Orthopedics

Feature Article 

Risk Factors for Failed Closed Reduction of Pediatric Supracondylar Humerus Fractures

John D. Beck, MD; John T. Riehl, MD; Blake E. Moore, MD; John H. Deegan, BS; Jennifer Sartorius, MS; John Graham, PhD; William M. Mirenda, MD

Abstract

The purpose of this retrospective study was to examine pediatric supracondylar humerus fractures at a Level I trauma center. Data were analyzed to identify risk factors associated with closed reduction failure. Closed pediatric supracondylar humerus fractures that were treated at the authors’ trauma center between October 1997 and January 2009 were reviewed. The main outcome variable was necessity of open reduction. To determine which factors were independently associated with a failed closed reduction, a multivariate logistic model was fit predicting open reduction status.

A total of 174 patients required operative treatment. Of these, 23 underwent open reduction and 151 underwent with closed reduction and percutaneous pinning. For patients who required open reduction, 39.1% had an associated injury compared with 14.6% of patients treated with closed reduction (P=.008). Average time from presentation to surgery was 4.1 hours in the open reduction and 6.3 hours in the closed reduction group (P=.049). Risk factors that significantly predicted failure of closed reduction were the presence of an associated injury, initial fracture displacement, and Gartland type III fracture (P=.008, .03, and .023, respectively).

Associated injury, large initial fracture displacement, and Gartland type III factures were statistically significant independent risk factors for closed reduction failure. Increased time from injury to presentation demonstrated a trend toward open reduction. Consideration should be given to the expedient transfer of patients with type III supracondylar humerus fractures with associated injuries when definitive care will be provided at another institution.

Drs Beck, Riehl, Moore, Graham, and Mirenda, Mr Deegan, and Ms Sartorius are from the Department of Orthopaedics, Geisinger Medical Center, Danville, Pennsylvania.

Drs Beck, Riehl, Moore, Graham, and Mirenda, Mr Deegan, and Ms Sartorius have no relevant financial relationships to disclose.

Correspondence should be addressed to: John D. Beck, MD, Department of Orthopaedics, Geisinger Medical Center, 100 N Academy Ave, Danville, PA 17822 (jdbeck1@gmail.com).

Abstract

The purpose of this retrospective study was to examine pediatric supracondylar humerus fractures at a Level I trauma center. Data were analyzed to identify risk factors associated with closed reduction failure. Closed pediatric supracondylar humerus fractures that were treated at the authors’ trauma center between October 1997 and January 2009 were reviewed. The main outcome variable was necessity of open reduction. To determine which factors were independently associated with a failed closed reduction, a multivariate logistic model was fit predicting open reduction status.

A total of 174 patients required operative treatment. Of these, 23 underwent open reduction and 151 underwent with closed reduction and percutaneous pinning. For patients who required open reduction, 39.1% had an associated injury compared with 14.6% of patients treated with closed reduction (P=.008). Average time from presentation to surgery was 4.1 hours in the open reduction and 6.3 hours in the closed reduction group (P=.049). Risk factors that significantly predicted failure of closed reduction were the presence of an associated injury, initial fracture displacement, and Gartland type III fracture (P=.008, .03, and .023, respectively).

Associated injury, large initial fracture displacement, and Gartland type III factures were statistically significant independent risk factors for closed reduction failure. Increased time from injury to presentation demonstrated a trend toward open reduction. Consideration should be given to the expedient transfer of patients with type III supracondylar humerus fractures with associated injuries when definitive care will be provided at another institution.

Drs Beck, Riehl, Moore, Graham, and Mirenda, Mr Deegan, and Ms Sartorius are from the Department of Orthopaedics, Geisinger Medical Center, Danville, Pennsylvania.

Drs Beck, Riehl, Moore, Graham, and Mirenda, Mr Deegan, and Ms Sartorius have no relevant financial relationships to disclose.

Correspondence should be addressed to: John D. Beck, MD, Department of Orthopaedics, Geisinger Medical Center, 100 N Academy Ave, Danville, PA 17822 (jdbeck1@gmail.com).

