Current concepts in the management of radial head fractures: a national survey and review of the literature

Abstract

Introduction

This study aims to report on current practice in the management of radial head fractures (RHFs) in the United Kingdom and to review the literature to identify areas for future investigation.

Methods

A 12-question online survey was sent to 500 surgeon members of the British Elbow and Shoulder Society in 2021. Questions focused on clinical assessment, indications for surgical treatment and willingness to participate in future studies. Descriptive statistical analysis summarised the responses.

Results

The response rate was 20.4% (n = 102). For minimally displaced RHFs, non-operative management with immediate mobilisation was reported by 90.2% (n = 92) as opposed to 9.8% (n = 10) for initial immobilisation in plaster or brace. The most cited indication for radial head arthroplasty as opposed to fixation was increased patient age or low functional demand in 69.9% of responses (n = 71). In total, 41.2% (n = 42) indicated the need for a future randomised controlled trial (RCT) concerning the management of RHF. Only five RCTs were returned by the literature review.

Discussion

There is considerable variability in the management of RHF among an experienced cohort of surgeons. With the interest declared by the participating surgeons, there is a call for a well-designed sufficiently powered RCT.

Introduction

Radial head fractures (RHFs) are the most common elbow fractures constituting one-third of elbow fractures and 4% of all fractures.¹ The most common mechanism of injury is a fall on an outstretched hand, usually with a degree of elbow flexion and forearm pronation leading to radial head impact against the capitellum and resultant radial head/neck fractures. Such injuries have the potential to have a significant impact on upper limb function because the radial head acts as a critical stabiliser for the elbow against valgus stresses secondary to the medial collateral ligaments. The most adopted classification system in clinical practice is that standardised by Mason in 1954.² This comprises three types: Mason I is a non-displaced marginal or fissure RHF; Mason II is displaced; and Mason III is a comminuted RHF. Despite proving consistent good inter- and intra-rater reliability, the definition of displacement was not clearly defined in this system. Broberg and Morrey³ further defined Mason II as fractures with more than 2mm of displacement, involving >30% of the radial head or that cause a mechanical block to forearm rotation. An additional type IV was added, which involved RHF associated with an elbow dislocation.

Treatment for RHF ranges from non-operative management to surgical intervention involving surgical fixation, arthroplasty or radial head excision (RHE). Despite the ubiquity of RHFs in the sphere of elbow trauma, there is a lack of consensus regarding their management. This is driven, in part, by few previous well-designed randomised controlled trials (RCTs). The aim of this study was to establish an overview of current UK practice in the management of RHFs as well as reviewing the existing literature.

Methods

Survey contents

The survey comprised 12 multiple-choice questions (Appendix 1 – available online). The questions were designed to assess various aspects of the entire clinical management of these injuries. This included approaches to initial assessment and emergency management of RHFs, indications for operative intervention, preferred methods for operative vs non-operative management, and how fracture morphology as well as patient factors are implicated in decision-making. As an adjunct to this information-gathering process, we assessed the willingness and interest of the respondents to participate in a future RCT to further our experience in the management of RHFs.

Survey administration

Five hundred surgeon members of the British Elbow and Shoulder Society (BESS; consultant and senior specialty trainees) were invited via email to participate in an online survey using Qualtrics software (Qualtrics, Provo, UT, USA). Participation was voluntary and no consent was required. The survey was designed to take no more than 10min to complete with no minimum number of responses required. The response rate was defined as the number of respondents divided by the number of eligible people invited (only surgeon BESS members excluding allied health professionals). Descriptive statistical analysis summarised the responses either quantitatively or narratively, as appropriate (using Microsoft Excel version 16.12 and Prism version 7.0).

The email invitation was sent out on 1 August 2021 with subsequent reminder survey emails sent and the survey closed on 15 September 2021. The survey was approved by the BESS Research Committee beforehand. Duplicate answers were excluded automatically by the survey software.

Results

The response rate was 20.4%. Of the 102 respondents, 64 (62.7%) were identified as consultant orthopaedic surgeons. There were 5 (4.9%) fellows and the remaining 33 (32.3%) respondents submitted their responses anonymously.

