Orthopedics

Review Article 

Outcomes Following Radial Head Arthroplasty

John R. Fowler, MD; Sarah E. Henry, MD; Peter Xu, BA; Robert J. Goitz, MD

Abstract

Most current series of radial head arthroplasty include small numbers of patients with short- to medium-term follow-up and significant heterogeneity in patients, treatments, and outcome measures. The purpose of this systematic review was to review outcomes for radial head arthroplasty based on injury chronicity, injury pattern, and type of implant used. The authors systematically searched electronic databases for studies containing radial head arthroplasty or radial head replacement and identified 19 studies for inclusion in the analysis. For each included study, a composite mean was obtained for Mayo Elbow Performance Score (MEPS) and range of motion. Outcomes were said to differ significantly if their confidence intervals did not overlap. The MEPS for acute treatment (90) was higher than that for delayed treatment (81). There was no difference in the pooled MEPS between the isolated (89) and complex injury pattern (87) groups or implant material. There was no difference in range of motion between the acute and delayed or isolated and complex groups, but the average degree of pronation was higher in patients treated with titanium implants (76°) compared with cobalt chromium implants (66°). This systematic review suggests that outcomes are improved following acute arthroplasty for treatment of radial head fractures compared with delayed treatment, based on MEPS. The lack of other significant differences detected is likely due to the significant heterogeneity and inadequate power in current studies. Further prospective studies isolating the different variables will be needed to determine their true effect on outcomes. [Orthopedics. 2016; 39(3):153–160.]

Abstract

Most current series of radial head arthroplasty include small numbers of patients with short- to medium-term follow-up and significant heterogeneity in patients, treatments, and outcome measures. The purpose of this systematic review was to review outcomes for radial head arthroplasty based on injury chronicity, injury pattern, and type of implant used. The authors systematically searched electronic databases for studies containing radial head arthroplasty or radial head replacement and identified 19 studies for inclusion in the analysis. For each included study, a composite mean was obtained for Mayo Elbow Performance Score (MEPS) and range of motion. Outcomes were said to differ significantly if their confidence intervals did not overlap. The MEPS for acute treatment (90) was higher than that for delayed treatment (81). There was no difference in the pooled MEPS between the isolated (89) and complex injury pattern (87) groups or implant material. There was no difference in range of motion between the acute and delayed or isolated and complex groups, but the average degree of pronation was higher in patients treated with titanium implants (76°) compared with cobalt chromium implants (66°). This systematic review suggests that outcomes are improved following acute arthroplasty for treatment of radial head fractures compared with delayed treatment, based on MEPS. The lack of other significant differences detected is likely due to the significant heterogeneity and inadequate power in current studies. Further prospective studies isolating the different variables will be needed to determine their true effect on outcomes. [Orthopedics. 2016; 39(3):153–160.]

Radial head fractures are the most common fracture around the elbow, with an incidence of 30 per 100,000 persons per year, accounting for 1% to 5% of all long bone fractures and 20% of proximal forearm fractures.1 The radial head plays a critical role in elbow mechanics, transmitting up to 90% of the force across the elbow and acting as an important secondary stabilizer to valgus force.2,3 Radial head arthroplasty has become an accepted treatment option for unreconstructible radial head fractures, especially those with associated ligamentous and/or bony injuries, and for treatment of complications after failed internal fixation or radial head excision.4–18 However, most series reporting outcomes following radial head arthroplasty include a single cohort with no control or comparison group and consist of a heterogeneous group of patients, treatments, and outcome measures.4,6,8–11,13,18 These series do not differentiate between acute or delayed treatment of the injury, isolated radial head fractures or those combined with associated injuries, or use of different implants.

