Indications and outcomes of radial head excision: A systematic review

Abstract

Background

Radial head excision has historically been a common surgical procedure for the operative management of radial head fractures and post-traumatic conditions. With recent advances in other surgical techniques, controversy exists regarding its indications. This review evaluates the indications and outcomes of radial head excision in traumatic and non-traumatic elbow pathology.

Methods

Multiple databases were searched for studies involving radial head excision. Screening and data abstraction were conducted in duplicate. Only studies reporting outcomes for radial head excision were included.

Results

Twenty-seven studies with 774 radial head excision patients were included. The most common indications involved acute excision of comminuted radial head fractures (n = 347) and rheumatoid arthritis (n = 201). Post-operative functional scores after acute excision were reported to be good to excellent. In the chronic setting of rheumatoid disease, radial head excision resulted in improved range of motion, although pain was not effectively relieved.

Discussion

Outcomes of radial head excision for acute fracture are good to excellent; however, it should not be performed when concurrent or ligamentous injuries are present. Although some studies compared excision to open reduction and internal fixation or replacement, more data are needed to make proper conclusions. The strength of these conclusions is limited by the quality of included literature.

Introduction

The radial head plays an important role in the biomechanics of the elbow joint. The articulation between the head of the radius and the capitellum stabilizes axial and valgus loading of the forearm and elbow, preventing dislocation and excessive valgus displacement.^1–4 When the radial head is excised, the biomechanics of the elbow are affected. It is well established that radial head excision results in an increase in valgus laxity of the elbow and proximal migration of the radius.^1,3,5 However, these findings do not necessarily correlate with functional results.⁴ Furthermore, with radial head excision, the loads originally borne by the radiocapitellar joint are transferred to the ulnohumeral articulation, increasing the forces across that joint, and ultimately leading to early osteoarthritis.^3,6

The radial head has a close relationship with the ligaments of the elbow and forearm; particularly, it is an important secondary constraint to the medial collateral ligament (MCL).^3,4 With ligamentous injury, excision of the radial head can result in significant instability. Concomitant MCL disruption or combined MCL and lateral ulnar collateral ligament injuries may result in valgus instability or posterior elbow dislocation.^1,7–9 Moreover, radial head excision in the setting of injury to the central band of the interosseous membrane can result in instability, longitudinal radioulnar dissociation, proximal migration of the radius, and ulnar impaction at the wrist.¹⁰ Thus, radial head excision is contraindicated in the setting of complex elbow instability, which may necessitate radial head arthroplasty to preserve both acute and long-term stability.^2,3,11

Indications for radial head excision are controversial. Excision without replacement has been advocated for isolated unreconstructable radial head fractures when the degree of comminution prevents internal fixation, especially in those with low demand of their upper extremities.^2,3 For example, Radin et al. suggested that displaced comminuted fractures involving more than two thirds of the radial head warrant radial head excision.¹²

Radial head excision is not only used acutely to treat elbow fractures but also in the chronic setting for rheumatoid and post-traumatic arthritis.¹³ Indeed, Smith- Petersen et al. suggest that radial head excision can be used to relieve any condition interfering with normal function between the radius and the ulna or capitellum of the humerus.¹³ The removal of the radial head can restore motion not only by relieving pain but also by removing a mechanical block caused by a deformed radial head.¹⁴

Therefore, the objective of this systematic review is to evaluate the indications and outcomes of radial head excision in patients with a variety of elbow pathologies.

Materials and methods

This systematic review was conducted according to the methods of the Cochrane Handbook for Systematic Reviews of Interventions.¹⁵

Search strategy and eligibility

EMBASE, MEDLINE, PubMed, Cochrane Library and CINAHL were searched for literature related to radial head excision from database inception to 13 January 2019. The search strategy was based on the research question and eligibility criteria which were established a priori. Manual review of the reference lists of key papers was performed. The search terms used were: “excision”, “resection”, “radi*”, “head”, “neck”, and “fracture” (online Appendix Table I). MeSH and EMTREE search terms were utilized in various combinations to increase search sensitivity.