Surgical indications

Question 1 asked how any mechanical block to pronation/supination is assessed. Two-thirds of respondents (64.7%, n = 66) choose to re-examine the patient 1–2 weeks post-injury thus minimising the effect of pain inhibition on clinical examination. About one-third of respondents (31.3%, n = 32) opt for aspiration of haemarthrosis and injection of local anaesthetic acutely to assess objective mechanical block to forearm rotation. Only three respondents (2.9%) prefered formal examination under anasthesia in a theatre setting, proceeding to open reduction and internal fixation (ORIF) if indicated.

The second question assessed indications for the operative management of RHF (Figure 1). RHF as part of a terrible triad injury was the determinant factor in opting for surgical management according to 97% (n = 99) of respondents. This was followed by mechanical block to rotation in 88.2% (n = 90) of responses. There was less consensus on the cut-off for the degree of fracture displacement that would mandate surgical intervention with 49% (n = 50) of respondents choosing to operate on a fracture with any displacement or comminution, 40.1% (n = 41) for complete cortical discontinuity when present and 32.3% (n = 33) of respondents having a clear cut-off of at least 2mm of displacement for a surgical decision. Only 4.9% (n = 5) of the respondents would opt for surgical fixation of all type II fractures, regardless of presence of a mechanical motion block.

Figure 1 .

Bar chart demonstrating different surgical indications as per the survey responses

Operative treatments

Patient age and mode of surgical management were assessed by further questions (Figure 2). For comminuted RHF in middle and older-aged patients, radial head arthroplasty (RHA) was the most common treatment as evidenced by 56.7% (n = 58) of the responses. This was followed by RHE in 19.6% (n = 20), screw fixation in 15.7% (n = 16) and least commonly by plate fixation in 7.8% (n = 8) of responses.

Figure 2 .

Bar chart demonstrating age implications in choice of surgical treatment of radial head fracture

Conversely, procedures conserving the radial head such as screw fixation were more favoured in younger patients in 56.7% (n = 58) or by plate fixation in 20.6% (n = 21) of respondents. Only 14.7% (n = 15) and 6.9% (n = 7) of respondents would opt for RHA in or RHE, respectively, in this younger cohort.

The most popular treatment for radial neck fractures (two-part) was screw fixation in a tripod configuration in 43% (n = 54) compared with plate fixation in 29.4% (n = 30), RHA in 11.8% (n = 12) and RHE 2.9% (n = 3) of responses.

The most cited indication for RHA instead of ORIF was increased patient age or low functional demand in 69.6% (n = 71) of the responses, followed by injury as part of a complex elbow fracture dislocation in 58.8% (n = 60) (Figure 3).

Figure 3 .

Bar chart demonstrating different indications of radial head arthroplasty as opposed to open reduction and internal fixation as per the survey responses

The number of fracture fragments was a factor in opting for RHA in 40.2% (n = 41) of responses where typically three or more fragments was the threshold in favour of RHA as opposed to ORIF. Among the respondents, 53.9% (n = 55) did not consider fracture fragment number in their clinical decision-making as regards mode of surgery. Only 26.5% (n = 27) of respondents stated that discontinuity of the radial head fragments with the shaft is an indication for RHA. The various RHA systems or procedures used are summarised in Table 1 by order of popularity of response.

Table 1 .

Summary of RHA systems used

Radial head arthroplasty system	n (%)
Anatomic metal arthroplasty	42 (41.2)
Circular monopolar metal	35 (34.3)
Circular bipolar metal	14 (13.7)
Pyrocarbon prosthesis	6 (5.9)

Only 29.4% (n = 30) of respondents would routinely immobilise post-surgery for a terrible elbow triad or when the injury was part of a more complex elbow dislocation. Some 70.6% of respondents (n = 72) reported no routine immobilisation regimen following surgery regardless of the severity of the initial injury.

