The treatment of patients with arthritis of the glenohumeral joint with an associated massive irreparable cuff tear is challenging. Since these patients usually have proximal migration of the humerus, the CTA extended head allows a surface with a low coefficient of friction to articulate with the acromion.

Between 2001 and 2006 a total of 48 patients with arthritis of the shoulder joint associated with a massive cuff tear, were treated with a CTA head. The indications for use being Seebauer Type 1a and 1b appearances on x-ray and active abduction of the arm to more than 60° with appropriate analgesia. Preoperatively, a Constant score and an ASES pain and function score were completed as well as standard radiological assessment. These were repeated at follow up. Paired t tests were carried out for all the variables. A Kaplan-Meier survival analysis was performed.

Follow up varied between 2 and 8 years. Improvements in pain, function and all movement parameters were significant at p< 0.001. There was no change in the strength component. Survival analysis showed 94% survival at 8 years (95% CL 8%) there were 2 revisions and 5 deaths. Radiological assessment at follow up revealed no evidence of humeral stem loosening. In 5 (17%) cases however there was evidence of erosion in the surface of the acromion and in 13 (45%) erosion of the glenoid. Finally one component was also seen to have subluxed anteriorly.

This head design has been in use for a number of years. To date there appears to be no reported outcome of their use. This series shows that in an appropriately selected patient a satisfactory clinical outcome can be maintained in the short to medium term. The presence of erosion of the glenoid but also the under surface of the acromion does require continuing monitoring.

Purpose: To measure any observed migration and rotation of humeral and ulnar components using radiostereometric analysis.

Methods: From 2002–2004 in a prospective study, twelve elbows in patients treated with either a linked(3) or unlinked(9) Acclaim total elbow prosthesis were included in a radiostereometry study. Six tantalum markers were introduced into the humerus another 3 markers were located on a humeral component. Four markers were placed in to ulna and three markers located on the ulnar component. RSA radiographs were taken postoperatively, six, twelve and twenty-four months. The radiographs were digitised and analysed using UmRSA software. The relative movement of the humeral and ulnar implants with respect to the bone was measured.

Results: At twelve months, the largest rotation of the humeral component was anteversion/retroversion with a mean of 20, anterior tilt had a mean of 1.10 and varus/valgus tilt was minimal mean 0.60. Mean difference between twelve and 24 months segment rotation was no more than 0.50. Rotation in unlinked humeral implants reached a plateau at 12 months this did not occur in linked implants. Paired t-tests between twelve and 24 months segment translation data showed the mean differences to be no more than 0.04mm. In contrast, humeral tip motion produced a mean of 1.3mm at 12 months dominated by movement in the horizontal plane with a mean difference at 24 months of 0.2mm. Mean micro-motion of the ulna implant tip was 0.32mm at 12 months with a mean difference at 24 months of 0.1mm.

Conclusions: Early micromotion of the Acclaim humeral implant occurs mostly by rotation about the vertical axis accompanied by anterior tilt. In unlinked implants this motion reaches a plateau at 12 months after operation but does occur in linked implants.

