Surgical management of radial head fractures

Objective: To compare the outcome of various surgical options exercised in the management of different types of radial head fractures. Method: A retrospective study of 29 patients with radial head fractures managed surgically was undertaken. Case notes were retrieved, and final assessment was performed through a telephone questionnaire. The results were analysed according to the classification of Wesley et al (1983), and compared with the patients’ own rating. Results: The best results were obtained in Mason type II fractures, followed by type III and type IV fractures. Comparing different operations, the best outcome was observed with screw fixation, followed by excision of the radial head, Kirschner wire fixation, partial excision, silastic implant, and plating, in that order. Conclusion: The less comminuted a radial head fracture is, the better the outcome. Screw fixation is to be preferred, if technically possible. Our method of outcome assessment closely reflected in patient satisfaction.

Keywords: functional assessment, internal fixation, prosthesis, radial head fractures, surgical management

Fractures of the radial head are the most common elbow fractures in adults, and can result in major disability.¹ They account for 17% to 44% of elbow injuries.^2,3 The radial head has an incompletely understood mechanical role in the forearm, and presents a special technical challenge to fix securely.⁵ Because of increased recognition of the problems after radial head excision, there is more emphasis on its preservation after fracture if technically feasible.⁶ The decision to excise or internally fix the radial head fracture is influenced by the presence of concomitant injuries and their severity.

Biomechanical studies have not shown improved stability or reduction of radial migration after trauma with the use of silastic implants, but an implant may give some support while soft tissue healing occurs.⁷ Despite the use of a prosthesis, in some patients more than four millimeters of proximal migration of the radius has been demonstrated.⁸ With increasing use of silastic prostheses, more reports of material failure, dislocation, and particulate synovitis have been published.^9,10

Elbow pain, cubitus valgus, subluxation of the distal radio-ulnar joint with wrist pain, and loss of strength have all been described as complications of radial head fractures.^11,12Capitellar rarifaction has been observed, and attributed to the lack of force transmission across the radiocapitellar joint.¹³

Other uncommon early complications include compartment syndrome in the forearm and posterior interosseous nerve palsy.^14,15 Arthroscopic debridement for arthrofibrosis may be considered in Mason type I fractures, if the range of motion has plateaued with significant residual loss of mobility after aggressive physio-therapy.¹⁶

Thus, there are numerous options available in the management of these fractures, a testimony to the diversity of options regarding the best method of management. Conservative management is uniformly accepted for Mason type I fractures. Although most authors agree that Mason type II fractures, with the fracture fragment involving more than 25% of radial head, are best treated by reduction and fixation, opinions differ as to the best method of fixation. Management of Mason type III fractures is equally controversial, with management options including conservative management, simple excision, excision and prosthetic implant replacement, and open reduction and internal fixation (ORIF).

In this study, we assessed the outcome of the surgical management of radial head fractures, comparing results of various surgical procedures according to the different Mason types of fractures in a large teaching hospital.

The 45 patients with radial head fractures undergoing surgery at Aberdeen Royal Infirmary, Aberdeen, from January 1985 to December 1994, were eligible for this study. Four of these patients died, and two patients were overseas tourists, who left the country after their initial treatment. One patient, who sustained an Essex-Lopresti injury, had radial shortening with marked ipsilateral wrist joint pain, was still under treatment with the Ilizarov technique for radial lengthening and, therefore, was excluded from the study. Of the remaining 38 patients, 29 patients were contactable, the remaining nine patients being lost from follow-up. Hence, the present study comprises 74% of the original patient population. The assessment was performed according to the recommendations of the Ethical Committee through case notes review, review of the available radiographs and telephone questionnaires to the patients.