Supracondylar humerus fractures are common injuries, accounting for more than 30% of all fractures in children younger than 7 years.1 Neurovascular injuries are common and can result from the fracture itself or from an iatrogenic complication from treatment. Associated neurovascular injuries commonly involve the anterior interosseous nerve, brachial artery, median nerve, or radial nerve.2 Complications from fracture or surgery include pin loosening, pin-site infection, Volkmann ischemic contracture, progressive cubitus varus with rotational malalignment, and, most commonly, reduced range of motion.3–5

The timing of definitive treatment of supracondylar humerus fractures has been debated. Some studies have demonstrated an increased risk of morbidity and complications with delayed treatment of supracondylar humerus fractures, specifically an increased need for open reduction.6–12 Open reduction results in the increased morbidity associated with open procedures, such as infection, scar formation, and an increased risk of iatrogenic neurovascular injury.11 Aside from time to treatment, little in the literature addresses other factors that may influence whether open reduction is performed.

The authors’ institition is a rural tertiary referral hospital with Level I trauma status. Subsequently, many supracondylar fractures are transferred from great distances for definitive management, leading to an inherent treatment delay. On review of these injuries, the authors’ open reduction rate was higher than some published standards.6–8,10,11,14,21 The current study examined the authors’ experience with supracondylar humerus fractures over an 11-year period. The hypothesis was that commonalities exist among patients who required open treatment that may be used to predict the need for open treatment in future patients.

Materials and Methods

Patient Population

Institutional review board approval was obtained for this study. The authors reviewed the records of consecutive patients with operatively treated pediatric supracondylar humerus fractures who were admitted between October 1997 and January 2009. The authors’ institution is located in a rural setting and is a referral center for pediatric trauma for a large geographic region. All patients with open growth plates who were treated for a supracondylar humerus fracture during the study period were included. Patients with open fractures were excluded from the study.

Patients were identified by the EPIC medical record system (Epic Systems, Verona, Wisconsin) using the International Classification of Diseases, Ninth Edition code for supracondylar humerus fracture. In addition, patients with a Current Procedural Terminology code for closed or open reduction of supracondylar humerus fracture during the study period were collected. Medical records for those patients were reviewed to confirm correct documentation. A total of 183 patients with operatively treated supracondylar humerus fractures were identified; 9 of these patients were excluded from the study because they had open fractures, leaving a final study population of 174 patients.

Data Collection

The medical records were reviewed, and data collected for each patient included demographic variables; mechanism of injury (high energy, fall from a height or bike; low energy, fall from standing); transfer from another hospital; open vs closed reduction and pin fixation; Wilkins modification of the Gartland fracture classification; flexion vs extension displacement; time of injury; time of presentation to the emergency department; time of surgery; presence of associated injuries, including nerve injury or palsy, ipsilateral upper extremity fractures, and vascular injury; attending specialty on call at time of patient arrival; and attending specialty that performed the operation.

Calculated data included time from injury to presentation to the emergency department, time from injury to surgery, and time from presentation to surgery. The main outcome variable was the need for open reduction. In the 119 patients with presentation radiographs, the direction and maximum displacement of the distal fracture fragment was evaluated and calculated in millimeters. This number was then normalized by dividing the maximum displacement in millimeters by the width of the humerus just proximal to the fracture site.

Indications for open reduction included a fracture irreducible by closed reduction, vascular compromise not improved with closed reduction, and nonacceptable reduction. A nonacceptable reduction was defined as excessive rotational or translational malalignment. Greater angular malalignment was accepted because this has a greater potential for correction through remodeling. Rotational malalignment was defined on a lateral radiograph when a 2- to 3-mm difference existed in the width of the bone at the fracture site between the proximal and distal fragments (Figure 1). Translational malalignment was defined on a lateral radiograph when less than 50% cortical contact existed between the proximal and distal fragments (Figure 2).

Lateral radiograph showing rotational malalignment.

Figure 1: Lateral radiograph showing rotational malalignment.

Lateral radiograph showing translational malalignment.

Figure 2: Lateral radiograph showing translational malalignment.

