Predictors Associated with the Need for Open Reduction of Pediatric Supracondylar Humerus Fractures

Authors: M. Bryant Transtrum, Diego Sanchez, Shauna Griffith, Brianna Godinez, Vishwajeet Singh, Kyle J. Klahs, Amr Abdelgawad, Ahmed M. Thabet

Abstract

Background:

Supracondylar humerus (SCH) fractures are some of the most common fractures in pediatric patients with surgery typically consisting of either open or closed reduction with internal fixation. The aim of this meta-analysis was to identify patient, injury, and administrative factors that are associated with treating pediatric SCH fractures with open techniques.

Methods:

Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, PubMed and CINAHL database searches were conducted for studies from 2010 to 2023 that made direct comparisons between open reduction and internal fixation (ORIF) and closed reduction and percutaneous pinning (CRPP) for treating SCH fractures in the pediatric population. The search terms used were “pediatric” AND “SCH fracture” OR “distal humerus fracture.” Screening, quality assessment, and data extraction were performed by 4 reviewers. After testing for heterogeneity between studies, data were aggregated using random-effects model analysis.

Results:

Forty-nine clinical studies were included in the meta-analysis. Summated, there were 94,415 11,329 treated with ORIF and 83,086 treated with CRPP. Factors that were significantly associated with greater rates of ORIF included obesity (p = 0.001), Gartland type IV fractures (p < 0.001), general neurological deficits (p = 0.019), and ulnar nerve deficits (p = 0.003). Gartland type II (p = 0.033) and medially displaced fractures (p = 0.011) were significantly associated with lower rates of ORIF. Secondary analysis showed cross-pinning constructs (p = 0.033) and longer hospital stays (p = 0.005) are more likely to be observed in patients undergoing ORIF compared with CRPP.

Conclusion:

This meta-analysis demonstrates that factors such as obesity, fracture displacement, and concomitant nerve deficits are more likely to require ORIF as opposed to CRPP.

Level of Evidence:

Therapeutic Level III.

Introduction

Supracondylar humerus (SCH) fractures are some of the most common fractures in pediatric patients and can present within a wide spectrum of severity^1^. If displaced, distal humerus fractures typically require surgical intervention to reduce the fracture and provide internal stability while healing^2-4^. Traditionally, closed reduction and percutaneous pinning (CRPP) is the first-line treatment for SCH fractures, with the more invasive open reduction and internal fixation (ORIF) being considered only after closed methods prove unsuccessful^3,5,6^. Factors such as patient demographics, injury characterization, and treatment administration likely all influence the ultimate treatment modality.

Patient demographics, such as age, race, body mass index (BMI), and gender, regardless of fracture pattern, may predispose to either a successful CRPP or conversion to ORIF. Previous studies have shown that age may be directly proportional to the likelihood of converting to open techniques^7^. Obese patients (defined as BMI ≥ 30.0) have also been found to undergo ORIF at higher rates than their nonobese cohorts^8-11^. Historically, male gender was believed to be positively associated with high-energy SCH fracture frequency and ORIF^12,13^; however, recent studies have challenged these findings^14,15^.

Injury patterns, such as open vs. closed fractures, fracture pattern, and extremity neurovascular status have all been implicated in surgical decision-making. The access to and visualization of the fracture site in these cases may result in the reasonable use of ORIF techniques as is seen in most open fractures treated in one previous study^16^. The Gartland classification system has been widely used to characterize SCH fracture displacement and may be a tool to easily predict the need to open^17-19^. A concomitant injury to surrounding nerves or vessels has also been associated with open exploration and ORIF^20,21^. Furthermore, the timing of injury presentation to the operative surgeon may also affect the ability to achieve a successful closed reduction^22^.

Other administrative factors, such as time to treat, treatment location, and surgeon experience, may also influence the treatment of SCH fractures. Community hospitals, tertiary referral centers, and dedicated pediatric hospitals are all sites of service for the treatment of SCH fractures. The potential correlation between treatment site and SCH fracture treatment type has not been previously investigated in the literature. The years of experience an orthopaedic surgeon has in practice as well as pediatric fellowship training may result in a nuanced understanding of and increased comfort with SCH fractures and their treatment. Several previous studies report that non–fellowship-trained surgeons used ORIF at a higher rate than those who underwent a pediatric orthopaedic surgery fellowship^23-28^.

