A series of 103 acute fractures of the coronoid process of the ulna in 101 patients was reviewed to determine their frequency. The Regan-Morrey classification, treatment, associated injuries, course and outcomes were evaluated. Of the 103 fractures, 34 were type IA, 17 type IB, ten type IIA, 19 type IIB, ten type IIIA and 13 type IIIB. A total of 44 type-I fractures (86%) were treated conservatively, while 22 type-II (76%) and all type-III fractures were managed by operation.
At follow-up at a mean of 3.4 years (1 to 8.9) the range of movement differed significantly between the types of fracture (p = 0.002). Patients with associated injuries had a lower Mayo elbow performance score (p = 0.03), less extension (p = 0.03), more pain (p = 0.007) and less pronosupination (p = 0.004), than those without associated injuries. The presence of a fracture of the radial head had the greatest effect on outcome. An improvement in outcome relative to that of a previous series was noted, perhaps because of more aggressive management and early mobilisation. While not providing complete information about the true details of a fracture and its nature, the Regan-Morrey classification is useful as a broad index of severity and prognosis.
There are few data available regarding the outcome after fractures of the coronoid process of the ulna. Based on a series of 37 fractures, Regan and Morrey1 proposed a classification according to the findings on lateral radiographs. Type-I fractures involve the tip, type-II < 50% of the height of the coronoid process and type-III > 50%. In the 32 patients available for follow-up, 92% of those with type-I, 73% of those with type-II and 20% of those with type-III fractures had a satisfactory outcome.1 Immobilisation was linked to a poorer outcome and the authors therefore advised early movement. Biomechanical studies and clinical observation have shown the importance of the coronoid process in the stability of the elbow and there has, therefore, been a trend towards operative stabilisation of more severe fractures.
Although biomechanical studies have documented the role of the coronoid process in providing stability to the elbow, simulation in vitro does not replicate this.2–4 Clinically, Doornberg and Ring5 highlighted the importance of recognising more complex, but clinically significant types of coronoid fracture and instituting appropriate treatment. In a series of 18 patients, they noted that a poorer result was seen in those who were thought retrospectively to have had inadequate treatment compared with those managed by adequate fixation.
We have analysed the outcome of patients with fractures of the coronoid process and correlated this with the pattern of the fracture, concomitant injuries and treatment, allowing further refinement of the system of classification. Based on the recommendations of the initial series of Regan and Morrey,1 we believed that the treatment of these fractures would subsequently be more aggressive in type-II and type-III fractures and that the outcome would be better than that seen in the original series.
Patients and Methods
After approval from the Institutional Review Board, all consecutive acute fractures of the coronoid process treated at our institution over a period of five years, with a follow-up of > 12 months, were reviewed to determine the Regan and Morrey classification, associated injuries, treatment and outcome.
Patients with nonunion or injuries more than four weeks old were excluded. Fractures which had received definitive treatment elsewhere were also excluded because details of the injury, management, and operative notes were lacking. Likewise, those with a follow-up of < 12 months were excluded. In total, 32 patients were excluded; five had earlier treatment or chronic fractures, three had died and 24 had inadequate follow-up.
This left 103 fractures in 101 patients (44 women, 57 men) who had a mean age of 45 years (10 to 83) (Table I⇓). The mean follow-up was for 3.4 years (1 to 8.9). The right hand was dominant in 78 patients, the left hand in ten, both in two, and the dominance was unknown in 11. The left arm was involved in 52% of fractures and the right in 46%. Bilateral injuries were present in two patients.
The plain radiographs were reviewed to determine the types of fracture and associated injuries. The fractures were classified according to the Regan-Morrey1 system and then subclassified into type A, with no associated dislocation of the elbow, or type B, with an associated dislocation. The records were reviewed to document the treatment, the postoperative course and complications, and the final result. The patients were evaluated at follow-up by retrospective review of the notes or by questionnaire and documented physical examination. The Mayo elbow performance score6 was calculated. Patients were questioned as to their level of pain (0, no pain; 1, mild pain with activity, 2; mild pain at rest or at night; 3, moderate pain limiting activity; 4, moderate pain at rest or at night; and 5, severe or disabling pain) and their ability to perform each of seven activities of daily living (feed oneself, comb hair, perform personal hygiene, button a shirt, open a door, tie shoes and rise from a chair).
Continuous and ordinal outcome variables were compared across groups using the Kruskal-Wallis rank test in the case of three or more groups and the Wilcoxon rank-sum test in the case of two groups. Associations between categorical variables were evaluated by chi-squared or Fisher’s exact tests, as appropriate. All the tests were two-sided and p-values ≤ 0.05 were considered to be statistically significant. Analysis was performed using JMP software version 6 (SAS Institute Inc., Cary, North Carolina).
According to the Regan-Morrey classification 51 fractures (49.5%) were type I; 34 (33.0%) type IA; 17 (16.5%) type IB; 29 (28.1%) were type II; 10 (9.7%) type IIA; 19 (18.4%) type IIB; 23 (22.3%) were type III; 10 (9.7%) type IIIA and 13 (12.6%) type IIIB. Dislocation of the elbow had occurred in 47.6% of cases. The characteristics of the patients and the fractures are summarised in Table I⇑.