Non-operative treatments

For minimally displaced RHF, as per responses to question 9, non-operative management with immediate mobilisation of the elbow was the most popular method of treatment in 90.2% (n = 92) of responses, as opposed to 9.8% (n = 10) who treated with initial immobilisation in plaster or a brace. Of those patients managed with immediate mobilisation, 65% (n = 60) of respondents would also commence physiotherapy at day 1 post-injury compared with 35% (n = 35) who would commence physiotherapy after at least 2 weeks of recuperation. Subsequent follow-up protocols also varied with 22.5% (n = 23) of respondents following up non-operatively managed patients for more than 6 weeks to review range of motion and clinical outcomes. Only 31.4% (n = 32) of respondents would routinely request radiographs after 1 week to assess for displacement.

Future research

The last three questions assessed the current gaps in the literature, as well as the willingness of our respondents to engage in a future RCT to help mitigate variations and disparity in the current practice. The research question most of the respondents (63.7%, n = 65) demonstrated interest in was for RHA vs excision for comminuted fractures without instability. This was closely followed by ORIF vs RHA in comminuted fractures when both are reasonable options (62.7%, n = 64), operative vs non-operative management of RHFs with >2mm displacement (53%, n = 54), and finally screw fixation vs plates when both are reasonable options (36.3%, n = 37). This highlights the paucity of high-quality evidence comparing surgical options for the management of RHF in the current literature. In addition to the demonstrated keen interest in engaging in a multicentric RCT, we also believe that our cohort would be highly reflective of the current practice in management of those common injuries, with 37.2% (n = 38) of respondents performing more than five RHA and 49% (n = 50) performing more than five RHF ORIFs per annum.

Discussion

RHFs are common elbow fractures but despite that there is a paucity of well-designed RCTs guiding their management. There is a consensus that a type I fracture, which is minimally displaced, lends itself well to non-operative management. In a prospective study of 103 Mason I RHFs managed non-operatively over a 1-year follow-up, all had excellent radiological and clinical outcome.⁴ One RCT including 98 participants compared early vs delayed (after 2 weeks) mobilisation of non-operatively managed RHF.⁵ Despite being the sole RCT to investigate non-operative options for RHF, it had the drawbacks of potential selection bias (no randomisation method was mentioned) and no information about the details of the intervention, such as exercise instructions and the level of supervision when mobilising the elbow, as well as an absence of any validated upper limb functional outcome score as a primary outcome.⁶ Another trial compared early vs delayed (after 5 days) mobilisation for type I RHF in 60 patients.⁷ It was found that early mobilisation led to lower pain and better elbow function at 1 week post-injury but similar outcomes at 4 weeks from injury. Despite addressing the critiques by integrating Broberg and Morrey functional score as a primary outcome, the downside is that this study did not follow any formal randomisation method as per the authors and that it was underpowered with a short duration of follow-up.

Others have recommended that repeat radiographs after initial presentation should be performed on clinical indication and are not routinely required given the exceedingly rare incidence of radiographic instability during follow-up (only 1 patient among a series of 237).⁴ The management of Mason II fractures lacks such consensus and is still debated with mixed results. The most agreed upon indication for surgical treatment of RHF is when concomitant with complex fracture dislocations or the presence of a mechanical block to motion, as also consistent with our survey.^8,9 However, this should be assessed carefully to exclude pain inhibition. In some cases, the mechanical block can be due to an interposed osteochondral lesion or even a cartilage flap not evident on the initial radiographs.¹⁰ Aside from the commonly cited 2mm displacement as an indicator for surgical management, other radiographic and intraoperative indicators have been subsequently proposed. Ring et al¹¹ reported unsatisfactory results in 93% of RHFs managed by ORIF when there were more than three fragments and 100% poor outcome in fracture dislocation injuries (type IV). Based on their data, Ring et al recommend that ORIF is better reserved for RHF which have minimal comminution with three or fewer fragments, and should be avoided in fracture dislocation-associated injuries. Rineer et al¹² reviewed 296 RHF and show that complete discontinuity of cortical contact with at least one fragment is associated with a more complex and unstable injury pattern by a factor of 21-fold.

