The management of clavicular fractures has changed substantially in the last fifteen years. In 1997, Hill et al. documented poor results and increased risk of nonunion in clavicular fractures with initial shortening of ≥20 mm1. McKee et al. confirmed this finding in a study involving more sophisticated outcome assessments including Constant and DASH (Disabilities of the Arm, Shoulder and Hand) scores as well as muscle strength testing2. However, the 2007 publication of the results of a randomized clinical trial conducted by the Canadian Orthopaedic Trauma Society3 showing improved outcomes with surgical management of displaced clavicular fractures appears to represent the “pivot point,” following which clinical practice began to be more commonly surgical.
Despite the accumulating evidence of the potential for compromised functional outcome and the increased risk of nonunion in patients with a displaced clavicular fracture, the role of surgery remains uncertain. Even if the nonunion rate is as high as 20%, it remains true that four of five patients with a displaced clavicular fracture will have healing of the fracture, treatment of a nonunion is relatively straightforward, and functional outcomes remain acceptable in many patients despite malunion. For these reasons, many investigators have tried to better define surgical indications by investigating risk factors for poor outcomes. Nowak et al. reviewed 245 adult patients with a clavicular fracture and noted that lack of osseous contact and fracture comminution with a transverse fragment were strong predictors of adverse sequelae whereas fracture location and shortening alone were not4. In 2004, even before the Canadian trial was completed, a group from Edinburgh reported study data that could be used to estimate the likelihood of clavicular nonunion5. In a consecutive series of 868 patients, the nonunion rate of diaphyseal clavicular fractures was just 4.5%, and advancing age, female sex, fracture comminution, and lack of cortical apposition were independent predictors of nonunion5.
Now that operative management of displaced clavicular fractures has become more common, it is even more important to be sure that surgery is appropriately utilized. Since displaced fractures are considered the fractures that are particularly at risk for adverse sequelae, the Edinburgh group has repeated their 2004 study but instead focused their attention solely on this group of injuries. Using much more sophisticated statistical analyses, the authors find that smoking status (yes/no), comminution (yes/no), and fracture displacement in millimeters can be used to predict the risk of nonunion. By calculating the absolute risk difference among different groups of patients and thereby defining the number needed to treat (NNT), the authors are able to provide clinicians with very clear indicators of the potential benefit of surgery in subsets of patients with various well-defined characteristics.
Clinicians treating fractures of the clavicle should familiarize themselves with the findings of this paper and incorporate the authors’ predictive model into their preoperative discussions. Data such as those in the study by Murray et al. provide a strong evidence base that lends itself to a shared decision-making process by presenting risks in a manner that is more quantifiable and easily understood by patients.