Mayo Elbow Performance Score (MEPS) and range of motion (ROM) have been reported to be superior following acute radial head arthroplasty compared with delayed reconstruction; however, other studies report no difference or do not directly compare the 2 groups.5,6,11,18–20 Although it is logical to hypothesize that patients with an isolated radial head fracture would have a better outcome than patients with a terrible triad injury pattern, this has not been proven in the literature.11,19,21,22 Since Speed23 performed the first documented radial head arthroplasty in 1941, various implant designs have been used. Silicone arthroplasty has fallen out of favor because of its suboptimal wear characteristics and implant failures.24 Various metals, including cobalt chromium (CoCr), titanium, and pyrocarbon, are currently being used in different designs, including monoblock, bipolar, and modular, without strong evidence of superiority of one implant metal or design. A large number of arthroplasties have been press-fit or left loose as a spacer, whereas a smaller number have been cemented, with no known difference in outcomes.11,19,25 The specific type of metal may be less important than the design, surgical technique, and indications, but the direct influence of each variable is currently unknown.25 The main goals of radial head arthroplasty are restoring the height of the radius, preventing proximal migration of the radius, and restoring stability and pain-free functional ROM to the elbow. If these goals are met, it may be difficult to detect smaller differences in functional outcomes related to other variables. The majority of studies include heterogeneous groups regarding injury chronicity, injury pattern, and type of implant; therefore, the effect of each variable on functional outcomes is unknown.4–8,10–12,15–22,25–27 Thus, the purpose of this systematic review was to define expected outcomes for radial head arthroplasty based on injury chronicity, injury pattern, and type of implant used.

Search Strategy and Criteria

The criteria of the Preferred Reporting Items for Systematic Reviews and Meta-analyses by the PRISMA group were used. The authors systematically screened the electronic databases PubMed, EMBASE, and the Cochrane Controlled Trials Register. For all studies selected, the reference lists were searched for additional studies, and only studies published between 1940 and 2013 were included. The following search terms were used: radial head arthroplasty and radial head replacement.

Two reviewers (J.R.F., S.E.H.) independently assessed all studies for inclusion. In cases of disagreement between the reviewers, the opinion of the senior author (R.J.G.) was decisive. To prevent investigator bias, reviewers scoring manuscripts were blinded to author and institute. Studies were included when radial head arthroplasty was performed with a non-silicone prosthesis and presented individual patient data on injury pattern, time to surgery, follow-up, ROM, type of implant used, and validated outcome scores. The authors of studies that did not publish individual patient data were contacted by email and asked to provide the data.

Trials were considered to be valid when they satisfied the inclusion and exclusion criteria and contained sufficient data for further analysis. The initial database search identified 1415 potential manuscripts (Figure 1). Title review resulted in the exclusion of 109 studies that were duplicates in the search and 1203 studies that were on topics unrelated to radial head arthroplasty. This left 103 studies for full manuscript review. The reference list for each of these manuscripts was examined for additional studies, but no additional studies were identified. After full manuscript review, 5 studies were excluded because they were review studies that did not contain any patient data; 3 were excluded because they were technique studies without patient data; 26 were excluded because individual patient data were not available and the corresponding authors did not provide additional data (because injury type and time to treatment were critical in the analysis, it was necessary to exclude studies where these factors were not clear for individual patients); 39 were excluded because the manuscripts were not available in English; 3 were excluded because, after full manuscript review, the studies did not discuss radial head arthroplasty; and 8 were excluded because the prosthesis used was silicone. This left 19 studies to be included in the analysis.


Search strategy flowchart.

Figure 1:

Search strategy flowchart.

The reviewers extracted the following data from the manuscripts: patient age; patient sex; type of implant used; acute (<4 weeks from injury) vs delayed (>4 weeks from injury) treatment; injury pattern (isolated, elbow dislocation, terrible triad, Monteggia fracture radioulnar longitudinal instability injury); time to treatment (months); follow-up (months); Disabilities of the Arm, Shoulder, and Hand (DASH) score (if available); MEPS (if available); visual analog scale pain score (if available); elbow flexion; elbow extension deficit; supination; pronation; grip strength; and presence of complications. The MEPS has a maximum of 100 points: pain (45 points), ROM (20 points), stability (10 points), and function (25 points) are evaluated. The DASH includes 30 questions investigating disability resulting from elbow problems.

Four comparison groups of interest were defined for the purpose of meta-analysis: acute vs delayed treatment, isolated vs nonisolated injury pattern, implant material (CoCr vs titanium vs polycarbon), and patient age (≤50 vs >50 years). Five outcome measurements were examined: MEPS, elbow flexion, elbow extension deficit, supination, and pronation. Due to the fact that most of the 19 studies reported outcomes within only 1 subset of these 4 comparison groups (eg, all of the patients of Brinkman et al5 received delayed treatment rather than a mix of acute and delayed, which would allow for direct comparison of acute vs delayed), meta-analysis was used to generate pooled mean values and confidence intervals (CIs) for each of the 5 outcomes within each of the comparison groups. Figure 2 shows an example forest plot and pooled 95% CI, which was generated for the MEPS outcome within studies that reported delayed treatment; all within-group estimates of mean outcome and all CIs were generated in this fashion.