The inclusion criteria were (1) studies reporting any outcome-related information on patients who had undergone radial head excision with minimum two-year follow-up, (2) patients above the age of 18, (3) studies published in English, and (4) studies on humans. Exclusion criteria were any nonsurgical treatment studies, studies including patients less than 18 years of age, studies with less than two years’ follow-up, and studies in which outcomes for patients undergoing radial head excision could not be separated from other patient outcomes. Case reports, editorials, reviews, and case series with less than five patients were excluded.

Study screening

Screening was performed according to PRISMA (Preferred Reported Items for Systematic Reviews and Meta-analyses), by two reviewers (AH, BZ) who independently screened the titles, abstracts, and full texts of the retrieved studies in duplicate. Discrepancies at any stage were resolved by consensus between the two reviewers. Duplicate articles were manually excluded.

Data extraction

The two reviewers independently abstracted relevant study data from the final included articles and recorded these data in a piloted Excel spreadsheet. Items on the form included source information (Study ID, author, year of publication, study design, and level of evidence), patient demographics (i.e. age, sex, clinical condition, etc.), and intervention information (i.e. number of intervention groups, details of procedure, etc.). Outcome information including Visual analogue scale (VAS) pain scores,¹⁶ functional outcome scores (Mayo elbow performance scores (MEPS),¹⁷ Disabilities of the Arm, Shoulder and Hand (DASH) scores,¹⁸ Broberg-Morrey functional rating scores,¹⁹ and range of motion. Complications and the number of patients requiring further surgery were also abstracted.

Assessment of risk of bias

The same two reviewers independently performed duplicate assessment of risk of bias of included studies using the Methodological Index for Non-Randomized Studies (MINORS) instrument for risk-of-bias assessment for non-randomized studies.²⁰ The maximum MINORS score is 16 for non-comparative studies and 24 for comparative studies.²⁰ Any discrepancies between reviewers were resolved by consensus.

Statistical analysis

Intraobserver agreement with respect to article eligibility was calculated with the Cohen κ coefficient. An intraclass correlation coefficient (ICC) was calculated to evaluate interobserver agreement for assessments of risk of bias. The κ and intraclass correlation coefficients were calculated using SPSS software (version 25.0). Descriptive statistics and measures of variance are presented where appropriate.

Results

Study identification

The initial electronic literature search yielded 1526 studies, of which 23 met the inclusion criteria for this review (Figure 1). Hand search of references yielded four additional studies to be included. There was substantial agreement among the two independent reviewers at the title and abstract (κ = 0.73; 95% CI 0.67–0.78) and moderate to substantial agreement at the full-text (κ = 0.61; 95% CI 0.49–0.74) screening stages.

Figure 1.

Flowchart of study selection.

Study characteristics

The characteristics of included studies can be found in online Appendix Table II. The studies were published between 1964 and 2018 and included a total of 1113 patients, with 774 patients who received 782 radial head excisions. Of these, 476 (62%) were primary acute traumatic excisions, 97 (13%) were secondary excisions after trauma, and 201 (26%) were excisions for rheumatoid arthritis. The mean sample size of the included studies was 41 patients (range, 5–151 patients). The patients treated across the studies were a mean age of 42.8 years (range, 18–89 years) with a mean follow-up of 13.6 years (range, 2–46 years). Of the 27 studies included, 19 were case series, and 8 were cohort studies (2 prospective cohort studies). Studies included patients with radial head or neck fracture (73%), fracture-dislocation of the elbow (2%), Monteggia fracture (4%), or rheumatoid arthritis (21%). The MINORs scores ranged from 6 to 20. Agreement between the two reviewers assessing risk of bias was high (intraclass correlation coefficient 0.92, 95% CI 0.81–0.96).

