CORR Insights®: What Are the Functional Outcomes and Pain Scores after Medial Clavicle Fracture Treatment?

Where Are We Now?

In 2007, the Canadian Orthopaedic Trauma Society developed a multicenter, prospective, randomized study that showed improved patient-reported outcome measures (PROMs) and decreased prevalence of symptomatic nonunion and malunion after open reduction and internal fixation compared to nonsurgical management for displaced, midshaft clavicle fractures [3]. Since the publication of that study, surgical treatment for this injury has become more common [5].

It is not clear, though, whether the purported benefits of surgical treatment are noticeable to patients. It appears that many of the reported PROM differences between patients whose fractures healed with or without surgery have not achieved the minimum clinically important difference (MCID) threshold for the assessment tools used [2]. Research on midshaft clavicle fracture treatment, therefore, has been focused on identifying patient- or fracture pattern–related risk factors for nonunion and malunion among patients treated nonoperatively [8]. In general, current smoking history and fracture patterns with comminution and/or significant fragment displacement were independently associated with the development of nonunion. Knowledge of these factors has helped to better delineate which patients would benefit from early surgical intervention.

There is less of an emphasis on treatment strategies and clinical outcome measurements for fractures of the medial clavicle, likely owing to the far lower prevalence of these injuries compared to their mid-shaft counterparts [1, 9]. Furthermore, I’m not aware of any comparative studies of surgical versus nonsurgical management for medial clavicle fractures, and the use of validated PROMs is rare even in case series [9]. Salipas et al. [9] reported excellent American Shoulder and Elbow Surgeons and Subjective Shoulder Value scores at a mean of 3 years postinjury, though it must be emphasized that the questionnaires were completed in < 50% of included patients, a percentage similar to the current study. Nearly 77% of the patients that were assessed for radiographic and/or clinical union were considered united. The authors found that the two instances of delayed union occurred in patients younger than 65 years with >1 cm of initial displacement. A subsequent systematic review based on seven case series and 10 case reports determined that nonsurgical management resulted in a nonunion rate of 5% [1]. It would seem that, similar to their midshaft counterparts, most medial clavicle fractures, even those with displacement, result in good clinical and radiographic outcomes.

Lindsey and colleagues [6], in a retrospective analysis of 31 patients with medial clavicle fractures treated predominately without surgery over a 9-year period, reported favorable functional outcomes as measured by the QuickDASH and visual analog pain scores at a mean of 5 years postinjury among surviving patients. The authors also found that more than one-third of patients with this injury died, which they attributed to the older age of patients who sustained these higher energy fractures. Finally, they were unable to find associations between patient- and fracture-related demographics and functional outcomes. While the only two instances of radiographic nonunion were found in patients who did not have surgery, only a small percentage of the patients had sufficient imaging to make a meaningful assessment of this outcome measure.

This study gives us good material to use when counseling patients with medial clavicle fractures: If they are older or have serious concomitant injuries, there is a real risk of death; however, if they can survive the initial trauma, there is every reason to expect good clinical and functional outcomes, regardless of whether surgical or nonsurgical management is chosen.

Where Do We Need To Go?

In light of the apparent lack of clinically important benefits of surgery with regard to pain or function for most patients with midshaft clavicle fractures, but with the risk of nonunion among those treated without surgery, we might focus next on how to anticipate which patients are at greatest risk of nonunion. We know that fractures with severe displacement are at risk, but we also know that most patients even with displaced fractures still will heal without surgery [2, 7]. The development of a symptomatic nonunion or malunion following nonsurgical management, which may affect up to 15% of patients with midshaft fractures, may necessitate surgical intervention that is more complex in the delayed setting [4, 7].

While clinical outcomes as measured by pain and PROMs are generally favorable following nonsurgical management of medial clavicle fractures, it is not entirely clear whether these injuries behave in a similar fashion to their midshaft counterparts with regard to fracture union. The current study raises some important questions that require further study: (1) Is there a notable difference in the prevalence of symptomatic nonunion and/or malunion for both operatively and nonoperatively treated displaced medial clavicle fractures? (2) If so, how can we predict the development of a symptomatic nonunion and/or malunion based on patient- and fracture-specific parameters? If we can progress toward answering these questions, the orthopaedic community can more objectively determine which patients would benefit from early surgical intervention.

The paper by Salipas and colleagues [9] provides some foundation for answering the above questions by categorizing initial fracture displacement, and then associating patient age and fracture displacement with the development of a delayed union. However, rigorous analysis of these and other patient- or fracture-related (comminution, shortening) demographics as predictors of nonunion was not performed. Some of the notable differences between the current paper and that of Salipas and colleagues [9] are the inclusion of medial clavicle fractures with midshaft and lateral extension in the latter study and the differing definitions utilized for both delayed union and nonunion. Furthermore, it is imperative that future studies evaluate the radiographic outcomes of the same fracture type so we are truly comparing “apples to apples.” It is clear that we need more work in defining objective parameters for offering patients early surgical intervention for medial clavicle fractures.

Asadollahi and Bucknill [1] were able to demonstrate that the majority of these fractures are minimally or nondisplaced and reported a nonunion prevalence of < 5% when treated without surgery. However, initial fracture displacement and pattern could not be universally determined from the included studies, precluding analysis of their effect on radiographic outcomes. While their study would seem to indicate that most medial clavicle fractures achieve union without surgery, the answer as to which factors were associated with nonunion remains elusive.

How Do We Get There?

There is little doubt that the path toward understanding which medial clavicle fractures would benefit from early surgical versus nonsurgical management is fraught with great difficulty, in no small part due to the low prevalence of these fractures, the high mortality rate precluding long-term follow-up, and the fact that many of these fractures are nondisplaced or minimally displaced. To determine which patient- and fracture-related demographics are associated not just with patient-derived outcomes as outlined by Lindsey and colleagues [6], but also with radiographic outcomes, there is no question that a randomized controlled trial would provide the most clinically useful data. However, for the reasons outlined above, such an endeavor is unlikely to be feasible and may not even be necessary to advance our current understanding of these fractures. A more practical solution would be to perform an observational study of non-surgically managed medial clavicle fractures such as that for midshaft fractures by Murray et al. [8], and statistically evaluate the association of both patient and radiographic factors with nonunion development. We need to focus on medial fractures without lateral extension, and perhaps more importantly, ensure that adequate radiographic follow-up is obtained among surviving patients. Standardized collection of known risk factors for clavicle nonunion would then allow potential pooling of data from future individual studies. Of course, this would undoubtedly need to be performed at a Level 1 trauma center to achieve sufficient patient numbers. The use of the proposed Fracture and Trauma Registry through the American Academy of Orthopaedic Surgeons as a central data repository may help overcome some of the logistical difficulties with this endeavor. Lindsey and colleagues [6] ought to be commended for providing some clarity regarding the clinical and radiographic outcomes for this relatively uncommon injury.