Pediatric SCH fractures are one of the most prevalent operative injuries in children and are, therefore, highly studied; however, findings of individual studies are often inconsistent and difficult to generalize. Consequently, there remains a paucity in the literature concerning predictive demographic, injury, and administrative factors that may influence the decision to convert from CRPP to ORIF.

Materials and Methods

Search Strategies

This systematic review and meta-analysis were structured and written in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We identified and retrieved relevant studies from electronic databases PubMed and CINAHL databases. Search terms included “pediatric” AND “SCH fracture” OR “distal humerus fracture.” Search filters included text full text; article 0 to 18 years; English; and publication January 1, 2010, to March 15, 2023. Four reviewers independently screened titles, abstracts, and article types for eligibility. Studies were selected for further review if they were relevant to the operative treatment of SCH fractures using open or closed reduction techniques.

Study Selection

Randomized controlled trials, cohort studies, and case series were included if they met the following predefined inclusion available data for the treatment of SCH fractures using ORIF and CRPP patient groups and sufficient results for data extraction, i.e., the number of subjects for each patient group was provided. Studies were excluded if they did not contain a pediatric focus, were written as a systematic review, or were only available in languages other than English. The same 4 reviewers independently performed full-text screening, excluding articles that were not relevant to the primary treatment of SCH fractures in pediatric populations or that did not make direct comparisons between ORIF and CRPP groups; only studies containing homogenous open and closed treatment groups could be included in the meta-analysis. In addition, studies were excluded if they described treatment protocols that did not meet North American standards of care. Most studies were retrospective comparative studies with level III evidence. The methodological items for nonrandomized studies (MINORS) instruments were used for the quality assessment of the included studies (Table I). This tool consists of 12 questions aimed at appraising multiple factors. Items are scored as 0 (not reported), 1 (reported but inadequate), or 2 (reported and adequate). The maximum score for noncomparative studies is 16 and for comparative studies is 24.

Preoperative Predictive Factors

Preoperative factors of interest included all reported demographic, injury, and administrative factors that were compared between CRPP and ORIF. Demographic factors included age, gender, ethnicity, race, and BMI. Injury characteristics comprised the affected arm sidedness and dominance, Gartland classification, open vs. closed injury status, fracture patterns, and neurovascular status. Finally, administrative factors included the time to treat, treatment location, and treating surgeon's pediatric fellowship status and years of experience.

Secondary Outcomes

In addition to preoperative factors, data related to intraoperative and postoperative outcomes were also collected. Intraoperative outcomes included operative time, surgical approach, and fixation constructs. Postoperative outcomes comprised duration of hospital length of stay, radiographic and functional outcomes, complications, and time to latest follow-up.

Data Extraction

A standardized data extraction sheet was used to gather information including the aforementioned preoperative factors and secondary outcome measures. Four authors independently extracted data. Any disagreement was adjudicated by a fifth author.

Data Analysis

Statistical software STATA (version 17; StataCorp) was used for the analysis. Data from 3 or more studies were required for any factor to be included in our meta-analysis. The extent of heterogeneity was performed based on the I^2^ statistics, where I^2^ ≤ 50% suggests moderate heterogeneity and I^2^ ≥ 50% indicates high heterogeneity. A random-effects model was used to obtain the pooled effect size values as odds ratio (OR) or standardized mean difference (SMD) with a corresponding 95% confidence interval (CI). A p value < 0.05 was considered statistically significant.

Results

After conducting a preliminary search with our search terms, a total of 4,217 articles were identified. 3,326 articles were eliminated from search filters for the following language other than English, not available in full text, outside of the publication range, and outside the pediatric age range. Screening resulted in 222 additional exclusions because of the removal of articles that were not relevant to our study or article types that were reviews, systematic reviews, or meta-analyses. After in-depth review of the remaining 669 studies, 621 were excluded for either having the wrong intervention, no reported ORIF vs. CRPP data, or unacceptable treatment protocol. The final total was 48 studies eligible for analysis. One study, Bell et al., separated out 2 groups each containing an open and closed treatment group, and so, it was counted as 2 separate studies, bringing the total to 49 study groups^7^. A total of 94,415 patients, with 11,329 in the ORIF group and 83,086 in the CRPP group, across all 49 studies were eligible for meta-analysis. A PRISMA flow diagram of studies from search through screening to inclusion and exclusion outlines each step of the review process (Fig. 1). Also, a quality assessment of the included studies using the MINORS instrument was conducted (Table II).