To compare the effect of standard non-offset and offset humeral heads on early glenoid loosening. From 2002–2004 in a prospective study, 12 patients with osteoarthritis and intact rotator cuff had a TSR with an offset humeral head matching normal anatomy. Five tantalum markers were introduced into the scapula and acromion, spaced widely apart; four markers were placed in a pegged glenoid component. RSA radiographs were taken postoperatively, three, six, twelve months. The micromotion of this offset head group was compared with a baseline series of TSRs with standard non offset humeral head implants which had previously been measured over 24 months. At 24 months, the largest segment translation for the offset head group was in the anterior/posterior direction with a mean of 0.24mm, the standard head group mean was 0.61mm, t=1.3,P=0.21. Proximal/distal translation gave means of 0.51mm (standard head) and 0.17mm (offset head), t=1.3,P=0.21. Medial/lateral translation was smallest for the offset group with mean 0.16mm this is less than half the standard group mean 0.57mm and is significant, t=1.4,P=0.19. The largest rotation was anteversion/retroversion with a mean of 1.50 (offset) compared to 4.40 (standard), t=2.3, P=0.04. Varus/valgus tilt differed significantly with means 0.90 (offset) and 2.50 (standard), t=3.8,P=0.001; and finally anterior/posterior tilt had smallest means of 0.70 (offset) and 1.00 (standard), t=1.2,P=0.25. Maximum total point motion was also significantly different at 24 months, means of 0.66mm (offset) compared to 1.65mm (standard), t=2.3,P=0.04. Early migration occurs in both groups of pegged components. A principal component micromotion index illustrates the complexity of micromotion. Offset heads move significantly less than the standard type, the significant differences in MTPM movement and in rotation (varus/valgus tilt) suggest that the offset head may decrease eccentric loading

We present the early results of 36 primary total elbow arthroplasties using the Acclaim prosthesis. The Acclaim prosthesis was used in 46 primary total elbow arthroplasties between July 2000 and August 2002. All operations were performed or directly supervised by the two senior authors (IAT and JKS). There were 32 females and 14 males. The mean age at surgery was 64 years (range, 34–93). The underlying pathology was rheumatoid arthritis in 39, osteoarthritis in five and post-traumatic arthritis in two. The early results of 36 cases are presented at a minimum follow-up of two years. Patients were assessed using the American Shoulder and Elbow Surgeons patient self assessment form and the range of movement of the elbow measured. The Wrightington method was used for radiographic analysis of lucencies. There was good relief of pain and range of movement improved. The mean preoperative pain score was 8.1 and decreased to 2.1 at latest follow up. The mean disability score increased from 34.2 to 66.1. The mean overall satisfaction rating following surgery was 9.3 on a visual analogue scale from zero to ten. The mean range of flexion increased from 83^oto105^o. The mean flexion gain was just over 10^o and the mean extension gain was just over 12^o. There were 11 cases of intraoperative fracture of the humeral condyle. One of these fractures failed to unite and required revision to a linked prosthesis because of persistent instability. There was one case of deep infection. There were three cases of ulnar neuropathy, one of which resolved. There was no evidence of loosening. The Acclaim total elbow arthroplasty gives good symptomatic relief and improvement in function according to the American Shoulder and Elbow Surgeons patient self assessment form. These early results are encouraging but the frequency of intra-operative fractures is of some concern.

To measure any observed migration and rotation of humeral and ulnar components using radiostereometric analysis. From 2002–2004 in a prospective ongoing study, twelve elbows in patients treated with either a linked or unlinked Acclaim total elbow prosthesis were included in a radiostereometry study. Six tantalum markers were introduced into the humerus another three markers were located on a humeral component. Four markers were placed in to ulna and three markers located on the ulnar component. RSA radiographs were taken postoperatively, six, twelve and twenty-four months. The radiographs were digitised and analysed using UmRSA software. The relative movement of the humeral and ulnar implants with respect to the bone was measured. At twelve months, the largest segment translation of the humeral component was in the anterior/posterior direction with a mean of 0.44mm followed by medial/lateral translation of 0.39 mm; there was minimal proximal/ distal translation or with a mean of 0.16mm. Paired t-tests between twelve and 24 months segment translation data showed the mean differences to be no more than 0.056mm. The largest rotation at twelve months was anteversion/retroversion with a mean of 2.40deg, anterior tilt had a mean of 1.20deg and varus/valgus tilt was minimal mean 0.60deg. Mean difference between twelve and 24 months segment rotation was no more than 0.30deg. In contrast, humeral tip motion produced a mean of 1.1mm at 12 months dominated by movement in the plane horizontal plane with a mean difference at 24 months of 0.06mm. No patients could be measured for segment micromotion of the ulnar component due to technical difficulty in visualising tantalum markers in the ulna. Early micromotion of the Acclaim humeral implant occurs mostly by rotation about the vertical axis accompanied by anterior tilt. This motion reaches a plateau at 12 months after operation.