The male to female ratio was 16:13, and the age ranged from 17 to 83 years with a peak incidence of between 20 and 50 years. The mean age was 42 and 39 years in female and male patients, respectively. High-energy injury was responsible for the fracture in 62% of patients. All patients were assessed clinically and radiographically by an independent assessor who had not taken part in the original management of the patients, and who was blinded as to the types of fracture and the operative management performed, with at least antero-posterior (AP) and lateral radiographs of the involved elbow. Twenty-seven fractures were closed (six type II, 15 type III and two type IV), and three were open (one Mason type III and two type IV). Aspiration of the elbow joint was performed in four patients in whom haemarthrosis was severe enough to cause significant discomfort. Total length of active out-patient follow-up ranged from 4 to 152 weeks, with a mean of 34 weeks. Patients were reviewed at an average of 6 years from the fracture, with a minimum follow-up of 4 years from the injury. All the patients treated by internal fixation of the radial head fractures were followed until bony union was complete. The authors examined all the radiographs of the injured elbow, and the fractures were classified according to the modified Mason classification, which includes a further class, type IV. The operations performed were Kirschner wires - two patients with Mason type II and III; screw fixation - two in type II and one with type III; plating - one each in type II and type III; silastic implant - five for type III; simple excision -six for type III and three unclassified old fractures, and partial excision - one each for type II and III and three in type IV fractures. Two patients had supplementary fixation of the distal radio-ulnar joint with temporary percutaneous pins.

The outcome of the operative management of these fractures was assessed through a telephone questionnaire, which included questions on pain in the operated elbow or ipsilateral wrist joint, and range of motion in the operated elbow. The enquiry also included the functional ability of the operated limb. Strength and range of motion of the operated elbow joint were assessed by asking the patient to compare them with that of the unoperated elbow. Functional assessment included ability to perform activities of daily living, and whether patients had returned to work and recreational activities. Patients were also asked to comment on the final outcome of their operation by rating it as excellent, good, fair, or poor. The results were classified according to Weseley et al (1983) (Appendix I).¹⁷A comparison of these results with that of patients’ own rating of the outcome was made. Analytical statistics could not be carried out given the small number of patients in each category.

The majority (20 patients, 69%) initially presented to the Accident and Emergency Department, while other patients were admitted either through the Fracture Clinic or on referral from a General Practitioner, and one patient was transferred from a nearby hospital.

Most (90%) of the patients were right handed. The prevalence of involvement of the dominant and the nondominant side was approximately the same (16 and 13 patients, respectively).

In addition to the radial head fracture, nine patients had other associated injuries involving the ipsilateral elbow or wrist joint.

Most of the operations (72%) were performed as primary procedures. Five operations were secondary procedures performed after initial conservative management for 2 to 4 weeks. Secondary excision of the radial head in three patients was performed for old fractures. The time lag between admission and operation varied between 12 hours to 7 days, with mean of 1.5 days. Hospital stay ranged from one to 10 days, with a mean stay of 2 days.

At final follow up, 13 patients (44%) had no pain in either of the joints. Thirteen patients had mild pain in the operated elbow, and two of these patients also had mild pain in the ipsilateral wrist joint. Two patients had moderate pain in the elbow, and they also had mild pain in the ipsilateral wrist joint. None of the patients in the present study had severe pain in either of the joints. Therefore, 16 patients (56%) had mild or moderate pain in one or both joints. The average loss of motion for all the procedures was extension 10°, flexion 5°, supination 14° and pronation 5°. Strength was equal to the opposite hand in 18 patients (62%), mild loss of grip was reported by nine patients (31%), and two patients had moderate loss. Thus, 11 patients (40%) had mild or moderate loss of grip. None of the patients reported severe loss of grip strength.

The least restriction of motion in Mason type II fractures was seen with partial excision, followed by screw fixation (Figures 1 and 2), Kirschner wire, and plating in that order (Table 1), in Mason type III fractures with Kirschner wire, followed by total excision, screw fixation, partial excision, silastic prostheses (Figures 3 and 4), and plating (Figures 5 and 6) in that order (Table 2). In Mason type IV fractures, partial excision resulted in better motion than total excision (Table 3). The majority (20 patients, 69%) had no limitation of their activities of daily living. Eight patients had minor symptoms such as occasional twinges or tiredness of the involved upper limb during activities such as dressing, washing, shaving, knitting, opening of jars, etc. Five patients noticed that they used the uninjured arm more often than before the injury due to pain in performing certain activities such as turning a screw, opening a jar, lifting weights, etc.