Statistical Analysis

Distributions of data were checked. Percentages were used for categorical data, and means or medians were used for continuous data (depending on distribution). The open reduction group was compared with closed reduction patients on collected variables. Categorical variables were tested using chi-square tests and continuous variables using t or Wilcoxon rank sum tests (depending on distribution). To determine whether factors were independently associated with risk of open reduction, a multivariate logistic model was fit predicting open reduction status. All variables were included as possible predictors; a final model was found using stepwise elimination. Final modeling results are reported with odds ratios (OR) and 95% confidence intervals (CIs). All analyses were performed using SAS version 9 software (SAS Institute Inc, Cary, North Carolina), and a P value less than .05 was considered statistically significant.

Results

A total of 174 patients met the inclusion criteria. Of these, 149 (85.6%) were transferred to the authors’ institution from another hospital (Table). Thirty-one (13.2%) patients had type II fractures and 140 had type III fractures. Twenty-three patients required open reduction. One hundred thirty-three (76.4%) patients presented between 5 pm and 7 am. The breakdown of attending specialty on call and specialty of the attending performing the surgery is presented in the Table.

Patient Characteristics

Table: Patient Characteristics

Of the 23 patients undergoing open reduction and fixation, 39.1% had an associated injury, whereas 14.6% of the patients undergoing closed reduction and percutaneous pinning had an associated injury; this difference was statistically significant (P=.008). In the open reduction group, associated injuries included 4 anterior interosseous nerve palsies, 2 ipsilateral distal radius fractures, 2 pulseless hands, and 1 radial nerve palsy (Table).

Initial displacement was a significant risk factor for open reduction (P=.03). When the displacement was normalized by cortical width, the open reduction group had a greater mean displacement, although the difference was no longer significant (Table). Eighty-seven percent of the patients undergoing open reduction had a high-energy injury mechanism compared with 68.7% of the closed reduction group (P=.08). Time from injury to surgery was not significantly different between the open and closed reduction groups (P=.91). Average time from injury to presentation was 8.1 hours in the open reduction group and 6.5 hours in the closed reduction group (P=.15). The average time from presentation to the emergency department to surgery was 4.1 hours in the open reduction group and 6.3 hours in the closed reduction group (P=.049) (Table).

Results from the logistic modeling showed that none of the following were significantly related to predicting the risk of open reduction: presentation type (direct to the authors’ institution or transfer from another hospital), time of presentation (7 am–5 pm vs 5 pm–7 am), attending specialty on admission, surgeon specialty, and time from injury to surgery. The significant factors in the logistic modeling were presence of an associated injury (OR=3.3 [95% CI, 1.4–7.7]; P=.008), Gartland classification (P=.023), and time from presentation to surgery (P=.049). Patients with an associated injury had significantly higher (3×) odds of having an open reduction than those with no associated injury.

Discussion

Although several authors have advocated early reduction and pinning of pediatric supracondylar humerus fractures,3,6–8,12,13 other studies have reported no increased benefit of early reduction and pinning.7–11 Proponents of early treatment claim less of a need to convert to open reduction for these fractures. The current study describes the experience at a rural tertiary referral center, using a multivariate logistic model to determine which factors were independently associated with an increased incidence of treating these fractures with open reduction.

The incidence of open reduction varies in the literature from 0.6% to 46%. 6–8,10,11,14–21 The current rate (13.2%) is consistent with previous reports.6–8,10,11,14,21 When the mechanism of injury was evaluated in the current patients, the high-energy injury rate was consistent with previous reports of supracondylar humerus etiology.16 Indications for open reduction included a fracture irreducible by closed reduction, vascular compromise not improved with closed reduction, and non-acceptable reduction.22 Standardization of indications of when to transition from closed to open reduction attempts to eliminate surgeon preference or comfort from biasing rates of open reduction and allows the regression analysis to determine other variables that predict closed reduction failure. As a tertiary referral center, the current authors’ instution sees various injuries that present with a multitude of patient variables. This allowed them to perform a statistical analysis on a large cohort to determine factors affecting the open reduction rate of pediatric supracondylar humerus fractures.