Aim: Patients with advanced rheumatoid arthritis who were managed with ipsilateral shoulder and elbow arthroplasties were reviewed to determine appropriate sequence of surgery, operative technique and the functional outcome.

Methods: Between 1992 and 2002, twenty-two patients underwent ipsilateral elbow and shoulder arthroplasties. Nineteen patients were available for final review. Clinical and radiological assessments were done on these patients.

Results: Twenty-four upper limbs in nineteen patients were reviewed. Mean age at final follow-up was 61.1 years (49.9 to 73.3yrs; SD 8.2). Mean duration of follow-up from the last operation was 55.5 months (12 to 129.3m; SD 33.5). The average interval between the operations was 40.1 months; it was 41.2 months when elbow operated first and 38.7 months when shoulder was operated first. This difference was not significant (p=0.82). All movements showed significant improvement after respective joint replacements. There was a significantly greater improvement in external rotation of the shoulder when it was operated first (p=0.48). The average improvement in Constant-Murley scores was 28.8 points; with no statistically significant difference between either sequence of operations (p=0.49). However, there was statistically significant improvement in the average Mayo elbow performance score after the elbow arthroplasty when it was operated first (p=0.03).

Two patients needed conversion of shoulder hemi-arthroplasty to total shoulder replacement due to subsequent erosion of the glenoid. One elbow replacement was revised because of recurrent dislocations. There were four patients who developed ulnar neuropathy, of which two were permanent. There were no peri-prosthetic fractures in this series. One patient needed custom-made short-stemmed shoulder prosthesis due to the presence of a long-stemmed humeral component of total elbow prosthesis in situ.

Conclusion: Ipsilateral shoulder and elbow replacements significantly improve pain and function of the limb, when there is advanced arthritis. The joint that appears clinically and radiologically worse should be replaced first. However if both the joints are equally involved we feel that elbow should be replaced first as the functional improvement seems to be better. Careful preoperative planning is required in choosing the type and size of prosthesis, to avoid potential complications.

The purpose of this study was to analyse the reason for failure of shoulder arthroplasties performed for osteoarthritis, looking specifically at factors associated with either hemi or total shoulder replacement.

Over an eight-year period over 80 consecutive patients undergoing a Global hemi-arthroplasty or total shoulder arthroplasty were looked at. All patients underwent at least two years of follow up. All patients were assessed pre and post operatively using the Constant and ASES scoring systems. A bio-engineer or physiotherapist performed these evaluations. X-rays were analysed for demarcation, loosening and glenoid erosion. Failure was defined as the need for revision surgery. Reasons for revision were predominately pain, loosening and instability. In these instances the differences between hemi and total joint arthroplasty were recorded. A distinct pattern was present with several early revisions being performed for continuing pain in the patient group who underwent hemi-arthroplasty.

In conclusion, this study supports total shoulder replacement as the operation of choice, in suitable patients, as compared to hemi-arthroplasty. This would appear to reduce the risk of the need for early revision on account of pain. This must be weighed against the potential for late glenoid loosening.

The purpose of the project was to develop a questionnaire for completion by patients with elbow pathologies which is short and practical, internally consistent, valid, responsive and sensitive to changes of clinical importance.

The first, pilot phase included 43 patients who each completed a 19 item questionnaire relating to elbow function. The 19 ADLs produced a total scale Cronbach Alpha of 0.96., two different groups of ADLs were identified by multivariate analysis. Group 1 consisted of ADLs requiring moderate to high isometric loading and Group 2 of ADLs requiring high flexion. From the 19 items the best 10 which represented both groups were selected. A summary score was used to create the Wrightington Elbow Disability Score (WEDS).