Indications

The indication for radial head excision was identifiable for 727 patients (97% of total radial head excision patients) from 25 of 27 studies. The most common indication for acute radial head excision was isolated comminuted radial head fracture (n = 347) followed by comminuted fracture with elbow dislocation (n = 84). The most common indication reported for radial head excision in the chronic setting was rheumatoid arthritis (n = 201).

In studies utilizing the Mason classification for radial head fractures treated with excision (n = 547), Mason type III fractures were the most common (52%), with Mason type II being the second most common (35%), and Mason type IV the third most common (7%). Six percent of patients who underwent radial head excision had fractures classified as Mason type I. In studies reporting whether patients had associated lesions (n = 713), isolated fractures were the most common (66%), with associated elbow dislocation being the second most common (20%) and Monteggia lesions being the third most common (3%). Other associated injuries present included concomitant ligament injury (2.7%), coronoid fracture (2.8%), olecranon fracture (0.7%), terrible triad (0.3%), humeral fracture (0.4%), capitellar fracture (0.1%), proximal ulnar fracture (0.1%), and other osseus injuries near the radial head (5%).

Outcomes

Radial head excision for acute fracture

Twenty-three of 27 studies in this review included patients who underwent acute radial head excision for a traumatic injury, most commonly a comminuted radial head fracture. These studies included a total of 476 primary radial head excision patients, or 62% of the total radial head excision patients included in this review. The clinical conditions of all patients who underwent acute excision can be found in Table 1.

Table 1.

Clinical condition of patients undergoing acute excision.

Condition	n (%)
Isolated radial head or neck fracture	405 (85)
Fracture-dislocation of the elbow	28 (6)
Monteggia lesions	7 (1)
Radial head fracture with ligamentous injury	19 (4)
Radial head fracture with other associated lesions	17 (4)
Total	476

Pain

Seventeen of 27 studies assessed pain post-operatively. The follow-up for these studies ranged from 4.6 to 21 years. Of these, five studies (n = 121 radial head excision patients) included the assessment of post-operative VAS pain scores after radial head excision. The mean VAS scores ranged from 0.2 to 19.3. Five studies (n = 118 radial head excision patients) assessed pain based on a 4 level rating scale (none, mild, moderate, or severe). Of the radial head excision patients in these studies, 89% rated their pain level as “none” or “mild”, and 11% rated their pain as “moderate” or “severe”. Eight of the 17 studies that assessed post-operative pain did not use a standardized scoring instrument, but described pain in subjective terms, with most finding that patients had minor, mild, or occasional pain.

Function

Five of 27 studies (n = 81) included post-operative assessment of the MEPS. The follow-up for these studies ranged from 16.9 to 21 years. The post-operative MEPS scores ranged from 86 to 96, demonstrating an excellent score. For studies reporting individual patient results for MEPS score (n = 40 patients), 98% had a good or excellent score, when outcomes were rated excellent if their MEPS score was 90–100, good if their score was 75–89, fair if 60–74, and poor if below 60.

Six of 27 studies (n = 128) included post-operative assessment of the DASH score. The follow-up of these studies ranged from 13.1 to 18 years. DASH scores ranged from 5 to 26, demonstrating very low disability.

Three of 27 studies (n = 81) included post-operative assessment of the Broberg and Morrey functional rating score. The follow-up for these studies ranged from 13.1 to 18 years. Scores ranged from 86 to 91, suggesting good outcomes. For studies reporting individual patient results for Broberg and Morrey scores (n = 42 patients), 95% had a good or excellent score, where outcomes are rated excellent if the score is 95–100 points, good if 80–94 points, fair if 60–79 points, and poor if less than 60 points.

Range of motion

Post-operative elbow flexion was assessed in 9 of 27 studies and found that mean post-operative flexion ranged from 112° to 140°. Six studies assessed post-operative extension in radial head excision patients, which ranged from −14° to 13°. Nine studies reported post-operative pronation and supination, with mean pronation ranging from 41° to 86°, and mean supination ranging from 44° to 85°. The follow-up of these studies ranged from 4.6 to 20.3 years.