A systematic review by Kaas et al¹³ on the operative vs non-operative management of Mason II RHF was able to identify only nine retrospective (level IV evidence) studies including 444 patients and found no firm evidence to back either management approach for such fractures. This systematic review called for a well-designed RCT for operative vs non-operative management of type II RHF and noted the paucity of high-quality evidence on this matter. In favour of non-operative treatment, Lindenhovius et al¹⁴ reported no appreciable advantage of operatively treated stable type II fractures over the published results for non-operative treatment after an average 22 years of follow-up. With only 19% of cases showing good long-term outcome in their series and a 44% complication rate, they advocated ORIF only for type II fractures that cause a mechanical block to forearm rotation. Herbertson et al¹⁵ reported 84% good outcomes in non-operatively managed type II fractures over a 19-year follow-up. Similarly, Akesson et al¹⁶ reported 82% good outcomes for non-operatively managed type II fractures after an average follow-up of 19 years.

Conversely, Khalfayan et al¹⁷ in a retrospective study including 26 type II fractures with a mean follow-up of 18 months, reported 90% good/excellent results for the ORIF group as opposed to only 44% with a similar outcome for the non-operative group. Their conclusion was that non-operatively managed type II fractures have a higher incidence of functional limitations and radiographic evidence of arthritis compared with fractures treated by ORIF. This was supported by other retrospective studies reporting 46% fair/poor outcomes following non-operative management of displaced RHF.¹⁸ A systematic review including 24 studies and 825 patients found that for Mason II fractures, non-operatively managed patients in comparison with the ORIF group had higher residual pain (42% compared with 32%, respectively) and lower proportion of good/excellent functional scores (52% compared with 88%, respectively).¹⁹

Only one RCT assessed different methods of fixation comparing biodegradable pins with metal screws in treating displaced RHF, and attained similar outcomes in terms of elbow function and adverse effects.²⁰ We could identify only one recently published RCT for type II RHF randomised to ORIF vs non-operative treatment in a sling (Radial Head – Amsterdam – Amphia – Boston – Others; The RAMBO trial). After 1-year follow-up, both patient groups had similar disability of arm, shoulder and hand (DASH) functional scores and complication rates.²¹ This trial only recruited 45 patients with a maximum follow-up of 1 year. Unlike Mason II fractures, Mason III/IV types have consistently been shown to fare better with operative management with limited studies supporting non-operative management.⁴

The decision whether to perform RHE (early or delayed) vs ORIF or RHA is multifaceted and the factors usually quoted are patient related, such as elderly or low activity levels, and fracture related, such as presence of mechanical block, degree of displacement, comminution, cortical contact and elbow stability. RHE has usually been reserved for isolated RHF without associated instability in low-demand elderly patients. Jansenn et al²² in a retrospective study on 21 patients with Mason III fractures treated acutely with RHE and with up to 30 years of follow-up found excellent results in 81% in terms of clinical, radiological and functional criteria. Broberg and Morrey³ reported good/excellent outcomes in 77% and functional improvement in 81% of their cohort after delayed RHE for comminuted fractures, but stressed reserving this as a salvage option. By contrast, Josefsson et al²³ reported an incidence of 63% severe osteoarthritis and recurrent instability after a follow-up of up to 34 years and recommended preserving the radial head whenever possible. This was supported by other authors reporting better function, strength and pain scores in favour of ORIF compared with radial head resection in Mason III cohort.²⁴

With the advent of RHA, RHE is gradually falling out of favour, and this is reflected in the national survey. The radial head in an important secondary valgus stabiliser and it has been shown in biomechanical studies that excision as opposed to RHA leads to an alteration in joint kinematics, load transfer and increased elbow instability.⁸ It has been suggested that RHA is superior to ORIF in the context of unstable elbow injuries. Watters et al²⁵ in their retrospective comparative study on 39 patients with terrible triad reported a higher rate of recurrent elbow instability in the ORIF group when compared with RHA at final follow-up (mean = 24 months). We could identify two RCTs that investigated complication rates and functional outcomes comparing ORIF vs RHA from multi-fragmented unstable RHF.^26,27 Both showed that significantly better functional outcomes and lower complication rates were attained in the RHA group. The caveat, however, is that they were significantly underpowered with only 67 sample sizes in both combined. There were also lack of reporting on randomisation used in these studies and assessor blinding.²⁸