Forest plot of mean Mayo Elbow Performance Score for studies reporting outcomes for delayed treatment. Abbreviations: CI, confidence interval; ES, elbow score; ID, identification.

Figure 2:

Forest plot of mean Mayo Elbow Performance Score for studies reporting outcomes for delayed treatment. Abbreviations: CI, confidence interval; ES, elbow score; ID, identification.

In the event that a study reported patients in both comparison groups of interest, such as acute and delayed treatment, then appropriate patients from that study could appear in the meta-analysis generation of pooled means for both comparison groups, facilitated by the fact that subject-level data were available for all included studies.

Statistical Analysis

For a given outcome, studies were included if there were at least 5 patients who had relevant data points to calculate a mean for that outcome. Furthermore, a comparison of 2 groups (eg, acute vs delayed treatment) was only analyzed if there were at least 3 studies per group with calculated means. Summary means within each comparison group for each of the 5 outcomes were pooled across studies by meta-analysis using the random-effects model. Meta-analysis was performed in Stata 13 (StataCorp, College Station, Texas), with use of the METAN Stata module. Differences between means were examined, and the lack of overlap of 95% CIs was considered suggestive but not conclusive of statistically significant differences.

An I2 value was calculated for each analysis to indicate the level of heterogeneity among studies included for that analysis. A higher I2 value indicates greater heterogeneity, and therefore decreased reliability, in the pooled data for the outcome. An I2 value of less than 25% is excellent, 25% to 50% is acceptable, 50% to 75% is weak, and greater than 75% is poor. For differences between comparison groups, to the authors' knowledge, no method exists to exactly quantify a significant difference between 2 pooled outcome means that were generated by separate meta-analyses, so the authors considered a lack of overlapping CIs to be roughly suggestive of a difference between comparison groups. For this reason, no exact P values were calculated, and multiple comparison corrections were not performed.

There were not enough data points to include pyrocarbon or vitallium (CoCr and molybdenum) in the MEPS analysis because at least 3 studies with at least 5 patients were required for inclusion in analysis. Visual analog scale pain scores and DASH scores were not analyzed because these 2 outcomes did not include a sufficient number of studies to yield statistically useful results.

Results

Nineteen studies were included and described 365 patients in regard to acute vs delayed treatment, 383 patients in regard to injury pattern, and 394 patients in regard to implant type. Of the 19 studies included in this analysis, 9 included only acute arthroplasty, 3 included only delayed reconstruction, and 7 included both (Table 1). Of these, 3 reported improved results following acute radial head arthroplasty, but 2 of these studies included patients with a variety of injury patterns and 1 did not compare the results to delayed reconstruction. Of the 19 studies included in this analysis, 2 included only complex injury pattern, whereas the remaining 17 included both isolated radial head fractures and complex injury patterns. Of the 19 studies included in this analysis, 8 used CoCr, 4 used titanium, 2 used vitallium, 2 used pyrocarbon, and 3 used mixed implants, and all included patients with a mixed pattern of injury. Of the CoCr studies, 4 were cemented bipolars, 1 was an uncemented bipolar, and 2 were modular. Of the titanium studies, 3 were monoblock and 1 was modular.


Injury Chronicity, Injury Pattern, and Implant Materials Reported in the Included Studies

Table 1:

Injury Chronicity, Injury Pattern, and Implant Materials Reported in the Included Studies

The pooled MEPS for acute treatment, 89.7 (95% CI, 87–92), was greater than that for delayed treatment groups, 80.9 (95% CI, 76–85) (Table 2). The pooled MEPS was not significantly different between the isolated (88.9) and nonisolated (86.9) groups (Table 3), age younger than 50 years (85.9) and age older than 50 years (88.6) (Table 4), or CoCr (86.9) and titanium (85.8) implant material (Table 5). Range of motion was not significantly different between the acute and delayed, isolated and nonisolated, and age younger than 50 years and older than 50 years groups. The average degree of pronation was found to be significantly different between patients with CoCr (66.3; 95% CI, 62–70) and titanium (76.3; 95% CI, 71–80) implants. No significant difference was found between CoCr and titanium groups in other measured ROM (flexion, extension, or supination).