The variables that significantly and independently influenced the need for open reduction in the current study were the presence of an associated injury (39.1% vs 14.6%, respectively; P=.008), Gartland type III fracture (P=.023), maximum initial fracture displacement (P=.03), and time from presentation to surgery (P=.049). The most common coexisting injuries were anterior interosseous nerve palsy and an ipsilateral forearm fracture. An associated injury increased the open reduction rate to 3 times greater than that for patients without associated injury. The presence of an associated injury or greater initial fracture displacement implies a higher-energy injury. This suggests that these children should be transferred to a hospital that can accommodate their total care. In addition, if an associated injury or a large degree of initial fracture displacement is present, the possibility of open reduction should be emphasized during preoperative planning for these fractures. The lack of fracture displacement in patients who have a Gartland II supracondylar humerus fracture with an associated injury may explain why they did not require open reduction. Although the need to convert to open treatment is a possibility for any surgically treated supracondylar humerus fracture, the treating surgeon should be aware of the increased likelihood in patients with associated injuries or a large degree of initial fracture displacement.

Similar to the findings of Farnsworth et al,16 the majority (72.7%) of the current patients presented between 5 pm and 7 am. The presentation time had no statistically significant effect on the need for open reduction. Similarly, the time from injury to surgery and injury to presentation had no statistically significant effect on the need for open reduction. This is consistent with the findings of Leet et al,8 who reported no increased incidence of complications (including open treatment) with a longer time from injury to surgery. However, the current authors found a statistically significant increase in the need for open reduction and pinning in the time from presentation to surgery. This decreased time was related to expedited care after patients presented due to their trauma level. The clinical significance is that in patients with greater trauma, such as Gartland type III fractures with associated injuries, for whom definitive treatment is expected to take place at another institution, early and expedient transfer should be arranged.

Limitations of the current study include the study design in which injury and results data were obtained retrospectively; therefore, the final numbers depended on the accuracy of documentation. Second, the study patients were identified using International Classifcation of Diseases, Ninth Edition and Current Procedural Terminology codes. However, this method of identifying patients excluded all patients with pediatric supracondylar humerus fractures that were coded incorrectly. Third, initial presentation radiographs were not always available for analysis because the hard copy radiographs had been lost or destroyed. Finally, a selection bias was likely present because patients with more difficult fracture patterns would have been sent to pediatric-trained orthopedic surgeons if available.

Conclusion

Gartland type III supracondylar distal humerus fractures, the presence of associated injury, and greater initial fracture displacement were significant risk factors for the need for open reduction. Treating surgeons should be aware of this increased incidence and be prepared to treat these injuries accordingly. In addition, consideration should be given to the early and expedient transfer of patients with Gartland type III supracondylar humerus fractures with associated injury or significant initial fracture displacement when definitive care will be provided at another institution.