Comparative studies

Excision vs. open reduction and internal fixation

Two retrospective cohort studies compared excision vs. open reduction and internal fixation (ORIF) for the treatment of radial head fractures (excision n = 39, ORIF n = 48).^21,22 Lindenhovius et al. (excision n = 15, ORIF n = 13) found no significant difference between the two interventions after evaluating post-operative pain, MEPS, Broberg Morrey functional rating score, American Shoulder and Elbow Surgeons (ASES) Elbow Evaluation Instrument, DASH score, range of motion, or radiographic changes in patients with unstable, displaced radial head fractures.²¹ Twenty of these fractures were Mason type III, 8 were mason type II, and a complex injury pattern was present in 15 of these patients.

Zarattini et al. (excision n = 24, ORIF n = 35) found statistically significant differences in various outcomes in favor of ORIF over radial head excision.²² More specifically, this study assessed patients with isolated Mason Type II radial head fractures and found that forearm rotation arc, grip strength, and Broberg and Morrey functional scores were significantly better for ORIF patients compared to excision patients (p < 0.001, p = 0.015, and p < 0.001, respectively). This study concluded that patients treated by ORIF had less residual pain, greater range of motion, better strength, improved functional scores, and lower incidence of severe posttraumatic arthritis compared to radial head excision patients.²²

Excision vs. replacement

Only one retrospective cohort study compared radial head excision to radial head arthroplasty (excision n = 50, replacement n = 28). Ünlü et al. found no significant differences between radial head excision and replacement when DASH, Mayo Wrist, and Oxford Elbow scoring systems were used to evaluate functional status. However, they did find that extension in the affected extremity was significantly more restricted in the replacement group (p = 0.014).²³ This review identified no other studies comparing excision and replacement in patients over 18 with at least two years’ follow-up.

Radial head excision in the chronic setting

Secondary excision after trauma

Two studies included outcomes for patients who underwent secondary or delayed radial head excision after traumatic injuries (n = 94). The average time between excision and initial injury ranged from 2 to 6 months, with mean follow-up ranging from 6 to 18 years. Initial injury included isolated comminuted radial head fracture (n = 68) and comminuted radial head fracture combined with elbow dislocation (n = 24). Only one study reported the reason for secondary excision, the most common reason being excision due to pain (n = 2).

Pain

The two studies that evaluated outcomes for secondary excision reported pain using a variety of outcome measures. Fuchs et al. (n = 92) found that only 41% of patients treated with secondary excision referred to their pain as minor, with the other 59% rating their pain as moderate (32%) or severe (27%).⁷ Herbertsson et al. assessed pain in a subjective manner and found that pain was present in one of two patients who underwent delayed radial head excision.²⁴

Function

In terms of functional outcomes, Fuchs et al. (n = 92) found that 30% of secondary excision patients had to change jobs, 62% had to give up sports activities, and only 44% were satisfied with their procedure.⁷ Furthermore, 73% of secondary excision patients had loss of strength post-operatively.⁷ In activities of daily living, 21% of secondary excision patients had some impairment.⁷ Of the two secondary excision patients in a study by Herbertsson et al., one patient had slight impairment of the injured elbow, reduced range of movement, a flexion deficit of 10°, a valgus increase of 5°, and an MEPS of 85, whereas the other patient had none of these deficits and an MEPS of 100.²⁴

Comparative studies

Primary vs. delayed radial head excision

Fuchs et al. (n = 151) compared primary excision (n = 59) to early and late secondary excision (n = 47 and 45, respectively). Primary excision was defined as within two weeks of initial trauma, early secondary as between the third week and six month after initial trauma, and late secondary as more than six months after the initial trauma.⁷ This study, while not reporting the statistical significance of results, found that the best results occurred after primary resection.⁷ More specifically, 69% of primary excision patients were satisfied with the procedure compared to only 44% of secondary excision patients. Furthermore, primary excision patients had less post-operative pain, loss of strength, restriction in daily activities, and work restrictions. In addition, when this study compared early vs. late secondary excision, they found that early secondary excision had worse results than late secondary excision, with more pain in the elbow and wrist joint, and more restriction in activities of daily living.⁷

Radial head excision for rheumatoid arthritis

Four of 27 studies in this review reported outcomes for patients who underwent radial head excision secondary to rheumatoid arthritis.^25–28 These studies included a total of 201 patients (209 elbows), or 26% of radial head excision patients included in this review. The follow-up of these studies ranged from 2 to 15 years.