The eligible 49 studies all reported the proportion of cases with open and closed reduction^7,10,25-50,51-70^. The pooled overall proportion obtained for ORIF was 0.16 (95% CI = 0.14, 0.18; Table III; Fig. 2). High heterogeneity was observed between studies reporting proportion of open reduction (I^2^ = 99.0%, p < 0.001). The factors that met our inclusion criteria for the meta-analysis are subsequently listed along with their study characteristics (Tables III–V).

Preoperative Factors

Obesity

Three studies were included in the analysis evaluating patient BMI^10,50,62^. Obese patients demonstrated a significantly higher likelihood of requiring ORIF (OR = 1.87; 95% CI = 1.27, 2.75; p = 0.001) with no heterogeneity (I^2^ = 0.0%, p = 0.72) (Table IV).

Fracture Displacement

Thirteen studies were included in the analysis evaluating Gartland type II fractures^31,34-36,41,43,45,49,56,58,60,66,68^. Patients with Gartland type II fractures demonstrated a significantly lower likelihood of requiring ORIF (OR = 0.31; 95% CI = 0.10, 0.91; p = 0.033) with high heterogeneity (I^2^ > 50%, p < 0.001) (Table IV).

Six studies were included in the analysis evaluating Gartland type IV fractures^28,34,40,48,57,68^. Patients with Gartland type IV fractures demonstrated a significantly higher likelihood of requiring ORIF (OR = 5.97; 95% CI = 2.96, 12.04; p < 0.001) with moderate heterogeneity (I^2^ = 44.2%, p = 0.11) (Table IV).

Three studies were included in the analysis evaluating medial fracture displacement^45,51,62^. Patients with medially displaced fracture patterns demonstrated a significantly lower likelihood of requiring ORIF (OR = 0.35; 95% CI = 0.15, 0.78; p = 0.011) with no heterogeneity (I^2^ = 0.0%, p = 0.79) (Table IV).

Neurological Deficits

Seven studies were included in the analysis evaluating general preoperative neurological deficits^32,35,45,46,62,63,70^. Patients with general preoperative nerve deficits demonstrated a significantly higher likelihood of requiring ORIF (OR = 3.43; 95% CI = 1.22, 9.65; p = 0.019) with high heterogeneity (I^2^ > 50%, p = 0.001) (Table IV).

Specifically, 3 studies were included in the analysis evaluating preoperative ulnar nerve deficits^39,62,70^. Patients with preoperative ulnar nerve deficits demonstrated a significantly higher likelihood of requiring ORIF (OR = 13.54; 95% CI = 2.39, 76.84; p = 0.003) with high heterogeneity (I^2^ > 50%, p = 0.021) (Table IV).

No statistical significance was observed in the relationships between ORIF and the following preoperative • Demographic: patient age (SMD = 0.17) and gender (OR = 0.96).• Injury: left-sided injury (OR = 0.73), Gartland type III fractures (OR = 1.60), flexion-type injury (OR = 2.35), extension-type injury (OR = 0.45), closed injury (OR = 0.18), and median nerve deficits (OR = 0.85).• Administrative: pediatric fellowship status of operating surgeon (OR = 0.92).

Secondary Intraoperative/Postoperative Outcomes

Fixation Constructs

Four studies were included in the analysis evaluating crossing k-wire configurations^28,35,37,70^. A significant, positive association was observed between crossing k-wire constructs and ORIF (OR = 15.87; 95% CI = 1.25, 201.39; p = 0.033) with high heterogeneity (I^2^ > 50%, p = 0.001) (Table V).

Hospital Stay

Three studies were included in the analysis evaluating hospital length of stay^29,67,70^. A significant, positive association was observed between hospital length of stay and ORIF (SMD = 0.41; 95% CI = 0.12, 0.69; p = 0.005) with moderate heterogeneity (I^2^ = 49.7%, p = 0.14) (Table V). The average hospital stay for patients treated with ORIF was 1.9 days compared with 1.2 days for those treated with CRPP.

No statistical significance was observed in the relationships between ORIF and the following intraoperative/postoperative operation duration (SMD = 0.48), Flynn functional criteria (OR = 91), Flynn cosmetic criteria (OR = 0.73), normal anterior humeral line (OR = 75), impaired range of motion (OR = 1.00), overall complications (OR = 0.85), malunion (OR = 1.16), infection (OR = 3.11), and revision surgery (OR = 1.73).