The more severe fractures tended to be associated with additional trauma to the elbow (Table II⇓), with concomitant injury occurring in 69% of type I, 76% of type II, and 83% of type III. The treatment, complications and additional operations are summarised in Table I⇑ by type of fracture. The outcome at the final follow-up is given by type of fracture in Table III⇓ and by type of treatment in Table IV⇓.
The Regan-Morrey classification was associated with the final arc of flexion (p = 0.01) (Table III⇑). Treatment was also related to the severity of the fracture by this classification. Patients who had plate fixation had significantly higher pain (p = 0.01) and less movement (p = 0.02) than those who had conservative treatment or excision of the fracture, but these patients also had more severe fractures with 22 of 23 (96%) being of type-IIA or worse. Conservative treatment or excision gave better movement than open reduction and internal fixation, but these patients tended to have less severe fractures. Patients with concomitant injuries had a lower Mayo elbow performance score (p = 0.03), less extension (p = 0.03), greater pain (p = 0.007) and less pronosupination (p = 0.004) than those without associated injuries. The presence of a fracture of the olecranon was significantly associated with additional operations (53% vs 23%, p = 0.02). A fracture of the head of the radius (36% vs 20%, p = 0.07) and use of a plate for fixation (43% vs 24%, p = 0.06) showed trends toward association with additional operations, although these results did not reach statistical significance. Injury to the medial collateral ligament, the lateral collateral ligament or both (34% vs 16%, p = 0.03) and treatment of the fracture by fixation with a plate (43% vs 18%, p = 0.009), were significantly associated with a complication.
There was a statistically significant increased range of pronation and supination in those patients who did not have a fracture of the olecranon compared with those who did (p = 0.04). The presence of any ligamentous injury correlated with worse pain scores (p = 0.002). Supination was poorer in those who had ligamentous injuries (mean 68° vs 78°, p = 0.05).
A fracture of the head of the radius had the greatest impact on outcome. The range of movement, including extension (mean 8.6° vs 14.1°, p = 0.008), pronosupination (mean 166° vs 146°, p = 0.004) and pronation (mean 86° vs 77°, p = 0.002) was significantly better in the absence of such a fracture. The level of pain was also significantly worse in those with concomitant fracture of the head of the radius (p = 0.006).
Few large series have recorded the outcome after fracture of the coronoid process. Regan and Morrey1 have defined the types of fracture and documented the outcome. Based on this series, prolonged immobilisation was avoided. Further biomechanical and clinical studies have documented the role of the coronoid process in the stability of the elbow.2–4 There has been a trend towards more operative fixation of more severe fractures. In the original series Regan and Morrey1 classified these fractures into three types and most of their patients were treated conservatively. In the current series of 103 fractures, 44 type-I fractures were treated conservatively or excised. All type-III and 76% of type-II fractures were managed by operation, often using a buttress plate. Prolonged immobilisation was avoided. The range of movement obtained decreased with the severity of the injury. However, the Mayo elbow performance score in patients with type-II and type-III fractures was higher than that in the original series, perhaps because of the more aggressive treatment and avoidance of prolonged immobilisation.
Since patients with more severe types of fracture were more likely to be treated by more complicated surgery, it is difficult to determine the outcome and complications based on the method of treatment. However, four patients treated by plate fixation required capsular release and excision of heterotopic ossification. One had delayed union of the fracture which necessitated a revision procedure. This failed and he ultimately required a total elbow replacement. One developed avascular necrosis of the coronoid, and another required arthroscopic capsular release and debridement, open decompression and transposition of the ulnar nerve but symptomatic arthritis later led to total elbow replacement. One patient developed stiffness and required capsular release and heterotopic ossification excision. Other patients treated by buttress-plate fixation experienced failure of the internal fixation, recurrent instability requiring revision, complications with wound healing, ulnar neuropathy and formation of heterotopic bone. Given the wide spectrum of associated injuries, it is difficult to comment on the complications related to the patterns of fracture or treatment in isolation. However, our series highlights the difficulty in defining the best treatment for a particular fracture given that these injuries tend not to occur in isolation, but as a part of a range of injuries to the elbow.
Limitations relating to the current series include the retrospective nature and reliance on information retrieved from the notes for documentation of associated injuries and outcomes. We relied on the treating physician to diagnose associated injuries and it is possible that not all were detected. Likewise, the impact of a fracture of the coronoid process and its treatment on the overall outcome is complicated by the presence of other injuries. An isolated fracture is rare and, therefore, such conflict is unavoidable if consideration of any sizeable series is to be undertaken. Plain radiographs may not be adequate in some cases to fully describe the patterns of the fracture. The Regan-Morrey classification system does not account for obliquity in the patterns of fracture, which has been documented by CT.3 Nevertheless, this classification system remains relevant and is widely used.
Our series describes the recent treatment and outcome of fractures of the coronoid process. Fractures with a higher Regan-Morrey type tended to have a poorer outcome, as did those with associated injuries. Most patients with more severe fractures were treated operatively. An improvement in the outcome as documented by the Mayo Elbow Performance score was seen compared with that of the earlier series,1 perhaps related to an increased appreciation of the role of the coronoid process and the need for treatment. We also observed the possibility of over-treatment of less severe injuries and recommend the avoidance of operative treatment of type-I injuries.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
- Received September 5, 2008.
- Accepted February 18, 2009.
- © 2009 British Editorial Society of Bone and Joint Surgery