Clavicle fracture management remains controversial in the orthopedic community. A multicenter randomized clinical trial by the Canadian Orthopedic Trauma Society in 2007 demonstrated improved functional outcomes with lower rates of nonunion and malunion with open reduction internal fixation (ORIF) compared with nonoperative treatment [1]. A more recent meta-analysis of randomized controlled trials demonstrated reduced risk for nonunion with ORIF, albeit with no improvement in functional outcomes [2]. At present, the American Academy of Orthopedic Surgeons Clinical Practice Guidelines (AAOS CPG) offer a moderate-strength recommendation for operative rather than nonoperative management of displaced midshaft fractures, based on evidence of higher union rates and improved early patient-reported outcomes with surgery. However, the AAOS CPG notes that either operative or nonoperative management may be considered, on an individual basis, given similar long-term patient-reported outcomes and satisfaction [3].

The degree of displacement often factors into decision-making for operative versus conservative treatment. Relative indications to operative management include 100% fracture displacement or ≥2 cm of radiographic shortening/displacement [3]. Therefore, multiple studies have evaluated the most appropriate method to determine displacement. Upright clavicle radiographs (UCRs) have demonstrated increased vertical displacement compared to supine radiographs due to gravitational forces and are routinely obtained at some institutions [4-6]. Drawbacks to obtaining these images include increased radiation exposure, healthcare costs, patient discomfort, and resource utilization [7,8]. Despite these drawbacks and evidence suggesting that UCR findings may have a limited influence on ultimate management decisions, they continue to be routinely performed at many institutions, including the study institution. Therefore, it is essential to obtain a clearer understanding of their actual impact on clinical decision-making to justify their continued use.

Recent studies have explored the utility of upright radiographs in the evaluation of clavicle fractures, particularly regarding their influence on surgical management. Hoogervorst et al. [5] demonstrated that upright chest radiographs more accurately revealed fracture displacement and shortening compared to supine imagine, suggesting they may better reflect functional deformity under the influence of gravity. Furthermore, studies have reported that upright radiographs influenced treatment decisions in nearly one third of midshaft clavicle fractures cases, underscoring their potential role in guiding operative versus nonoperative cases [6]. Similarly, Herman et al. [9] found that upright shoulder radiographs altered clinical management in a significant proportion of cases, with an increase in measured displacement and shortening often prompting a shift toward surgical treatment. Nevertheless, the degree to which these imaging findings independently drive surgical decision making, versus reflecting surgeon preference or institutional practice patterns, remains uncertain.

At our level 1 trauma center, clavicle fractures are commonly diagnosed with an initial supine clavicle radiographs (SCR) or chest/abdomen/pelvis computed tomography (CAP CT) imaging during the primary trauma evaluation. Once identified, routine inpatient UCR is typically obtained and evaluated using institutional protocol to determine further fracture displacement and potential indications for surgical intervention. Given the variability in practice, this study aimed to evaluate the frequency with which inpatient UCR altered clavicle fracture management at our institution and to determine whether these additional images meaningfully influenced treatment decisions. The primary hypothesis was that UCR would rarely demonstrate additional displacement exceeding the commonly cited 2-cm surgical threshold. The secondary hypothesis was that UCR findings would rarely lead to changes in operative versus nonoperative management at our facility.

This study was conducted in compliance with the principles of the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Review Board of Saint Louis University (No. 33810). Informed consent was waived given the retrospective nature of the study.

We conducted a retrospective review of all clavicle fracture patients presenting to a single level 1 trauma center from July 2021 to November 2023. All patients were initially identified from a hospital database for orthopedic trauma consults. Patients who met the inclusion criteria were aged over 18 years and under 89 years (per institutional protocol) with clavicle fractures identified on initial SCR or CAP CT during primary trauma evaluation. To meet the inclusion criteria, patients must have been admitted and must have received UCR as inpatients. Patients with bilateral clavicle fractures were included, and either injury was considered on an individual basis. Exclusion criteria included patients with chronic, pathological, or open fractures, inadequate imaging, or incomplete documentation.