Pain

In three studies reporting pain outcomes (n = 93), 53 patients (57%) had relief from pain after excision.^26–28 However, for the remaining 43% of patients, pain remained unchanged or worsened. In two studies that carried out a survival analysis of radial head excision and synovectomy for rheumatoid elbows (n = 140), 30 elbows (21%) were deemed as failures due to pain, 6 (4%) due to desire for further surgery, 21 (15%) for both these reasons, and 9 (6%) due to revision surgery for unclear reasons.^25,28 Woods et al. (n = 45 elbows) found that the cumulative survival rate at 10 years was 69% for radial head excision synovectomy.²⁸ Gendi et al. (n = 95) found that the overall failure rate was 46% during an average follow-up period of 6.5 years.²⁵

Function

Woods et al. (n = 45) assessed function according to three tests (hand to mouth, hand to back of head, and turn door handle) and found that performance improved in all but one excision patient.²⁸ Similarly, Herold et al. (n = 11) found that all but one patient were pleased with the result of their operation.²⁶ However, Gendi et al. found that short-term results of excision and synovectomy for rheumatoid arthritis were good, but that the long-term outcomes were generally poor.²⁵

Range of motion

Range of motion improved from pre-operative to post-operative measurement in all four studies. Gendi et al. found this improvement to be significant for both supination-pronation (p = 0.0001) and flexion-extension (p = 0.001).²⁵ The other three studies did not report on statistical significance of pre-operative vs. post-operative measures.

Radiographic degenerative changes

In two studies that reported pre-operative and post-operative radiographic grading of arthritic changes (n = 61), Rymaszewski et al. showed no change in radiographic findings at an average of six years (range, 2–15) follow-up,²⁷ while Herold et al. showed worsening in radiographic grading at 14 years’ follow-up.²⁶ However, these radiographic findings correlated poorly with clinical results, as range of motion and patient satisfaction were not affected by this radiographic deterioration.²⁶

Comparative studies

One prospective cohort study by Woods et al., with a historical control, compared radial head excision and synovectomy (historical group) with total elbow arthroplasty (TER) in patients with rheumatoid arthritis (excision n = 45, TER n = 38 (45 elbows)).²⁸ This study found that TER was significantly better than radial head excision and synovectomy at reducing pain (p = 0.02).

Complications and revision rate

Complications were reported in 22 of 27 studies. The most common complication was radiographic degenerative changes of the elbow (n = 82), followed by positive ulnar variance (n = 50), and periarticular ossification (n = 46) (Table 2). Only four (8%) of the patients who experienced positive ulnar variance were reported to be symptomatic. Only one (1%) patient with radiographic degenerative changes was reported to be symptomatic, with most studies finding that degenerative changes do not correlate with pain or clinical outcome. Five (10%) of patients with periarticular ossification had reduced elbow movement.

Table 2.

Complications following radial head excision.

Complication	n (%)
Ulnohumeral osteoarthritis	82 (24.0)
Unspecified osteoarthritic changes	48 (14.0)
Wrist osteoarthritis	8 (2.3)
Positive ulnar variance	50 (14.6)
Valgus laxity	24 (7.0)
Radioulnar subluxation	30 (8.8)
Periarticular ossification	46 (13.5)
Partial regrowth of radial head	10 (2.9)
Radioulnar synostosis	2 (0.6)
Synovitis	1 (0.3)
Loose bodies	1 (0.3)
Ulnar nerve lesion	27 (7.9)
Osteoporosis of capitellum	7 (2.0)
Avulsion of bone chip at olecranon process	1 (0.3)
Calcification of joint capsule	1 (0.3)
Transient wound discharge	1 (0.3)
Residual bone fragments	3 (0.9)
Total	342

Seven of the 27 studies included in this review reported revision rate after radial head excision (Table 3). The average revision rate among these studies was 2.7% (range, 0–7.7%). Revision surgeries varied but included TER and ulnar shortening osteotomy.