Mean Values for Outcome Measures Analyzed by Injury Chronicity

Table 2:

Mean Values for Outcome Measures Analyzed by Injury Chronicity


Mean Values for Outcome Measures Analyzed by Injury Pattern

Table 3:

Mean Values for Outcome Measures Analyzed by Injury Pattern


Mean Values for Outcome Measures Analyzed by Patient Age

Table 4:

Mean Values for Outcome Measures Analyzed by Patient Age


Mean Values for Outcome Measures Analyzed by Implant Metal

Table 5:

Mean Values for Outcome Measures Analyzed by Implant Metal

Discussion

Most current series of radial head arthroplasty include small numbers of patients with short- to medium-term follow-up and significant heterogeneity in injury chronicity, pattern, type of implant, and outcome measures. The purpose of this systematic review was to review expected outcomes for radial head arthroplasty based on injury chronicity, pattern, and type of implant used based on pooled analysis of the individual patients from each included study. The MEPS is widely used and a validated measure of elbow function.28 In the current review, the pooled MEPS score was found to be significantly higher for patients treated acutely compared with those treated in a delayed manner, and the average pronation was found to be significantly higher for patients treated with titanium implants compared with CoCr implants. No significant differences were found in other analyzed comparisons.

Regarding injury chronicity, the MEPS was higher in patients treated acutely vs in a delayed fashion, but no difference in ROM was found between groups. Zunkiewicz et al20 retrospectively reviewed 29 patients with a mean follow-up of 34 months. Mean MEPS was higher for the acute group (95) compared with the delayed group (85) (P=.03). These results are comparable with the current study's mean MEPS for acute (89.7) and delayed (80.9) groups. Chapman et al6 reviewed 16 patients with a mean follow-up of 33 months. Average flexion contracture was significantly less and arc of motion significantly greater in the acute group (9° and 125°, respectively) compared with the delayed group (21° and 104°, respectively). The authors also noted worse DASH scores for the delayed group (30) compared with the acute group (24).6 Burkhart et al19 reviewed 19 patients with a mean 8.8-year follow-up and reported no statistically significant difference between acute and delayed treatment.

Regarding injury pattern, the current systematic review found no difference in MEPS or ROM in patients with isolated vs complex injury patterns. Celli et al21 reviewed 16 patients treated acutely, including 9 isolated radial head fractures and 7 associated elbow dislocations, with a mean follow-up of 41 months and found no difference in the mean arcs of motion between groups. Although it may seem logical that a patient with an isolated radial head fracture may have a better outcome than a patient with associated injury patterns, if all of the injuries are addressed intraoperatively, studies report that patients achieve good functional results.4,8,11,16,19,22 It is important to recognize that the lack of detectable differences between outcomes may be a function of combining a variety of injury patterns and combining causes for delayed treatment due to the small sample sizes. Treatment of radial head fractures with associated ligamentous and/or bony pathology has been reported to have good or excellent outcomes in 76% to 94% of patients at long-term follow-up.11,19 Winter et al22 reviewed 13 patients who underwent acute radial head arthroplasty for terrible triad injuries with a mean follow-up of 25 months. Mean flexion contracture was 11°, and 84% of patients were very satisfied or satisfied.22 Heijink et al27 reviewed 8 patients who underwent delayed treatment for radioulnar longitudinal instability injuries with mean follow-up of 5.7 years and found an average MEPS of 71, significantly lower than reported in other injury types. Delayed treatment of radioulnar longitudinal instability injuries represents a challenging patient group. The current study included 29 patients with radioulnar longitudinal instability and found an average MEPS of 80, average flexion of 133°, and average extension deficit of 9°.