For patients meeting the inclusion criteria, electronic medical records (EMR) were queried to collect data, including patient age at the time of injury, gender, fracture laterality, and fracture location. Clavicles were measured on UCR and divided into medial, middle, and lateral thirds by measuring the overall clavicle length and dividing it into three equal segments. Injury data collected included the mechanism of injury and the presence of polytrauma. Orthopedic polytrauma patients were defined as those who sustained a clavicle fracture with at least one concomitant orthopedic fracture of the axial or appendicular skeleton, including spine, pelvis, long bone, scapula, hand, and foot fractures. Isolated hand and/or foot fractures not limiting weight-bearing were not classified as polytrauma. These patients were analyzed together with isolated fracture cases to better reflect the real-world heterogeneity of trauma presentations and to assess whether concomitant injuries influenced the likelihood of operative intervention. For example, patients with ipsilateral lower extremity fractures, such as femur or acetabular injuries, may have been more likely to undergo clavicle fixation to facilitate early weight-bearing with assistive devices on the affected side. Polytrauma patients and their management are discussed in detail later in the study. However, it is important to note that the management of polytrauma patients may be different from the management of those with isolated fractures and that the emphasis of our study was on the routine use of UCR for the latter.

Initial radiographic and CT imaging were reviewed by a fellowship-trained orthopedic traumatologist and an orthopedic surgery resident. Clavicle fractures were classified by laterality and location (medial, middle, and lateral). Absolute vertical displacement was measured and compared for the SCR and UCR of each patient by research team members and averaged. Inter and intra-observer reliability values were not calculated. Absolute vertical displacement was quantified using the methodology described by Hoogervorst et al. [10], based upon the distance between the superior cortex of the medial fragment and the superior cortex of the lateral fragment. The methodology is shown in Fig 1. The initial displacement on SCR and UCR was recorded, along with the increase in vertical displacement between SCR and UCR. Additionally, the number of patients who had an increase in shortening from <2 cm on the SCR to ≥2 cm in the UCR was recorded in each group, as this, along with 100% fracture displacement, has commonly been used as a relative indication for surgical management in the literature [11-13].

After the overall data were recorded, it was subdivided according to whether patients were treated with nonoperative or operative management. Among patients who underwent ORIF, whether the procedure occurred during their index hospitalization or as an outpatient was recorded. Data was additionally collected to assist in identifying the indication for ORIF. This included an increase in vertical displacement from SCR to UCR, presence of polytrauma, impending open fractures, concern for scapulothoracic dissociation, additional shoulder injuries with double disruption to the superior shoulder suspensory complex (SSSC), or persistent pain in the follow-up clinic setting. EMR documentation was evaluated for narrative descriptions of surgical indications.

Additional subgroup analysis was performed using a 1-cm threshold of initial displacement on supine radiographs. This threshold was chosen as a cutoff with the assumption that 1 cm roughly equals a 100% displacement. The cohort was stratified into two groups: those with <1 cm and those with ≥1 cm of initial displacement on SCR. Average displacement on supine and upright radiographs, as well as the mean change between them, were compared between the two groups.

Statistical analysis was performed using Microsoft Excel version 16.83. Categorical data were presented by number and percentage. Continuous data were reported as a mean of vertical displacement in millimeters. Comparisons between groups of continuous data, particularly the increases in vertical displacement of clavicle fractures from SCR to UCR for patients who underwent nonoperative management versus ORIF, were made using the unpaired t-test. A P-value of <0.05 was used to determine statistical significance. Chi-square analysis was used for the evaluation of categorical data. In addition, 95% CIs for mean differences were calculated to better characterize the magnitude of observed differences and strengthen statistical interpretation.

Ultimately, 230 charts of patients with clavicle fractures were reviewed, with 160 patients meeting the inclusion criteria. Of these, 70 patients were excluded, most commonly due to the absence of SCR/UCR imaging and inadequate documentation. For those patients meeting the inclusion criteria (Table 1), the mean age was 47.0 years, 112 (70%) were males, and 48 (30%) were females. The most common mechanisms of injury included motor vehicle collision (40%) and falls (26%). Left-sided clavicle fractures were found in 84 patients (52%). All clavicle fractures were unilateral. Medial, middle, and lateral fracture locations were 8 (5%), 117 (73%), and 35 (22%), respectively. Of the 160 patients, 78 patients (49%) were found to have additional fractures or were considered polytrauma patients. Among the polytrauma patients, the most common concomitant injuries were long bone fractures, as they occurred in 29 of the 78 patients (37%). Concomitant pelvic fractures were found in 15 patients (19%), and concomitant spine injuries were found in 22 patients (28%). The remaining 12 patients (16%) had a variety of upper and lower extremity injuries. These concomitant injuries were important to note because they imposed weight-bearing restrictions that necessitated the use of assistive devices for ambulation.

