Supporting Regional Upper Limb Arthroplasty: The Impact of Establishing an Upper Limb Orthopaedic Network

Abstract

Background and Objectives

The management of complex upper limb arthroplasty has received national guidelines supporting the use of a regional network. An upper limb network was established for both elbow and shoulder arthroplasty. This study evaluates the impact of establishing this network over a 5-year period.

Methods

Data were collected from network meetings from June 2017 to December 2022. Hospital-level National Joint Registry data were obtained for analysis of case volume.

Results

A total of 243 cases were discussed. Network discussion changed the management plan in 53% of cases. Only 8% of cases required transfer to the tertiary center. The proportional caseload at either hub or spoke hospitals did not change after creation of the network.

Discussion

Regional network discussion aids decision-making for complex cases, with further management options realized in over half of the cases presented. The discussion allowed every patient to gain the advantage of regional expertise while being managed conveniently at their local hospital.

Introduction

As the cumulative number of patients living with upper limb arthroplasty increases so will the need for high-quality revision surgery. The National Health Service is under unprecedented pressure to provide high-quality care to patients with increasingly complex needs. ¹ The Getting It Right First Time Report (GIRFT) has highlighted a worrying trend for primary and revision arthroplasty to be undertaken by centers and surgeons with low volume. ² This is echoed in the National Joint Registry, where the median number of primary shoulder replacements completed per unit is 12 (interquartile range [IQR] 5–23) and 2 (IQR 1–4) for primary elbow replacements. ³ Subsequent British Elbow & Should Society guidelines propose that both primary and revision elbow replacements are performed at a centralized regional hub to ensure an adequate yearly caseload. ⁴ Our region provides care for ∼2 million people. The hub of the network is located in the regional city and the spoke units are up to 60 miles from the hub, complete centralization of services would have a significant impact to patients and families with travel time to the regional unit being in excess of an hour from several of the peripheral hospitals.

Following the GIRFT report and the success of the lower limb arthroplasty model, ⁵ it was proposed to set up a regional upper limb network to support shoulder and elbow arthroplasty surgeons in the region. It aimed to allow surgeons to share experience and best practices when tackling complex primary procedures and revision upper limb arthroplasty cases. This has the potential to improve outcomes for patients and minimize complications and morbidity while providing a convenient, cost-effective local service for patients and families.

Centralization of complex care has been shown to improve outcomes in many specialties. ⁶ However, there are concerns about the impact this will have on both regional centers due to additional workload and patients having to travel significant distances for care. The aim of this study is to assess the impact an upper limb network has had on the provision of upper limb arthroplasty in the region in terms of any additional burden of cases on the tertiary center and the impact of network discussion on clinical decision-making.

Methods

The infrastructure for the upper limb network is similar to the previously reported lower limb network. ⁵ Meetings were initially held monthly with Nottingham University Hospitals acting as the hub and 4 district general hospitals as the spokes. Meetings are chaired from the hub hospital by a revision arthroplasty surgeon, other specialities (such as vascular, plastics, and microbiology) are available if required.

The meeting is coordinated by a dedicated multidisciplinary facilitator. Electronic referrals to the meeting are completed using a standardized template to ensure all required clinical information is recorded and submitted using a secure @nhs.net email account (Supplemental Appendix 1). Teleconferencing is provided using the Microsoft Teams platform (Microsoft) and radiographs are shared using a regional picture archiving and communication system (PACS, GE Healthcare). The technology allows all users to share a single computer desktop so that each surgeon can control the selection and presentation of radiographs to the network.

A prospectively maintained database of patients who have been discussed has been maintained since the inception of the upper limb regional meeting. Additional information for this study has been extracted from PACS and patient case notes. Data were extracted from this database from its inception in June 2017 to December 2022. Hospital-level National Joint Registry (NJR) data were obtained for the same time period to allow an analysis of case volume at all centers.

Results

A total of 243 cases have been discussed from June 2017 to December 2022, this consisted of 156 revision cases and 35 primaries. On average 4 cases were discussed per meeting (median, IQR 2–5). The majority (196, 81%) were shoulder arthroplasty cases, with the remaining 19% being elbow arthroplasty cases. A year-on-year increase in overall caseload was seen, 30 cases in 2017, 51 cases in 2018, and 66 cases in 2019. A subsequent decline was seen in the following years with reduced elective activity as a consequence of COVID-19.

A total of 192 (79%) cases were referred for discussion from the hub center and 51 (21%) cases from the spoke centers. The majority of cases were discussed due to component loosening or wear requiring revision, the reasons for referral are summarized in Table 1. Only 19 cases (8%) required transfer to the tertiary center, the remaining cases were managed in the referring hospitals after discussion. Discussion within the meeting altered management plans in 48% (n = 118) of cases. In 43%, this was a change in the surgical strategy, a suggestion of further investigation prior to surgery in 30%, or a different choice of components in 21% of cases.

Table 1.

Reasons for Referral to the EMSON Meeting.