Table 3.

Reported revisions following radial head excision.

Study	Number of patients requiring revision (%)	Time from primary excision to revision	Revision operations
Boulas and Morrey29	0 (0)
Herbertsson et al.24	1 (7.7)	18 Months	Ulnar nerve transposition
Iftimie et al.30	2 (7.4)	12 Months (n = 1), 14 months (n = 1)	Ulnar shortening osteotomy due to symptomatic proximal translation of the radius greater than 5 mm with persistent wrist pain
Karlsson et al.31	0 (0)
Zarattini et al.22	1 (4.2)	18 Months	Removal of numerous osseous loose bodies
Gendi et al.25	9 (0)	Not reported	Eight total elbow replacements and one repeat excision-synovectomy
Postacchini and Morace32	0
Total: 7	Mean: 2.8

Discussion

The indications for primary excision of the radial head most commonly reported in the literature involved the acute comminuted fracture, as well as rheumatoid arthritis in the chronic setting. Most patients reported good or excellent functional scores when treated with radial head excision in the setting of acute trauma. However, the literature comparing radial head excision to ORIF in the acute setting is less conclusive—of the two such studies included in this review, one found no significant differences, and one found significant differences favoring ORIF only in certain outcome measures such as grip strength and forearm rotation. Therefore, there is not enough data to support the favoring of one procedure over the other, even though from a clinical point of view ORIF is typically performed for those fractures involving less than three fragments that are amenable to fixation.³³ In the chronic setting, conclusions were also unclear. The results of secondary excision following trauma showed a relatively high rate of post-operative pain, low satisfaction, and a high rate of giving up jobs and sports; however, this is supported by only two included studies. Furthermore, one study comparing primary excision (within three weeks) to secondary excision found that the best results occurred after primary excision. Radial head excision in patients with rheumatoid arthritis also had variable outcomes, and although post-operative function was generally good, this tended to deteriorate in the long-term, with residual pain and a high failure rate due to pain being commonly reported.

While the majority of studies concluded radial head excision resulted in favorable long-term outcomes when used to treat isolated displaced and comminuted radial head fractures, it is critical to evaluate the elbow for associated injuries prior to resection, particularly ligamentous injuries, as outcomes in this group were reported to be poor.^{4,7,25,28,29,31–39} As such, radial head excision is contraindicated in the setting of MCL or interosseous membrane injuries unless these injuries are concurrently treated.

The long-term results of radial head excision were described by several studies in this review with a mean follow-up of greater than 14 years,^{4,24,30,31,34,35,37,40} and all found that radial head excision was a viable treatment option with good functional results. Despite being a viable treatment option in certain instances, radial head excision has been shown to have long-term complications. Notably, advanced osteoarthritic changes was uniformly present in a number of these studies but typically was a radiographic finding and not associated with functional impairment.^30,40 Other studies have also suggested that degenerative grading does not necessarily correlate with clinical outcome.²⁷ Therefore, the importance of these radiographic changes in determining the role of radial head excision for fracture treatment is unclear. Also, some long-term follow-up studies demonstrated valgus instability, stiffness, or proximal migration of the radius.^41,42 Overall, excision leads to a very low complication rate, with low-clinical-impact radiographic arthritis being the most significant one.