As illustrated in Table 2, the average initial vertical displacement measured on SCR and UCR for all patients was 6.0±7.1 mm (95% CI, 4.9–7.1 mm) and 10.5±9.4 mm (95% CI, 9.1–11.9 mm), respectively. The average change in vertical displacement from SCR to UCR for all patients was 4.5±6.3 mm (95% CI, 3.6–5.5 mm) and found to be statistically significant (P<0.001). On further analysis, 144 patients (90%) underwent nonoperative management and 16 patients (10%) underwent ORIF. The mean age of patients managed nonoperatively was 48.9 years (range, 18–89), which was statistically older than the mean age of patients who underwent operative management with a mean age of 29.4 years (range, 18–48) (P<0.001). Of the 16 patients managed surgically, 8 underwent ORIF as inpatients and 8 underwent surgical intervention on an outpatient basis. As illustrated in Table 2, the average change in vertical displacement from SCR to UCR was 4.5±6.5 mm (95% CI, 3.5–5.6) for patients managed nonoperatively and 5.1±4.5 mm (95% CI, 2.7–7.5) for those managed operatively. The change within the nonoperative group was found to be statistically significant (P<0.001), whereas within the operative group it was not (P=0.06). Subsequent analysis comparing the average change between the nonoperative group and the operative group was found to be not statistically significant (P=0.71). Of the 160 patients, UCR demonstrated that 20 (12.5%) had a change of vertical displacement from <2 to ≥2 cm. Of those 20 patients, four (20%) underwent ORIF. Only one of these four patients underwent ORIF based on the change in vertical displacement as the documented primary indication, while the others underwent operative interventions primarily based on the need for mobilization related to polytrauma or persistent pain.

As shown in Table 3, 16 patients (10%) underwent surgical management for their clavicle fractures. Four patients managed surgically underwent inpatient ORIF based on indications independent of UCR findings. These findings included scapulothoracic dissociation, impending open fracture, and double disruption to the SSSC (Table 3). An additional two patients managed surgically underwent inpatient ORIF due to difficulty with mobilization in the setting of polytrauma, requiring the use of their upper extremities for ambulation with an assistive device. Both patients were able to tolerate weight-bearing immediately after ORIF.

Two out of a total of 160 patients (1.3%) underwent inpatient ORIF due to changes in vertical displacement on UCR. One patient had 10.4 mm of vertical displacement on SCR and 19.3 mm of vertical displacement on UCR, with a change of 8.9 mm. The other had 6.5 mm of vertical displacement on SCR and 20.3 mm of vertical displacement on UCR, with a change of 13.8 mm. Ultimately, these were the only two patients who underwent surgical intervention based on a change in vertical displacement on UCR as the primary documented indication. The specific displacement threshold used to indicate surgery was not documented. The eight patients who underwent outpatient ORIF had a mean SCR vertical displacement of 15.0 mm and a UCR of 18.9 mm with a mean change of 3.9 mm. All were documented as having persistent pain as the primary indication for ORIF.

As summarized in Table 4, we performed a subgroup analysis of patients with <1 cm and ≥1 cm of initial displacement on SCR, and found that 115 patients (72%) had initial displacement of <1 cm and 45 (28%) had initial displacement of ≥1 cm. For patients with an initial displacement of <1 cm on SCR, the average displacement found on upright radiographs was 7.08±6.94 mm (95% CI, 5.81–8.35). For patients with an initial displacement of ≥1 cm, the average displacement found on upright radiographs was 16.78±9.36 mm (95% CI, 14.05–19.51). The average change in displacement between the two groups showed no statistical difference (P=0.440). The proportion of patients who underwent operative management was significantly higher in the group with ≥1 cm of initial displacement (P=0.001).

The primary purpose of this study was to assess whether findings on routine inpatient UCR led to changes in operative intervention for patients sustaining clavicle fractures. Consequently, only two out of 160 (1.3%) clavicle fractures proceeded to surgical intervention following documented changes in vertical displacement on routine inpatient UCR. Although several studies have investigated the change in displacement between SCR and UCR of clavicle fractures, there is comparatively less literature on the impact of routine UCR on fracture management on an inpatient basis [4-6]. One such exception was a study by Herman et al. [9], who published data on 365 patients evaluated with supine radiographs only versus supine and upright radiographs. They found both groups had similar rates of operative intervention. However, the supine radiograph only group more commonly had delayed surgery based on persistent pain, whereas the supine and upright radiograph group more commonly had earlier operative intervention based on the increase in vertical displacement noted on upright radiographs. In comparison, our study found similar reasons for delayed surgery based on persistent pain on an outpatient basis, but minimal occurrences of inpatient surgery due to findings of vertical displacement on UCR.

