Historical evolution in the treatment of midshaft clavicle fracture

Article information

J Korean Shoulder Elbow Soc. 2021;.cise.2021.00346

Publication date (electronic) : 2021 August 19

doi : https://doi.org/10.5397/cise.2021.00346

Steven J. Frediani, Mariano E. Menendez, Nicholas R. Pagani, Scott P. Ryan

Department of Orthopedic Surgery, Tufts Medical Center, Tufts University School of Medicine, Boston, MA, USA

Corresponding Author: Nicholas R. Pagani Tufts University School of Medicine, Department of Orthopedic Surgery, Tufts Medical Center, 800 Washington St, TMC Box #306, Boston, MA 02111, USA Tel: +1-61-7636-5172, Fax: +1-***, E-mail: npagani@tutsmedicalcenter.org

Received 2021 June 1; Accepted 2021 July 5.

Abstract

INTRODUCTION

For decades, nonoperative techniques have dominated the treatment approach for clavicle fractures. In a landmark study published in 1960, Dr. Neer reported that the nonunion rate for midshaft clavicle fractures treated by closed methods was only 0.1% [1]. This, along with a study by Rowe [2] showing a nonunion rate of 0.8% after closed treatment, established a standard that went unchallenged for decades. However, more recent high-quality evidence has altered grading of treatment success by challenging the assumptions that union occurs in most cases and that nonoperative treatment is the only option.

UNDERSTANDING THE ORIGINAL LITERATURE

Neer’s study was a large retrospective cohort study that identified 18 cases of nonunion after either open or closed treatment of midshaft clavicle fractures. The study reported the following conclusions: (1) improper initial operation is more likely to cause bone nonunion than closed treatment in middle-third fractures, (2) only 0.1% of middle-third clavicle fractures treated nonoperatively result in nonunion, and (3) nonunion of outer-third fractures is caused by disruption of the coracoclavicular ligaments and is not seen when these ligaments remain intact.

The report by Neer did not include explanations for how or when nonunion was determined, the criteria used to determine operative versus nonoperative management, or how operative treatment was performed. This is likely because Neer’s study was not focused on the incidence of nonunion of midshaft clavicle fractures in operative and nonoperative cohorts, and less than a paragraph was dedicated to explaining that finding. The primary focus was classifying and treating cases of nonunion after initial treatment.

A study conducted eight years later by Dr. Carter Rowe provided a more in-depth methodology discussion, including indications for open fixation and criteria for determining non-union. More specifically, Rowe stated that the typically agreed upon period for determining nonunion is between 4 and 6 months, although he did not clearly state how nonunion was determined, described intramedullary fixation as his method of choice for internal fixation, and described the use of figure-8 straps and spicas for closed fixation. Two findings were inferred from his study: that open reduction is more likely to result in nonunion than is closed treatment, and that closed treatment results in a very low nonunion rate (less than 1%).

Conducting clinical research in the 1960s was challenging and might have contributed to these findings and subsequent clinical assumptions. Both Neer and Rowe published seminal studies on clavicle fractures that have generated curiosity and stimulated higher quality research on treatment outcomes for such injuries.

A NEW PERSPECTIVE

Around the turn of the century, researchers began to question the old literature, and the discussion has yielded some surprising results. While the old literature used radiographs to determine treatment success or failure, the newer literature focused on patient satisfaction in conjunction with objective measurements of range of motion and strength. For instance, Hill et al. [3] in 1997 interviewed a group of 52 patients with a displaced mid-clavicle fracture that was treated nonoperatively at a mean follow-up of 38 months after injury. Surprisingly, they found that 15% of patients had developed nonunion, and 31% reported unsatisfactory results. More specifically, 25% had residual pain that required medication, and 54% found the end result cosmetically displeasing. This report in 1997 was one of the earliest studies to consider the patient perspective when evaluating the treatment result.