Regarding type of implant, this systematic review found improved pronation in patients treated with titanium implants compared with those treated with CoCr implants but found no differences in MEPS or flexion or extension motion. A variety of implant materials and designs are currently used, without strong evidence of superiority of one type of implant (Table 1).4–8,10–12,15–22,25–27 Metals include CoCr, titanium, vitallium, and pyrocarbon, and designs include monoblock, bipolar, or modular and may be press-fit or cemented. Burkhart et al19 reported an average MEPS of 91, average flexion contracture of 21°, average pronation of 64°, and average supination of 64° at a mean follow-up of 8.8 years using a CoCr cemented bipolar implant. Harrington et al11 reported an average MEPS of 88, average flexion contracture of 17°, average pronation of 74°, and average supination of 65° at a mean follow-up of 12 years using a titanium uncemented monoblock implant. Conflicting data have been reported on the function of bipolar prostheses compared with monoblock designs.29,30 Rotini et al25 reviewed 30 patients treated with either monoblock (n=12) or bipolar (n=19) CoCr uncemented implants at a mean follow-up of 2 years. There were no significant differences in outcomes (MEPS, ROM, or complications) between patients treated with monoblock vs bipolar implants.25 Clinical short- to medium-term outcomes studies have shown favorable results with both implant designs and all types of metals, but most of these studies have relatively short follow-up and do not report detailed analysis of complications and revisions.4–8,10–12,15–22,25–27

The current study has some limitations. First, by excluding studies that did not have individual patient data available, selection bias was introduced. The authors attempted to contact the corresponding authors for all studies that met other criteria but lacked the individual patient data. Two of the corresponding authors provided the individual data, and these studies were included in the analysis.8,20 Requiring individual patient data was the only way to account for the heterogeneity from these series by specifically defining the injury type, implant type, and time to treatment for each patient. Although a noted limitation, the authors feel that this novel approach strengthens the analysis.

The meta-analysis performed in this study also has limitations. The studies included were often retrospective case series, and the literature search was limited by publication bias. Most of the included studies did not report comparative measures, such as effect size differences between the comparison groups of interest; instead, most articles included patients from only 1 of the comparison groups of interest. As such, the most feasible meta-analysis was an estimation of pooled mean and 95% CI for each outcome within each comparison group. Examination of CI overlap between comparison groups is only an approximate way to assess potentially significant differences. Given that the authors made many comparisons, there is the possibility that the CIs may not overlap on one of the comparisons by chance alone. Another limitation is the lack of a Bonferroni correction in the statistical analysis. In addition, due to existing heterogeneity between the available studies, several of the meta-analyses generated I2 values above 75%; this is simply a limitation of the existing studies for this subject. Finally, the reason for the difference in MEPS between acute and chronic injuries is unclear. Although it may be related to surgical timing, it could also be related to unknown patient factors associated with chronic injuries. Given that none of the studies included in the analysis were randomized, the authors are unable to answer this question.

Conclusion

The analysis performed in this systematic review suggests that outcomes are improved following acute arthroplasty for treatment of radial head fractures compared with delayed treatment based on MEPS. Pronation was significantly higher in patients treated with titanium implants compared with CoCr implants. However, this difference, as well as the lack of other significant differences, is likely due to the significant heterogeneity in patients, treatments, and outcome measurements; small sample size; and short- to medium-term follow-up. Further prospective studies isolating the different variables are needed to determine the true effect on outcomes of acute vs delayed treatment, outcome of isolated fractures vs those with associated injuries, and type and design of various implants. There are few data available to address these specific questions; thus, sufficiently powered, prospective, randomized, multicenter trials need to be performed with specific inclusion criteria to assist in answering these important questions.