Among the eight patients who underwent inpatient ORIF, six of the indications for operative intervention were reasons other than a change in displacement on UCR. One patient demonstrated clinical and radiographic concern for scapulothoracic dissociation and persistent fracture distraction. While guidelines for management are not definitive, scapulothoracic dissociation is recognized as a limb-threatening entity that typically is managed with surgical intervention [14-16].

An additional patient had skin tenting with an impending open fracture, which is also traditionally considered a relative indication for operative intervention [17,18]. Two patients had a concurrent displaced clavicle and scapula fracture, indicating a possible double disruption to the SSSC. While controversial, some advocate for clavicle ORIF to reestablish clavicular length and indirectly stabilize the scapula fracture [19]. Regardless, this was recognized on SCR and CT before UCR and documented as the indication for clavicle ORIF. Two patients underwent clavicle ORIF in the setting of polytrauma to aid with mobilization. One patient had a concomitant pelvic ring injury and the other a concomitant lower extremity long bone fracture.

At our institution, displacement alone rarely results in the surgical management of clavicle fracture patients, despite the relative radiographic indications of 100% displacement or ≥2 cm of shortening/displacement [3]. Although data suggest a higher nonunion rate in clavicle fractures with displacement treated nonoperatively, a meta-analysis of randomized controlled trials by Woltz et al. [2] found similar long-term functional outcomes and rate of reoperation compared to patients treated surgically.

Building upon these conclusions, our study found that only four of 20 patients with an increase in vertical displacement from <2 cm on SCR to ≥2 cm on UCR were managed with ORIF. One of these four patients underwent ORIF based primarily on displacement the other three patients had surgery related to mobilization in the setting of polytrauma and persistent pain. Although there was a statistically significant difference in vertical displacement between the average SCR and UCR overall (P<0.001), the core finding of our study was that there was no significant difference between operative and nonoperative patients (P=0.71). This supports the notion that routine use of upright radiographs contributes minimal additional value in guiding management decisions for clavicle fractures. Moreover, the statistically significant change in vertical displacement was only 4.5 mm, which is small in the context of clavicle fracture management, where relative indications for surgery are about 20 mm of displacement, nearly five times greater [5,6]. As such, although the 4.5-mm increase reached statistical significance, it is unlikely to reflect a clinically meaningful alteration in fracture alignment that would influence treatment decisions for most patients.

A subgroup analysis was conducted using an initial vertical displacement threshold of <1 cm versus ≥1 cm on SCR. For the 115 (72%) patients with initial displacement of <1 cm on SCR, the average displacement found on upright radiographs was 6.98 mm, well below the relative indications of 2 cm or 100% displacement. Furthermore, the ≥1 cm group had an average initial vertical displacement of 15.55 mm, most likely negating the need for additional imaging with UCR to dictate fracture management.

It is worth mentioning that of the 115 patients who demonstrated less than 1 cm of displacement on initial supine radiographs, 37 (32%) were found to have greater than 1 cm of displacement on UCR and 7 (6%) had greater than 2 cm of displacement on subsequent UCR imaging. Out of the 160 patients, 45 (28%) showed displacement changes on UCR that may have changed fracture management, given the relative indications for surgical management of clavicle fractures. UCR may add diagnostic value when patients are approaching relative indications based on fracture displacement and clinical assessment, but routine use may not be justified. Previous studies have reported that upright radiographs have a strong influence on management decisions. For instance, Hoogervorst et al. [5] and Herman et al. [9] both demonstrated that UCR findings led to operative intervention in a significantly larger proportion of cases, suggesting that institutional practice patterns and surgeon preference may play a substantial role in whether UCR results translate into clinical action.