Figure 1: (a) Antero-posterior and (b) lateral views of a Mason type II fracture in a 35-year-old male

Figure 2: Same patient as in Figure 1. (a) Antero-posterior and (b) lateral views of a Mason type II fracture managed by screw fixation 6 years after fixation

Figure 3: (a) Antero-posterior and (b) lateral views of a Mason type III fracture in a 21-year- old female

Figure 4: Same patient as in Figure 3. (a) Antero-posterior and (b) lateral views of a Mason type III fracture managed by excision silastic prosthesis 51 months after the injury

Figure 5: (a) Antero-posterior and (b) lateral views of a Mason type II fracture in a 31-year-old male

Figure 6: Same patient as in Figure 5. (a) Antero-posterior and (b) lateral views of a Mason type III fracture managed by plate and screws fixation 4 years after fixation.

Procedure	Patient numbers	Average loss of motion (degrees)				Pain	Loss of strength (%)	Return to work (weeks)	Results (E+G+F) (%)
Procedure	Patient numbers	Ext	Fle	Sup	Pro	Pain	Loss of strength (%)	Return to work (weeks)	Results (E+G+F) (%)
Kirshner wire	2	12.5	0	12.5	0	50%	0	1	100
Screws	2	5	5	0	0	50%	0		100
Partial excision	1	0	0	0	0	0%	0	7	100
Plate	1	20	0	20	0	Mild	100	26	100
Average (for all procedures)	6	9.6	2	8	2	33%	17	8.6	100

E-excellent; G-good; F - fair; P - poor; Ave - average; Ext - extension, Fle - flexion; Sup - supination and Propronation

Procedure	Patient numbers	Average loss of motion (degrees)	Pain	Loss of strength (%)	Return to work (weeks)	Results (E+G+F) (%)
Ext	Fle	Sup	Pro
Silastic	5	10	8	17	5	80%	60	5.6	80
Excision	6	8	4	8	2	43%	14	6	100
Kirshner wire	2	10	3	0	3	50%	0	8	100
Screw	1	15	0	20	0	Mild+W	Mild	8	100
Plate	1	0	20	30	0	Mod	Mod	26	0*
Partial excision	1	25	10	45	0	Mild	Mod	16	100
Average (for all procedures)	16	10	7	14.3		64%	41	8.3	88

* Removed; W Wrist Joint; Results: E-Excellent; G-Good; F-Fair; P-Poor; Ext- extension; Fle- flexion; Sup- supination and Pro- pronation

Procedure	Patient numbers	Average loss of motion (degrees)	Pain (%)	Loss of strength (%)	Return to work (weeks)	Results (E+G+F) (%)
Ext	Fle	Sup	Pro
Partial excision	3	8	3	13	13	33	0	13	66
Excision	1	30	0	30	0	0	Mild	52	100
Average (for both procedures)	4	14	3	18	10	25	25	26	75

E-Excellent; G-Good; F-Fair; Ave- average; Ext- extension; Fle- flexion; Sup- supination and Pro-pronation

The time period for return to work varied from one week to one year. Twenty-six patients (90%) returned to their previous job. One patient, with an associated open fracture of the ipsilateral ulna, returned to his plumbing job after one year. Three patients changed their jobs following their fracture due to persisting elbow pain. Seven patients returned to their leisure activities in less than 6 weeks, seven patients between 6 to 12 weeks, six patients between 13-26 weeks, nine patients between 27-52 weeks. One patient, whose sporting activities were karate and golf, reported limited activity due to pain. Four patients could not pursue their usual leisure activities due to pain in the elbow joint. Their original leisure activities included motor cycle riding, aerobics, bicycle riding, and push-ups. These patients had to switch over to activities such as swimming and jogging.

There was a wide variation in the time period to achieve maximum benefit from the operation. This ranged from 4 weeks to 15 months. Ten patients achieved it in 3 months, eight patients in 4-6 months, ten patients in 7-12 months, and one patient took 15 months. This last patient had an associated open fracture of the ipsilateral olecranon.