More evidence is needed to make conclusions about studies comparing radial head excision to replacement for acute treatment of radial head fracture with greater than two years’ follow-up. This review included only one study that compared radial head excision to replacement. This study found no difference between the two treatments in terms of functional scoring systems, but did find that extension was significantly more restricted in the replacement group.²³ In other studies in the literature that assessed outcomes of radial head replacement for fracture with a mean follow-up of eight years or more, all reported sustained good or excellent clinical results with few complications.^43–46 Ultimately, the contention of whether radial head replacement or excision is a better treatment for acute trauma cases remains an area of further study, except where excision is contraindicated such as when valgus or axial instability is present.³⁷

In chronic situations, this review found that in patients with rheumatoid arthritis, 43% (for whom pain outcomes were reported) had unchanged or worsened pain after excision. In addition, range of motion improved in all patients, although radiographic grading of degenerative changes generally worsened or stayed the same. Nevertheless, degenerative grading may not correlate with clinical outcome.²⁷ It has been suggested that when non-operative treatment such as pharmacologic agents is insufficient for the treatment of rheumatoid arthritis, arthroscopic synovectomy is the surgical procedure of choice, with elbow arthroplasty to be considered for more advanced joint destruction.⁴⁷ This review provides some evidence that radial head excision is an effective treatment option for rheumatoid arthritis of the elbow in terms of range of motion, but is not as effective at relieving pain. It could be that patients with rheumatoid arthritis do not thrive after radial head excision due to disease presence persisting in the ulnohumeral joint. Patients may benefit more from TER rather than radial head excision alone, according to one comparative study.²⁸ However, surgical intervention for rheumatoid arthritis is becoming increasingly infrequent due to the existence of modern disease-modifying anti-rheumatic drugs (DMARDs).^48,49 Therefore, although certain surgical interventions have been shown to be a viable option to treat rheumatoid elbows, DMARDs are reducing the need for such interventions.

Some benefits of radial head excision mentioned in the included studies was its simple and rapid nature, low learning curve and technical demand, high rate of satisfactory long-term clinical results, and permission of early postoperative motion.^35,40 Furthermore, radial head excision can also be performed arthroscopically in chronic situations, with similar results to open excision.^50,51 Wijeratna et al. even suggested that arthroscopic excision may provide additional short-term benefits in terms of speed of healing and rehabilitation.⁵¹ Ultimately, in the setting of confirmed elbow stability in patients who have lower demand to their upper extremities, radial head excision for isolated comminuted radial head fracture remains a viable option due to its simplicity.

Strengths and limitations

One strength of this review is its rigorous methodology. A comprehensive search strategy was used to search several large databases, which helped to ensure that no relevant studies were missed. Furthermore, all stages of article screening were done by two independent reviewers, which minimized reviewer bias. An analysis of agreement during the screening process also adds strength to this study. Finally, the review captured a total of 51 studies with 1277 radial head excision patients, a relatively large sample size for this pathology.

Limitations of this systematic review include the low quality of included studies. Most studies were of Level IV evidence, with some Level III studies. Included studies were often retrospective observational studies such as case series with no comparative control group or cohort studies with no randomization, which increases the risk of bias. Due to significant heterogeneity in outcome reporting among included studies, a meta-analysis was not possible. One final limitation is that studies were excluded if they were not published in English lending a potential for language bias in our findings.

Conclusions

The primary indications for radial head excision reported in the literature are fracture comminution in the acute setting, and rheumatoid arthritis in the chronic setting. Outcomes were generally good or excellent following radial head excision for radial head fractures. There is insufficient evidence to recommend ORIF or replacement compared to radial head excision in the setting of complex radial head fractures. Radial head excision should not be performed when concurrent or ligamentous injuries are present. Secondary radial head excision in posttraumatic situations showed inferior results and patients often continued to have residual pain. In patients treated with radial head excision for rheumatoid arthritis, the range of motion improved, but pain was not effectively relieved and degenerative grading generally worsened. The strength of these conclusions is limited by the quality of the included literature and the inconsistencies in outcome reporting among studies.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Graham King has financial involvement with Lawson Health Research Institute.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material is available at: http://journals.sagepub.com/doi/suppl/10.1177/1758573219864305.