We found that patients who underwent nonoperative clavicle fracture management were significantly older than those treated with ORIF, with a mean age of 48.9 years and 29.4 years, respectively (P<0.001). There is sparse literature investigating the impact of age on clavicle treatment. Medina Perez et al. [20] investigated factors affecting patient decision-making for midshaft clavicle fractures through patient surveys and found that respondents who were ≥33 years of age were more likely to choose nonoperative treatment, which was significantly higher than patients <33 years of age (P<0.001). Additional studies have also suggested that older patients are less likely to pursue operative intervention, potentially due to lower functional demands, comorbidities, or differences in perceived surgical risk [21]. Patient preference may have been a contributor to the discrepancy in age between nonoperative and operative treatment in this study.

In addition to limited utility, there is the added cost, radiation exposure, and pain related to routine upright imaging to consider. The U.S. Food and Drug Administration has advised a reduction in unnecessary radiation exposure by only ordering imaging to answer a medical question, help treat a disease, or guide a procedure [22]. Previous analyses have estimated that standard clavicle radiographs deliver approximately 0.1 mSv of radiation, roughly equivalent to one day of background radiation, while advanced imaging such as CT exposes patients to 3–4 mSv [23]. Although the individual risk is minimal, the cumulative cost and exposure across large trauma populations can be substantial, underscoring the importance of judicious imaging use. Beyond radiation, obtaining UCR may contribute to patient discomfort, particularly in the acute traumatic setting where arm mobility is limited and upright positioning can be painful. Additionally, the need for upright films increases demands on a radiology department that already operates under significant resource constraints at our institution, further impacting workflow and contributing to delays in care. Furthermore, we found routine use of inpatient UCR rarely altered treatment and resulted in unnecessary additional imaging in most patients. The cost of an upright 2-view clavicle radiograph costs 340 dollars at our institution for a patient without medical insurance [24]. Finally, undergoing upright imaging may cause additional pain related to fracture displacement and may be difficult and uncomfortable for patients who are acutely injured.

This study does contain several limitations. Only 58% of patients had a documented follow-up where repeat clavicle radiograph could be reviewed. While this limits evaluation of fracture healing with nonoperative management, it should not alter the conclusion from the primary purpose of this study, which was to assess the utility of routine inpatient UCR on operative management. Another potential limitation was variability in displacement measurements and the possibility of human error. Although radiographic measurements were obtained by multiple trained team members and averaged to help reduce the impact of individual variability, the absence of formal inter-observer and intra-observer reliability testing indicates that measurement inconsistency may still be present. Nevertheless, the average measurements provided a reliable general assessment of fracture displacement that supports the overall conclusions of the study. Furthermore, the retrospective design of our study introduces potential selection and information bias, and the absence of functional outcome measures such as union rate, shoulder strength, or patient-reported outcomes must be noted. These factors limit the ability to assess the broader clinical impact of UCR use and reduce external validity.

The most significant limitation of this study, however, lies in the inherent subjectivity of surgical decision-making. The determination of operative intervention is not based on rigid, universally applied criteria, but rather on each surgeon’s personal threshold for what constitutes clinical significance of displacement, patient demand, or soft-tissue compromise. As a result, the decision to proceed with surgery may vary between providers, even for similar radiographic findings, introducing a potential source of confounding that cannot be fully controlled in a retrospective analysis. The authors recognize a more conservative approach to clavicle fracture management at our institution, despite a changing landscape pushing toward acute clavicle fracture ORIF, namely in the polytrauma scenario. Therefore, any apparent lack of influence of upright radiographs on surgical decision-making in our data may reflect institutional and individual surgeon preference rather than a purely objective assessment of radiographic displacement. Future studies incorporating multi-institutional data, standardized operative criteria, or blinded case review may help reduce this bias and better isolate the independent impact of radiographic findings on management decisions.

The purpose of this study was not to deter the decision for operative intervention. Rather, we aimed to demonstrate that routine inpatient use of UCR was unlikely to alter decisions for operative management of inpatient clavicle fractures and should be selectively used to guide treatment.

Inpatient UCR rarely changes the management of clavicle fractures. For patients who underwent inpatient ORIF, the increase in displacement seen on UCR was not significantly different from that observed in those managed nonoperatively. While these findings suggest that routine use of inpatient UCR may offer limited additional value in most cases, certain borderline or clinically equivocal fractures, particularly those nearing relative surgical thresholds, may still benefit from upright imaging to better inform management decisions. Overall, the use of UCR should be applied judiciously, with clinical judgment and individual patient factors guiding imaging selection rather than routine protocol.