As regards the different available prosthesis designs, a systematic review by Heijink et al²⁹ investigated different RHA prosthesis designs and identified 30 articles with 727 patients. The pooled outcome was good/excellent in 85% and it was found that revision rates and functional outcomes were not reliant on prothesis design such as polarity, material or fixation technique (cemented vs press-fit). The only salient finding was that silastic implants had consistently poor outcomes. Table 2 summarises the five RCTs identified in our literature review.

Table 2 .

Summary of identified randomised controlled trials

RCT	Population	Randomisation	Control	Intervention	Outcome measures	Results	Conclusion
Unsworth-White et al5	98 patients with non-operatively managed Mason I and II fractures	Not mentioned	Immediate mobilisation of the elbow post fracture (n = 29)	Immobilisation of the elbow in 90 degrees flexion (n = 29), vs immobilisation of the elbow in extension for 2 weeks (n = 23)	Functional outcomes as residual pain, range of elbow movements and patient’s satisfaction at an average 25 months	No significant difference between the three methods in terms of pain or ROM	No difference between immediate or delayed immobilisation for conservatively managed RHF
Helling et al20	164 operatively managed displaced RHFs	Telephone randomisation by an independent agency	61 RHFs treated with standard metal mini-fragment implants	74 RHFs treated with biodegradable polylactide pins	Complication rates, and Broberg and Morrey elbow score up to 2 years	Good or excellent results achieved by 92% of the control group and 96% of the polylactide group	Biodegradable polylactide pins have at least comparable outcomes as standard metal implants
Mulders et al21	45 patients with Mason II RHFs	Computer-generated block randomisation	Non-operative management by a pressure bandage (n = 22)	Operative management by screw fixation (n = 23)	DASH score up to 1 year	Both groups had similar functional outcomes and complication rates	Non-operatively treated isolated Mason II RHFs have similar functional outcomes
Ruan et al26	22 patients with Mason III RHFs	Not mentioned	ORIF (n = 8)	Cemented bipolar RHA (n = 14)	Broberg and Morrey score up to 15.9 months	Good or excellent results in 92.9% of RHA patients and in 12.5% of ORIF patients	RHA is superior in terms of functional outcomes compared with ORIF in Mason III RHFs
Chen et al27	45 patients with unstable, multi-fragmented RHFs	Block randomisation	ORIF (n = 23)	RHA using metal prothesis (n = 22)	Broberg and Morrey score and complications rate up to 2 years	91% good or excellent outcomes in RHA group compared with 65.2% in ORIF group	RHA is superior to ORIF in unstable multi-fragmentary RHFs

DASH = disability of arm shoulder and hand; ORIF = open reduction and internal fixation; RHA = radial head arthroplasty; RHF = radial head fractures; ROM = range of motion

Conclusion

This national survey highlights the lack of consensus in the management of RHF among an experienced cohort of participating orthopaedic surgeons who are regularly engaged in the management of these common injuries.

Few previous RCTs have sought to evidence best practice in the management of RHFs with the conclusion in the main being that further adequately powered RCTs are required. Of the five existing RCTs in the literature on the management of RHF, most have serious methodological weaknesses (Table 2). It comes as no surprise then that a recent survey by Mahmoud et al³⁰ on the management of RHF by 49 surgeons, found that only 4.1% of the treatment decisions were evidence based. There is therefore a need for high-quality robust RCTs focusing on this injury pattern with appropriate randomisation methods, sufficiently powered sample sizes and validated outcome measures to better guide future management of these common injuries.

Given this relative lack of quality evidence in the literature and in view of the interest declared by the participating surgeons in this survey, a future multicentre RCT would be well received and engender the requisite level of engagement to further our experience in management RHFs.

Author contributions

AB drafted the manuscript and performed the data analysis. CM revised the manuscript for submission. HS incepted the project and supervised survey approval and distribution.