Two patients had radial nerve involvement. One patient had a radial nerve palsy from the initial injury, while the other sustained a radial nerve injury at the time of operation. Both patients recovered well, although the latter patient had some persisting altered sensation. One patient who had a Kirschner wire fixation had superficial infection, which settled after its removal.

Mason type	Patient numbers	Average loss of motion (degrees)	Pain (%)	Loss of strength (%)	Return to work (weeks)	Results (E+G+F) (%)
Ext	Fle	Sup	Pro
Type II	6	10	2	8	2	33	17	8.6	100
Type III	16	10	7	14	3	65	41	8.4	88
Type IV	4	14	3	18	10	25	25	26	75
Unclassified	3	10	5	25	12	0	33	3.3	100
Average	29	11	4	16	7	31	29	11.5	91

E-Excellent; G-Good; F-Fair; Ave- average; Ext- extension; Fle- flexion; Sup- supination and Pro-pronation

Procedure	Patient numbers	Average loss of motion (degrees)	Pain	Loss of strength (%)	Return to work (weeks)	Results (E+G+F) (%)
Ext	Fle	Sup	Pro
Excision	10	8	5	13	4	30	13	13.5	100
Silastic	5	5	8	13	6	80	60	7	80
Kirshner wire	4	11	0	13	1	50	0	4.6	75
Screw	3	10	2.5	10	0	33	3	5.6	100
Plate	2	10	10	2.5	25	100	7	26	50
Partial excision	5	10	4	17	8	40	3	12	80
Average	29	9	5	11	7	56	38	11	81

Mason type II fractures of the radial head are the most controversial, as regards management. In displaced fractures, there is disagreement over the indications for conservative and operative treatment. Early excision of the radial head was favoured by Watson-Jones, whereas Charnley advised delayed excision, to be performed only if there was significant limitation of rotation at the end of the second week after injury.^{18, 19} Muller et al introduced the concept of open reduction and internal fixation of displaced wedge fractures of the radial head.²⁰For this group of fractures, Mason suggested that ‘If in doubt, resect’.⁴ By contrast, Johnston recommended ‘If in doubt, conserve’.¹

Type III Mason fractures result from a severe injury, and are accompanied by more soft tissue damage than type II fractures.²¹ Excision of the radial head for comminuted fractures is widely practised.^{4, 7, 22, 23} In the present study, total excision of the radial head was the most commonly employed surgical option for type III fractures, and it had the best outcome. Comparative results of implant arthroplasty in various studies were not uniformly good. Morrey et al, from a study of 17 patients who had silastic prosthesis with a follow up of 6.7 years, reported five failures, four of which necessitated removal of the prosthesis.²⁵Three other patients had radiographic failures. They concluded that the indications for use of the silastic radial head prosthesis after fracture of the radial head were extremely limited, and that its routine use cannot be justified. Of our patients who had silastic implants, four complained of pain, three of loss of strength, and one reported a poor outcome. All but two patients treated for type III fractures returned to their previous occupation, and two other patients had to change their recreational activities. The highest incidence of pain (19 patients, 65%) and loss of strength (12 patients, 41%) were both observed in this group, probably reflecting the fact that the majority of the patients occupation involved strenuous activities like manual labour, fishing, etc.

Dislocation of the humeroulnar joint with fracture of the radial head is a complex injury that involves soft tissues, articular cartilage and bone. This combined lesion has been reported to have a worse prognosis than either of the single injuries.^1,25 McLaughlin emphasised the need for early excision, and considered type IV radial head fractures an ‘orthopaedic emergency, to be measured in a matter of hours.’²⁶ Alder and Shaftan expressed the opposite view, and felt that it is better to treat these injuries non- operatively with early mobility rather than to introduce further trauma through surgery.²⁵ They recommended surgery only when there was an obvious mechanical block to mobility, and that, when indicated, surgery be undertaken within 24 hours after the injury.

In the present study, partial excision of the radial head was the most commonly employed option for these fractures, while the remaining patient was managed by total excision of the radial head. Our good results could partly be due to the fact that the majority of the patients did not have severe radial head injury, and these patients were managed by closed reduction of the elbow joint and partial excision of the radial head, without compromising the stability of the elbow joint.