In the second phase 89 patients completed the new WEDS form, reliability studies produced a Cronbach’s alpha value of 0.91. Internal validity of the groups of ADLs all correlated at p< 0.001 level with strength (Group1) and flexion (Group 2). A sub set of 40 patients undergoing total elbow arthroplasty were assessed for sensitivity to change in disability, the WEDS indicated a significant improvement at the p< 0.001. Convergent validity was demonstrated by the correlation with the ASES-e score at p< 0.001 level. The WEDS was significantly correlated with the ASES-e but not the DASH score.

Our study confirms that the WEDS questionnaire which is short and practical, is internally consistent, valid, responsive and sensitive to changes of clinical importance.

To measure any observed differences in migration and rotation between keeled and pegged glenoid components using roentgen stereo-photogrammetric analysis.

Between 2000–2001 in a prospective randomised study, 20 patients with osteoarthritis had a TSR with roentgen stereo-photogrammetric analysis. Five tantalum markers were introduced into the scapula and acromion, spaced widely apart. Another 4 markers were placed in either a pegged or keeled glenoid component.

RSA radiographs were taken postoperatively, three, six, twelve and eighteen months. The radiographs were digitised and analysed using dedicated software (UmRSA). The relative movement of the glenoid with respect to the scapula was measured.

The largest translation for the keeled components was in the proximal/distal direction with a mean of 1.3mm, the pegged group mean was 0.27mm. This difference was significant, P = 0.001. Both other translation axes showed no significant difference between the two component types. Medial/lateral migration gave means of 0.38mm(keel) and 0.52mm(peg) and the anterior/ posterior translation with means of 0.54mm(keel) and 0.41mm(peg). Maximal total point motion mean values at 18 months were 2.6mm for keeled and 1.1mm for pegged glenoid components. This difference was also significant P=0.001

The largest rotation was anterior/posterior tilt with means of 3.5°(keel) and 1.1°(peg) this difference was significant p=0.005; varus/valgus tilt with a mean of 3.2°(keel) compared to 20(peg) was again, significant P = 0.002. and finally anteversion/retroversion means were 3.3°(keel) and 2.7°(peg). Multivariate analysis identified 2 principal components from the keeled data each accounting for 35% of the variation in the data. The first included transverse and saggital translation together with rotation about the longitudinal axis; the second component consisted of translation on the longitudinal axis with rotation about the transverse axis.

Between 1992 and 2000, 120 shoulder arthroplasties have been inserted in patients with osteoarthritis at this hospital. Shoulder function was assessed using the ASES patient self-evaluation form and the Constant-Murley clinical index as well as evaluation including routine measurements of shoulder motion and pain. The mean follow-up was 4.4years (2 to 10years).

Movement, pain and both shoulder function scores significantly improved after operation with t-test values significant at P< 0.001 level. Three sub-groups of primary osteoarthritis (OA), avascular necrosis (AVN) and post traumatic arthritis (PTA) were analysed separately. Although each group was significantly improved after arthroplasty the AVN and PTA patients had significantly poorer results than OA patients for pain and the functional scores, except the AVN group for the Constant Murley score. General linear model analysis indicated that both Constant-Murley and ASES shoulder function scores were shown to be sensitive to change after treatment and that improvement was maintained for up to 8 years post-operation.

Kaplan-Meier analysis showed a significant difference in survival between the 3 sub-groups of OA, PTA and AVN with 5 of 16 PTA operations needing revision. This was reflected in function measurement with PTA values 20 points lower in both clinical (P=0.014) and self assessed functional status (P=0.004).

At each visit a true AP radiograph was taken. Radiological lucency was graded using the method of Matsen. The humeral components were inserted without cement and there were no signs of loosening. Just four components had lucency at one of 7 zones. There was a noticeably higher incidence of glenoid lucency, Proportional hazards analysis was used to model the hazard rate of glenoids at risk. A hazard model for the glenoid components consisted of only one variable – presence or absence of glenoid erosion noted at operation. The model is significant (P= 0.02) with erosion having a hazard rate 6 times greater than that if not eroded.