Reason for referral	N	%
Loosening	59	24
Wear	39	16
Primary	36	15
Infection	34	14
Fracture	23	9
Altered biomechanics/cuff failure	20	8
Dislocation/subluxation of implant	12	5
Unknown	11	5
Pain? Cause	1	1
Fall (no fracture)	1	1

Analysis of NJR data from June 2017 to December 2022 is summarized in Figure 1 and Table 2. There is a steady increase in overall caseload until 2020 when a significant reduction in elective activity occurred as a result of the COVID-19 pandemic. Since then the regional caseload has returned to prepandemic levels. Overall, there does not appear to have been an increase in proportional caseload at the regional hub, or a decrease in caseload at spoke hospitals after implementation of the upper limb network.

Figure 1.

Line graph plotting hub and spoke arthroplasty volume in comparison to network discussion caseload.

Table 2.

Summary of National Joint Registry (NJR) data for regional upper limb arthroplasty caseload.

Year	Hospital	Elbow	Shoulder	No. of cases	% of regional volume
2017	Hub	18	120	138	47
	Spokes	21	133	154	53
2018	Hubs	10	98	108	37
	Spokes	25	160	185	63
2019	Hubs	11	123	134	44
	Spokes	17	154	171	56
2020	Hub	8	54	62	46
	Spokes	11	62	73	54
2021	Hub	9	47	56	33
	Spokes	15	99	114	67
2022	Hub	6	103	109	45
	Spokes	11	120	131	55

Discussion

The upper limb orthopedic network has delivered positive results for both the hub and spoke units within the network by providing a forum to discuss a wide mix of shoulder and elbow cases. A significant number of management plans have been altered following network discussion taking advantage of the combined experience of surgeons within the network.

While there is growing evidence relating to minimum numbers per surgeon and unit for hip and knee surgery,^7,8 there is less evidence relating to upper limb arthroplasty. In the literature, high-volume surgeons have been defined as those completing more than 11 total shoulder replacements ⁹ or 30 “any-type” arthroplasties.^10,11 At present, the GIRFT report has recommended that individual surgical societies develop their own recommendations. ² It seems likely that individual surgeon volume is only a part of the picture, and that intuitional experience has a significant impact on patient outcomes. The creation of the network has allowed the experience and expertise of surgeons in the region to be combined, almost half of cases had their proposed management plan altered after discussion.

The lack of clinical outcome data for patients and inclusion of patients prior to the network significantly limits the study and hence the conclusions that can be drawn. In the absence of this, the impact of the network is inferred from the effects the discussion had on proposed management plans and its subsequent effect on that patient's care. However, the majority of these alterations were changes in either surgical technique or implant choice, both of which could be considered significant management changes and hence the patient's outcome. The effect of these changes cannot be quantified based on this study alone and is an area for further research. Despite these limitations and given only 8% of cases required transfer of care, it can be concluded that the vast majority of patients would have benefitted from a consensus expert opinion with the convenience of being managed locally.

The restructuring of lower limb arthroplasty care brought about by GIRFT has generated concern that the centralization of revision arthroplasty will result in tertiary centers having additional workloads and clinical practice skewed by solely absorbing the revision needs of the network. ¹² However, over the period that the network has been running, the overall caseload at the tertiary center has not disproportionately increased in relation to the spoke units (Figure 1). This may be a reflection of the overall lower revision volume in upper limb arthroplasty in comparison to lower limb not having a significant effect on the overall workload at a high-volume center.

Only 21% of the network caseload originated from the spoke centers, given that NJR data show an approximately equal division in overall arthroplasty caseload between the hub and spokes centers, this figure is lower than expected. One possible cause for this discrepancy is that the majority of revision arthroplasty cases that would necessitate discussion are referred directly to the revision hub from primary care, hence the revision caseload is primarily generated by a single unit. Alternatively, there may be other factors such as logistical issues or timetabling constraints preventing more cases at the spoke centers from being discussed.

While the network has been a success clinically, there have been several difficulties in its establishment. The most common feedback from all consultants involved is the difficulty in identifying a mutually convenient time for the meeting. The meeting has been scheduled in a lunchtime slot rather than a morning or afternoon session and consultants have found it difficult to gain appropriate recognition in their job plans. Technological faults occasionally prevent the PACS system and videoconferencing software from functioning fully, but the inclusion of an experienced facilitator has helped to find solutions when required.

Conclusion

We feel that this upper limb network has been a success in bringing together surgeons from 5 hospitals to share their combined experience of more than 250 annual arthroplasties. In almost 50% of cases, discussion resulted in a change to the clinical plan, and hence the care received by the patient.

Its introduction has not resulted in an increased surgical workload at the hub hospital with most referred patients being retained by spoke hospitals. The specialist network is an ideal method to ensure that every patient gains the advantage of shared experience while retaining the ability to be cared for conveniently at their local hospital.

Supplemental Material

sj-pdf-1-sea-10.1177_24715492241258623 - Supplemental material for Supporting Regional Upper Limb Arthroplasty: The Impact of Establishing an Upper Limb Orthopaedic Network

Supplemental material, sj-pdf-1-sea-10.1177_24715492241258623 for Supporting Regional Upper Limb Arthroplasty: The Impact of Establishing an Upper Limb Orthopaedic Network by Ben Oakley, Ben Marson, Malin Wijeratna, Paul Manning, John Geoghegan and Ben Gooding in Journal of Shoulder and Elbow Arthroplasty