References

  1. Duckworth AD, Clement ND, Jenkins PJ, Aitken SA, Court-Brown CM, McQueen MM. The epidemiology of radial head and neck fractures. J Hand Surg Am. 2012; 37(1):112–119. doi:10.1016/j.jhsa.2011.09.034 [CrossRef]
  2. Bryce CD, Armstrong AD. Anatomy and biomechanics of the elbow. Orthop Clin North Am. 2008; 39(2):141–154. doi:10.1016/j.ocl.2007.12.001 [CrossRef]
  3. Morrey BF, An KN, Stormont TJ. Force transmission through the radial head. J Bone Joint Surg Am. 1988; 70(2):250–256.
  4. Ashwood N, Bain GI, Unni R. Management of Mason type-III radial head fractures with a titanium prosthesis, ligament repair, and early mobilization. J Bone Joint Surg Am. 2004; 86(2):274–280.
  5. Brinkman JM, Rahusen FT, de Vos MJ, Eygendaal D. Treatment of sequelae of radial head fractures with a bipolar radial head prosthesis: good outcome after 1–4 years follow-up in 11 patients. Acta Orthop. 2005; 76(6):867–872. doi:10.1080/17453670510045516 [CrossRef]
  6. Chapman CB, Su BW, Sinicropi SM, Bruno R, Strauch RJ, Rosenwasser MP. Vitallium radial head prosthesis for acute and chronic elbow fractures and fracture-dislocations involving the radial head. J Shoulder Elbow Surg. 2006; 15(4):463–473. doi:10.1016/j.jse.2005.09.010 [CrossRef]
  7. Chien HY, Chen AC, Huang JW, Cheng CY, Hsu KY. Short- to medium-term outcomes of radial head replacement arthroplasty in posttraumatic unstable elbows: 20 to 70 months follow-up. Chang Gung Med J. 2010; 33(6):668–678.
  8. Doornberg JN, Parisien R, van Duijn PJ, Ring D. Radial head arthroplasty with a modular metal spacer to treat acute traumatic elbow instability. J Bone Joint Surg Am. 2007; 89(5):1075–1080. doi:10.2106/JBJS.E.01340 [CrossRef]
  9. Flinkkila T, Kaisto T, Sirnio K, Hyvonen P, Leppilahti J. Short- to mid-term results of metallic press-fit radial head arthroplasty in unstable injuries of the elbow. J Bone Joint Surg Br. 2012; 94(6):805–810. doi:10.1302/0301-620X.94B6.28176 [CrossRef]
  10. Grewal R, MacDermid J, Faber KJ, Drosdowech DS, King GJ. Comminuted radial head fractures treated with a modular metallic radial head arthroplasty: study of outcomes. J Bone Joint Surg Am. 2006; 88(10):2192–2200. doi:10.2106/JBJS.E.00962 [CrossRef]
  11. Harrington IJ, Sekyi-Otu A, Barrington TW, Evans DC, Tuli V. The functional outcome with metallic radial head implants in the treatment of unstable elbow fractures: a long-term review. J Trauma. 2001; 50(1):46–52. doi:10.1097/00005373-200101000-00009 [CrossRef]
  12. Judet T, Garreau de Loubresse C, Piriou P, Charnley G. A floating prosthesis for radial-head fractures. J Bone Joint Surg Br. 1996; 78(2):244–249.
  13. King GJ. Management of comminuted radial head fractures with replacement arthroplasty. Hand Clin. 2004; 20(4):429–441. doi:10.1016/j.hcl.2004.06.003 [CrossRef]
  14. Knight D, Rymaszewski L, Amis AA, Miller JH. Primary replacement of the fractured radial head with a metal prosthesis. J Bone Joint Surg Br. 1993; 75(4):572–576.
  15. Lim YJ, Chan BK. Short-term to medium-term outcomes of cemented Vitallium radial head prostheses after early excision for radial head fractures. J Shoulder Elbow Surg. 2008; 17(2):307–312. doi:10.1016/j.jse.2007.07.020 [CrossRef]
  16. Moro JK, Werier J, MacDermid JC, Patterson SD, King GJ. Arthroplasty with a metal radial head for unreconstructible fractures of the radial head. J Bone Joint Surg Am. 2001; 83(8):1201–1211.
  17. Ricon FJ, Sanchez P, Lajara F, Galan A, Lozano JA, Guerado E. Result of a pyrocarbon prosthesis after comminuted and unreconstructable radial head fractures. J Shoulder Elbow Surg. 2012; 21(1):82–91. doi:10.1016/j.jse.2011.01.032 [CrossRef]
  18. Shore BJ, Mozzon JB, MacDermid JC, Faber KJ, King GJ. Chronic posttraumatic elbow disorders treated with metallic radial head arthroplasty. J Bone Joint Surg Am. 2008; 90(2):271–280. doi:10.2106/JBJS.F.01535 [CrossRef]
  19. Burkhart KJ, Mattyasovszky SG, Runkel M, et al. Mid- to long-term results after bipolar radial head arthroplasty. J Shoulder Elbow Surg. 2010; 19(7):965–972. doi:10.1016/j.jse.2010.05.022 [CrossRef]
  20. Zunkiewicz MR, Clemente JS, Miller MC, Baratz ME, Wysocki RW, Cohen MS. Radial head replacement with a bipolar system: a minimum 2-year follow-up. J Shoulder Elbow Surg. 2012; 21(1):98–104. doi:10.1016/j.jse.2011.05.012 [CrossRef]
  21. Celli A, Modena F, Celli L. The acute bipolar radial head replacement for isolated unreconstructable fractures of the radial head. Musculoskeletal Surg. 2010; 94(suppl 1):S3–S9. doi:10.1007/s12306-010-0062-y [CrossRef]
  22. Winter M, Chuinard C, Cikes A, Pelegri C, Bronsard N, de Peretti F. Surgical management of elbow dislocation associated with non-reparable fractures of the radial head. Chir Main. 2009; 28(3):158–167. doi:10.1016/j.main.2009.02.003 [CrossRef]
  23. Speed K. Ferrule caps for the head of the radius. Surg Gynecol Obstet. 1941; 73:845–850.
  24. Vanderwilde RS, Morrey BF, Melberg MW, Vinh TN. Inflammatory arthritis after failure of silicone rubber replacement of the radial head. J Bone Joint Surg Br. 1994; 76(1):78–81.
  25. Rotini R, Marinelli A, Guerra E, Bettelli G, Cavaciocchi M. Radial head replacement with unipolar and bipolar SBi system: a clinical and radiographic analysis after a 2 year mean follow-up. Musculoskeletal Surg. 2012; 96(suppl 1):S69–S79. doi:10.1007/s12306-012-0198-z [CrossRef]
  26. Allieu Y, Winter M, Pequignot JP, de Mourgues P. Radial head replacement with a pyrocarbon head prosthesis: preliminary results of a multicentric prospective study. Eur J Orthop Surg Traumatol. 2006; 16(1):1–9. doi:10.1007/s00590-005-0041-4 [CrossRef]
  27. Heijink A, Morrey BF, van Riet RP, O'Driscoll SW, Conney WP III, . Delayed treatment of elbow pain and dysfunction following Essex-Lopresti injury with metallic radial head replacement: a case series. J Shoulder Elbow Surg. 2010; 19(6):929–936. doi:10.1016/j.jse.2010.03.007 [CrossRef]
  28. Cusick MC, Bonnaig NS, Azar FM, Mauck BM, Smith RA, Throckmorton TW. Accuracy and reliability of the Mayo Elbow Performance Score. J Hand Surg Am. 2014; 39(6):1146–1150. doi:10.1016/j.jhsa.2014.01.041 [CrossRef]
  29. Moon JG, Berglund LJ, Zachary D, An KN, O'Driscoll SW. Radiocapitellar joint stability with bipolar versus monopolar radial head prostheses. J Shoulder Elbow Surg. 2009; 18(5):779–784. doi:10.1016/j.jse.2009.02.011 [CrossRef]
  30. Yian E, Steens W, Lingenfelter E, Schneeberger AG. Malpositioning of radial head prostheses: an in vitro study. J Shoulder Elbow Surg. 2008; 17(4):663–670. doi:10.1016/j.jse.2007.12.008 [CrossRef]