In the early 2000s, more studies were undertaken to better characterize the rate of nonunion of midshaft clavicle fractures. In 2004, Robinson et al. [4] performed a prospective study of 868 patients with clavicle fractures that were treated nonoperatively. Of the 662 patients with 24-week follow up, the prevalence of nonunion was 6.2%. In 2004, Zlowodzki et al. [5] performed a systematic review of 2,144 fractures and compared the results of different treatments for managing midshaft clavicle fractures, with specific emphasis on completely displaced fractures. They found a nonunion rate of 5.9% in all fractures treated nonoperatively versus that of 2.5% in all operatively treated fractures. When looking specifically at completely displaced fractures, the nonunion rate was 15.1% in the nonoperative group and 2.2% in the operative group. These results contrasted those reported by Drs. Rowe and Neer four decades prior.

Around this same time, a number of randomized controlled trials that focused on operative versus nonoperative treatments was being published. In 2001, Smith et al. [6] reported a 24% nonunion rate in 100% displaced midshaft clavicle fractures treated nonoperatively and a 0% nonunion rate for the same fractures treated by plating. This study was conducted with 50 participants in both the nonoperative and operative groups. In 2007, a multicenter trial by The Canadian Orthopedic Trauma Society that randomized 132 patients with completely displaced clavicle fractures into operative and nonoperative groups showed a 3.2% nonunion rate in the operative group and a 14.2% nonunion rate in the non-operative group [7]. Furthermore, the authors found that the operative group had a significantly shorter mean time to union (16.4 weeks vs. 28.4 weeks; p=0.001) as well as superior Constant and Disability of the Arm, Shoulder, and Hand (DASH) scores at all time points.

The study by Hill et al. [3] was groundbreaking and methodologically rigorous at the time of its publication in 1997. It found remarkably high rates of residual pain, dissatisfaction, and nonunion that conflicted with previously reported results for nonoperative treatment of fractures. Their approach was more centered around patient satisfaction and return to full function with minimal pain, as evidenced by their use of questionnaires to better assess the patient’s perspective. These questionnaires might have set the precedent for using DASH and Constant scores to assess return to full function in later studies. While surgery for clavicle fractures has gained a more prominent role over the years, nonoperative treatment continues to be the most common treatment approach.

Although clinical management of clavicle fractures has changed over the past few decades, it is important to ask if the current approach is too extreme and could be further improved. Operative fixation of clavicle fractures is not without risk or complication, and we must account for this in future decisions to best adapt such treatment. In a systematic review of 11 studies, Wijdicks et al. [8] documented the prevalence of complications following plate fixation of displaced midshaft clavicle fractures. They found that wound infection rates, both deep and superficial, were at or below 10% in almost all studies surveyed, and all resolved with oral antibiotics. Furthermore, they found that neurovascular complications that included brachial plexus symptoms and regional pain syndrome ranged between 0% and 38%, although only seven of the studies surveyed reported on these symptoms. Additionally, the most common post-operative complications reported were implant related. Implant complications were reported in almost every study, with rates of irritation noted between 9% and 64% and rates of plate debridement/removal/revision falling between 9% and 53%. This second point is alarming because, across the 11 studies reviewed by Wijdicks, at best approximately 1 in 10 patients required a second operative intervention after fixation.

A handful of more recent studies has corroborated the results of the Wijdicks study. In 2018, Sawalha and Guisasola [9] published a retrospective study that reviewed 90 patients treated operatively for midshaft clavicle fractures. Of those 90 patients, 23 (25.5%) developed complications ranging from plate removal to superficial infection. Of those who experienced complications, nine underwent an operation to remove hardware and three underwent an operation to remedy a fixation failure. In all, 12 of the 90 patients (13.3%) required a follow-up operation. In a randomized controlled trial, Woltz et al. [10] published the outcomes of 86 midshaft clavicle fractures that were treated by operative fixation. By the one-year follow-up, hardware removal was performed in six patients for implant failure, two patients for deep infection, one for nonunion, and 14 patients for elective removal. Of the 84 patients that were available at follow-up, 9 (10.7%) underwent a second operation for adverse events, and 14 (16.7%) underwent elective hardware removal, for a total of 23 patients (27.4%) who underwent a second operation10. In recent decades, clinicians have opted for plate fixation over conservative treatment; however, based on the results of this report, more caution should be employed when considering operative treatment.

The historical evolution for treating midshaft clavicle fractures highlights the importance of the scientific method in limiting preconceived notions and biases. As surgeons and scientists, it is important to continue to cultivate curiosity, critically appraise the literature, and evaluate outcomes to provide the best approach for patients.

Article information Continued