Our results compare well with published results, although the time to return to work was higher, with an average of 11.4 weeks. This could partly be due to the fact that a larger proportion of our patients were undertaking heavy work (manual labour, fishing, working on oil rigs, etc). Our rating was exactly the same as the patients’ own rating in 18 instances (60%). Only six patients (20%) felt that their out-come was worse than our assessment. In all, the 12 patients in whom the gradings differed, the difference was only by one grade, except for one patient, in whom the difference was by two grades. Our method of outcome assessment, therefore, closely reflects the patients’ point of view, although it cannot easily be compared with other studies.

A number of confounding variables, including differences in outcome instruments, fracture classification schemes, treatment regimens, patient co-morbid conditions, the severity of injury, and the skill of the surgeon, make it difficult to compare published studies. Numerous self-administered questionnaires quantitatively measure function and quality of life, and have been validated for research purposes.^{1,27,28,29,30,31} However, none of these studies measured range of motion through a questionnaire. Our method of measurement of motion by telephone interview was not fully objective and, hence, could have influenced the final grading of the results. Thus, given these limitations, comparison with other studies is not entirely appropriate. Although this method may not give accurate measurements of these variables in absolute terms, it enabled us to have a good assessment of the function of the elbow, which is an important issue.³² Impaired physical, social and emotional function is an important end-result of all acute and chronic conditions.³¹ This mode of patient contact gave us a response rate of 74%. This was a good response considering that the study spanned 11 years, that the catchment area was vast, and included the Shetland and Orkney Islands, and that nearly 30% of the patients were above 50 years of age. In comparison with postal questionnaires, telephone questionnaires provide an opportunity to discuss and clarify difficult aspects, and it also helped us to cover a greater number of issues than would have been feasible by a postal questionnaire.

The present study has several limitations. For example, it is a retrospective review, based in a single teaching centre where several orthopaedic consultants applied their own beliefs to the management of the condition, for which there was no established protocol. Also, the use of a telephone questionnaire can only give a partial view to the problem, as some categories, such as joint motion, are notoriously difficult to assess even in a hospital or office setting. However, we were able to trace a significant proportion of the eligible patients, and a single trained researcher, not involved in the initial management of these patients, administered all the telephone questionnaires, thus, minimising interviewer bias and variability.

In conclusion, we observed our best results in Mason type II injuries, followed by type III, and type IV, respectively. This confirms that the increasing grades of fractures are associated with increased soft tissue, articular cartilage and bone damage. We found residual pain in the injured elbows the commonest complaint in these patients, as in other studies. Our method of outcome assessment closely reflected the patients’own assessment.

Appendix I: Weseley et al Classification for outcome assessment (J. Trauma, 1983)

Results	Criteria
Excellent (E)	Loss of less than five degrees of combined flexion-extension No loss of pronation-supination No pain
Good (G)	Less than 200 loss of combined flexion-extension Minimum loss of pronation-supination Mild pain (normal activity, no medication)
Fair (F)	At least 450 of useful flexion-extension or Minimum of 50% pronation-supination Moderate pain (with or after activity)
Poor (P)	Less than 450 of useful flexion-extension Less than 50% of pronation-supination Severe pain (at rest, constant medication)

We thank Mr GP Ashcroft, Mr RB Chesney, Mr J Gibson, Mr PH Gibson, Prof JD Hutchison, Mr A Johnstone, Mr D Knight, Mr WM Ledingham, Mr K Mills, Mr TR Scotland, Mr D Wardlaw, and the other consultants in the Department of Orthopaedic Surgery, Aberdeen Royal Hospitals Trust, who allowed us to include their patients in this study.

Many thanks are due to Miss Linda Lothian for her help given in typing this manuscript.

Correspondence: Mr Nicola Maffulli, Department of Orthopaedic Surgery, University of Aberdeen Medical School, Polwarth Building, Foresterhill, Aberdeen, AB25 2ZD U.K.

ORIGINAL ARTICLES

Surgical management of radial head fractures