Injury Chronicity, Injury Pattern, and Implant Materials Reported in the Included Studies

StudyInjury ChronicityInjury PatternImplant MetalImplant DesignImplant Fixation
Ashwood et al4MixedMixedTitaniumMonoblockUncemented
Brinkman et al5DelayedMixedCoCrBipolarCemented
Chapman et al6MixedMixedVitalliumMonoblockUncemented
Chien et al7MixedMixedTitaniumModularUncemented
Doornberg et al8AcuteMixedCoCrModularUncemented
Grewal et al10AcuteMixedCoCrModularUncemented
Harrington et al11AcuteMixedTitaniumMonoblockUncemented
Judet et al12MixedMixedCoCrBipolarCemented
Lim and Chan15AcuteMixedVitalliumMonoblockCemented
Moro et al16AcuteMixedTitaniumMonoblockUncemented
Ricon et al17AcuteMixedPyrocarbonModularUncemented
Shore et al18DelayedMixedMixedMixedUncemented
Burkhart et al19MixedMixedCoCrBipolarUncemented
Zunkiewicz et al20MixedMixedCoCrBipolarUncemented
Celli et al21AcuteMixedCoCrBipolarCemented
Winter et al22AcuteComplexMixedMonoblockUncemented
Rotini et al25AcuteMixedCoCrMixedUncemented
Allieu et al26MixedMixedPyrocarbonModularUncemented
Heijink et al27DelayedComplexMixedMonoblockUncemented

