
The median [25th and 75th percentile] Western Ontario Osteoarthritis of the Shoulder index score (WOOS) in patients diagnosed with degenerative shoulder disease who underwent surgery at Koege Hospital, Orthopaedic department, was 89.1 [60.4 and 95.8] which is 12.0 points higher than the national median of 77.1 [51.5 and 93.1] [The Danish Shoulder Arthroplasty Registry, Annually report, 2015]. Acromioclavicular joint (AC-joint) resection was performed in patients with signs of AC-joint arthrosis. AC-joint resection is not performed at all orthopaedic departments.
Purpose: We investigated the 1 year post-operative patient-reported outcome after Copeland RHA with or without AC-joint resection and the overall revision rate with 6 years follow-up.
Methods: We include all patients who had surgery with the Copeland prosthesis at Koege Hospital Orthopaedic department in a 6 years period from January 2008 to December 2013. The Danish Shoulder Arthroplasty Registry was searched to identify all patients with Copeland RHA surgery at our department in that period and to gain WOOS score one year post-operatively.
The WOOS score at one year was used to evaluate the patient-reported outcome. An audit of the medical journals in January 2015 was used to confirm the type of prostheses, to identify patients with AC-joint resection and if the patient has had a revision.
Results: In total 138 consecutive Copeland RHA in 132 patients were evaluated. The mean (SD) follow-up time was 3.2 (± 1.6) years. The cohort included 46 (35%) males and 86 (65%) females.
The mean age was 69.2 (± 8.8) years at surgery. AC-joint resection was performed in 51 (37%) shoulders. In total 89 patients returned the WOOS questionnaire at one year. The median WOOS score was 89.5 [60.4 and 95.9]. The median WOOS score for the group of patients with AC-joints resection was 89.7 [78.9 and 97.2] and for patients without AC-joint resection the median WOOS score was 89.0 [57.5 and 95.3]. There was no statistically significant difference in median WOOS score between the patient group with AC-joint resection compared to the patient group without AC-joint resection. Three patients had a revision and 13 patients died within the 6 years of follow-up.
Conclusion: Overall the cohort had high patient-reported outcome after one year compared to the national median outcome score. Thus, our results from the Copeland RHA are more in line with the results of the total shoulder arthroplasty than the results of the hemiarthroplasty, as reported in the Danish Shoulder Arthroplasty Register [Danish Shoulder Arthroplasty Registry, Annually report, 2015]. We found no difference in patient-reported outcome for patients with and without AC-joint resection. Finally we found a low revision rate. We have no explanation for the discrepancy between our results on both patient-reported outcome and revision rate and the ones reported in the literature.

Juha Paloneva, Vesa Lepola2, Ville Äärimaa3, Antti Joukainen4 and Ville M Mattila2,5

1Department of Surgery, Central Finland Hospital, Jyväskylä, Finland and University of Eastern Finland. 2Division of Orthopedics and Traumatology, Department of Trauma, Musculoskeletal Surgery and Rehabilitation, Tampere University Hospital, Tampere, Finland. 3Department of Orthopaedics and Traumatology, Turku University Hospital, Turku, Finland. 4Department of
Orthopaedics, Traumatology and Hand Surgery, Kuopio University Hospital, Kuopio, Finland. 5University of Tampere, Tampere, Finland.

Background: Increased rotator cuff repair incidences have been reported from the United States, England and Finland. We analyzed recent nationwide data relating to rotator cuff repairs recorded in the Finnish National Hospital Discharge Register (NHDR).

Methods: The NHDR was reviewed to identify adult patients who underwent rotator cuff repair between 1997 and 2014. Incidence rates per 105 person-years were calculated using the annual adult population size.

Results: During the 18-year time period, 68 140 rotator cuff repairs were performed on subjects aged 18 years or older. Mean age of the patients undergoing RCR was 56 years for men and 58 years for women in 2014. The incidence of rotator cuff repair showed an almost linear increase of 269 %, from 36 per 10000 person-years in 1997 to 133 per 105 person-years in 2011, after which the incidence decreased by 20 % to 106 per 105 person years in 2014. The incidence in men was approximately double compared to women. The incidence peaked at 345 per 105 in men aged 45 to 64 years in 2011. The most common concomitant procedure was acromioplasty, which was reported in approximately 38 % of the RCR operations. Tenotomy of the long head of the biceps tendon (CLBB) showed an increasing rate and was performed in 13 % of the cases. Other common concomitant procedures included tenodesis of the CLBB (7 %) and resection of the acromioclavicular joint (4 %).

Conclusions: This nationwide analysis revealed a remarkable increase in the incidence of rotator cuff repair between 1997 and 2011, after which the incidence has started to decline. We propose that this change in clinical practice is due to scientific evidence that shows no clear difference in outcome between surgical and nonsurgical interventions for degenerative rotator cuff tear. However, the exact cause of the declining incidence cannot be defined based solely on a registry study. The incidence in Finland is high compared to rates reported elsewhere.