Discussion

The purpose of this study was to determine preoperative predictive factors associated with the utilization of ORIF techniques as opposed to CRPP in pediatric SCH fractures. We conducted a systematic review and meta-analysis of all studies published and available on PubMed from January 1, 2010, to March 15, 2023, to determine whether identified preoperative factors contributed to operative modality as well as commonly reported outcomes. We found that obese patients were significantly more likely to undergo an ORIF compared with nonobese patients. Higher Gartland class and preoperative neurovascular deficits were also associated with a higher probability of needing ORIF. In addition, cross-pinning constructs and longer duration of hospitalization were positively associated with the ORIF group.

Obesity was positively predictive for ORIF likely because of a higher mechanism of energy based on increased patient mass as compared to those nonobese^71,72^. Childhood obesity also contributes to bone mineral metabolism, which increases not only the risk of fracture but also the displacement risk^73^. Multiple studies have identified pediatric obesity to generate more complex and adult-like fracture patterns^74-76^. Li et al. found obesity to be a risk factor for ORIF because of the increased difficulty with closed manipulation beneath a large, traumatized soft-tissue envelope^10^. The obese patient presents many challenges to the orthopaedic surgeon that may influence the decision to expand direct fracture visualization and ensure construct stability.

Fracture characteristics were significantly related to the ultimate surgical method used to treat SCH fractures. Higher Gartland class, especially type IV, where the posterior osseous hinge and periosteal stability were eliminated, was more likely to require ORIF to stabilize the fracture while securing with internal fixation. Simple, routine, SCH fracture patterns such as Gartland type II and III, were not associated with an increased need to open and were likely more easily close-reduced before internal fixation.

Preoperative general neurological deficits and specifically ulnar nerve palsies were associated with using open operative techniques. Nerve palsies typically represent a larger, more diffuse soft-tissue injury surrounding the fracture that occurs in a predictable pattern^77-79^. Larger soft-tissue injuries facilitate more edema, swelling, and potentially a more difficult environment for closed fracture manipulation and maintenance of reduction, requiring other, more invasive techniques. A study by Sun et al. investigating only flexion-type fractures identified ulnar nerve palsies as a predictive factor for ORIF^62^. Although flexion-type SCH fractures were not found to be predictive of ORIF in our study, ulnar nerve palsies, which are typically accompanied by flexion-type fractures, were associated with ORIF.

When assessing intraoperative techniques there was heterogeneity in describing and labeling certain configurations. Reported surgical approaches were almost equally divided between posteromedial and posterolateral approaches to the distal humerus. This likely represents large rotational deformity in the fracture pattern, necessitating opening manually and obtaining a reduction before internal fixation^80^. More than half of open techniques that reported on pin configuration used cross pins. Cross-pin configuration, although inherently imposes a higher risk of iatrogenic nerve injury, has been shown to be biomechanically superior to alternative pin configuration^81-83^. Within these evaluated studies, the surgeons were likely faced with very unstable fracture pattern types in those cases requiring open techniques and, therefore, took on more risk with cross-pins to achieve greater fracture stability.

There is an increasing trend to treat pediatric SCH fractures in an ambulatory setting^84-86^. It has been shown to be not only cost-effective, but also safe, with very low rates of postoperative infections and complications^84-86^. Our study demonstrated longer hospital stay for those patients undergoing ORIF, likely because of continued postoperative prophylactic antibiotics and evaluation of neurovascular recovery.

There were many limitations to this meta-analysis, to include limitation of eligible studies during our data collection period (2010-2023), a high heterogeneity on both preoperative demographic information as well as reported outcomes, and scarcity of adverse events making direct comparisons difficult. Most studies included in this meta-analysis were level III evidence retrospective comparative studies. Ideally, prospective studies would be preferred for meta-analysis; however, available prospective studies are limited in the literature. In addition, although we included studies on pediatric patients from 0 to 18 years of age, generalizing our findings to skeletally mature adolescents may be inappropriate because many surgeons may opt to directly treat this population with open reduction and fixation with plate constructs as they would with an adult injury. Studies on these patients would likely fall outside the limits of those included in our analysis because pinning constructs are typically used in younger, skeletally immature children. These limitations contributed to some loss of power within our analysis and prohibited the direct comparisons of some factors.

Areas for future research include expanding our data collection period to capture a larger number of potentially eligible prospective studies that directly compare SCH fractures treated with CRPP vs. ORIF.

Conclusion

This meta-analysis demonstrates that factors such as obesity, fracture displacement, and concomitant nerve deficits are more likely to require ORIF as opposed to CRPP.