Mean Values for Outcome Measures Analyzed by Injury Chronicity

Outcome MeasureNo. of Studies IncludedMean (95% CI)

AcuteDelayed
Elbow flexion14 acute, 8 delayed130.1° (125.9°–134.3°)119.0° (111.7°–126.2°)
Elbow extension deficit12 acute, 6 delayed14.3° (10.3°–18.2°)17.5° (12.6°–22.3°)
Supination13 acute, 7 delayed70.7° (65.0°–76.4°)70.6° (64.9°–76.3°)
Pronation14 acute, 8 delayed72.68° (68.3°–77.1°)67.5° (62.3°–72.8°)
MEPS10 acute, 7 delayed89.7 (87.4–92.13)a80.9 (76.5–85.4)a

Mean Values for Outcome Measures Analyzed by Injury Pattern

Outcome MeasureNo. of Studies IncludedMean (95% CI)

IsolatedComplex
Elbow flexion13 isolated, 14 nonisolated127.0° (121.1°–132.9°)128.2° (123.7°–132.8°)
Elbow extension deficit11 isolated, 11 nonisolated13.1° (7.9°–18.3°)17.2° (13.6°–20.9°)
Supination12 isolated, 13 nonisolated73.4° (68.7°–78.0°)67.3° (60.1°–74.4°)
Pronation13 isolated, 14 nonisolated70.5° (64.8°–76.1°)69.4° (64.1°–74.7°)
MEPS11 isolated, 12 nonisolated89.0 (85.1–92.8)87.0 (83.5–90.5)

Mean Values for Outcome Measures Analyzed by Patient Age

Outcome MeasureNo. of Studies IncludedMean (95% CI)

≤50 Years Old>50 Years Old
Elbow flexion15 <50 y, 15 >50 y128.4° (124.2°–132.7°)127.4° (122.3°–132.6°)
Elbow extension deficit13 <50 y, 13 >50 y13.3° (10.3°–16.4°)15.9° (11.9°–19.9°)
Supination14 <50 y, 14 >50 y72.0° (67.3°–76.7°)67.3° (59.6°–75.0°)
Pronation15 <50 y, 15 >50 y71.5° (66.5°–76.4°)70.6° (65.2°–76.0°)
MEPS13 <50 y, 11 >50 y86.0 (82.4–90.0)88.6 (86.1–91.2)

Mean Values for Outcome Measures Analyzed by Implant Metal

Outcome MeasureNo. of Studies IncludedMean (95% CI)

CoCrTitaniumPyrocarbon
Elbow flexion8 CoCr, 5 titanium, 3 pyrocarbon129.5° (126.5°–132.6°)132.8° (126.8°–138.7°)119.9° (108.0°–131.8°)
Elbow extension deficit8 CoCr, 4 titanium, n/a pyrocarbon18.4° (16.0°–20.9°)14.3° (6.8°–21.8°)(Not run; only 2 studies)
Supination8 CoCr, 4 titanium, 3 pyrocarbon69.7° (63.1°–76.3°)66.3° (56.3°–76.3°)76.7° (65.3°–88.2°)
Pronationa9 CoCr, 4 titanium, 3 pyrocarbon66.4° (62.4°–70.3°)a76.3° (71.9°–80.8°)a77.2° (69.8°–84.6°)
MEPS6 CoCr, 5 titanium, n/a pyrocarbon87.0 (82.5–91.5)85.8 (83.0–88.6)(Not run; only 1 study)
Authors

The authors are from the Department of Orthopedic Surgery, University of Pittsburgh, Pittsburgh, Pennsylvania.

The authors have no relevant financial relationships to disclose.

This study was supported by the National Institutes of Health (grants UL1-RR-024153 and UL1-TR-000005).

The authors thank CTSI for assistance with statistical analysis.

Correspondence should be addressed to: John R. Fowler, MD, Department of Orthopedic Surgery, University of Pittsburgh, 3471 Fifth Ave, Ste 1010, Kaufmann Bldg, Pittsburgh, PA 15213 ( fowlerjr@upmc.edu).

Received: August 11, 2014
Accepted: February 04, 2015

Posted Online: April 05, 2016

10.3928/01477447-20160324-06

Sign up to receive

Journal E-contents