References

  1. Cheng JC, Shen WY. Limb fracture patterns in different pediatric age groups: a study of 3,350 children. J Orthop Trauma. 1993; 7(1):15–22. doi:10.1097/00005131-199302000-00004 [CrossRef]
  2. Lyons ST, Quinn M, Stanitski CL. Neurovascular injuries in type III humeral supracondylar fractures in children. Clin Orthop Relat Res. 2000; (376):49–55.
  3. Celiker O, Pestilci FI, Tuzuner M. Supracondylar fractures of the humerus in children: analysis of the results in 142 patients. J Pediatr Trauma. 1990; 4(3):265–269.
  4. Cheng JC, Lam TP, Shen WY. Closed reduction and percutaneous pinning for type III displaced supracondylar fractures of the humerus in children. J Orthop Trauma. 1995; 9(6):511–515. doi:10.1097/00005131-199509060-00009 [CrossRef]
  5. Wilkins KE. The operative management of supracondylar fractures. Orthop Clin North Am. 1990; 21(2):269–289.
  6. Walmsley PJ, Kelly MB, Robb JE, Annan IH, Porter DE. Delay increases the need for open reduction of type-III supracondylar fractures of the humerus. J Bone Joint Surg Br. 2006; 88(4):528–530. doi:10.1302/0301-620X.88B4.17491 [CrossRef]
  7. Gupta N, Kay RM, Leitch K, Femino JD, Tolo VT, Skaggs DL. Effect of surgical delay on perioperative complications and the need for open reduction in supracondylar humerus fractures in children. J Pediatr Orthop. 2004; 24(3):245–248. doi:10.1097/01241398-200405000-00001 [CrossRef]
  8. Leet AI, Frisancho J, Ebramzadeh E. Delayed treatment of type 3 supracondylar humerus fractures in children. J Pediatr Orthop. 2002; 22(2):203–207. doi:10.1097/00004694-200203000-00014 [CrossRef]
  9. Sibinski M, Sharma H, Bennet GC. Early versus delayed treatment of extension type-3 supracondylar fractures of the humerus in children. J Bone Joint Surg Br. 2006; 88(3):380–391. doi:10.1302/0301-620X.88B3.17181 [CrossRef]
  10. Iyengar SR, Hoffinger SA, Townsend DR. Early versus delayed reduction and pinning of type III displaced supracondylar fractures of the humerus in children: a comparative study. J Orthop Trauma. 1999; 13(1):51–55. doi:10.1097/00005131-199901000-00012 [CrossRef]
  11. Mehlman CT, Strub WM, Roy DR, Wall EJ, Crawford AH. The effect of surgical timing on the perioperative complications of treatment of supracondylar humeral fractures in children. J Bone Joint Surg Am. 2001; 83(3):323–327.
  12. Segal D. Pediatric orthopedic emergencies. Pediatr Clin North Am. 1979; 26(4):793–802.
  13. Paradis G, Lavallee P, Gagnon N, Lemire L. Supracondylar fractures of the humerus in children: technique and results of crossed percutaneous K-wire fixation. Clin Orthop Relat Res. 1993; (297):231–237.
  14. Cramer KE, Devito DP, Green NE. Comparison of closed reduction and percutaneous pinning versus open reduction and percutaneous pinning in displaced supracondylar fractures of the humerus in children. J Orthop Trauma. 1992; 6(4):407–412. doi:10.1097/00005131-199212000-00002 [CrossRef]
  15. Danielsson L, Pettersson H. Open reduction and pin fixation of severely displaced supracondylar fractures of the humerus in children. Acta Orthop Scand. 1980; 51(2):249–255. doi:10.3109/17453678008990794 [CrossRef]
  16. Farnsworth CL, Silva PD, Mubarak SJ. Etiology of supracondylar humerus fractures. J Pediatr Orthop. 1998; 18(1):38–42. doi:10.1097/00004694-199801000-00008 [CrossRef]
  17. Hart GM, Wilson DW, Arden GP. The operative management of the difficult supracondylar fracture of the humerus in the child. Injury. 1977; 9(1):30–34. doi:10.1016/0020-1383(77)90046-8 [CrossRef]
  18. Mangwani J, Nadarajah R, Paterson JM. Supracondylar humeral fractures in children: ten years’ experience in a teaching hospital. J Bone Joint Surg Br. 2006; 88(3):362–365. doi:10.1302/0301-620X.88B3.16425 [CrossRef]
  19. Mehlman CT, Crawford AH, McMillion TL, Roy DR. Operative treatment of supracondylar fractures of the humerus in children: the Cincinnati experience. Acta Orthop Belg. 1996; 62(suppl 1):41–50.
  20. Peters CL, Scott SM, Stevens PM. Closed reduction and percutaneous pinning of displaced supracondylar humerus fractures in children: description of a new closed reduction technique for fractures with brachialis muscle entrapment. J Orthop Trauma. 1995; 9(5):430–434. doi:10.1097/00005131-199505000-00012 [CrossRef]
  21. Van Laarhoven CJ, Oosterhuis KJ. Operative treatment of supracondylar fractures of the humerus in children. Neth J Surg. 1990; 42(2):61–62.
  22. Reitman RD, Waters P, Millis M. Open reduction and internal fixation for supracondylar humerus fractures in children. J Pediatr Orthop. 2001; 21(2):157–161. doi:10.1097/00004694-200103000-00004 [CrossRef]

Patient Characteristics

CharacteristicOpen Reduction (n=23)Closed Reduction (n=151)P
Women, %47.849.7.86
Mean age at injury, y5.7±2.85.3±2.4.38
Extension injury, %95.696.5.85
Associated injuries, %39.114.6.008a
Transferred to current authors, %95.680.4.083
7 am–5 pm presentation time, %34.825.2.33
Gartland classification, %.023a
  Type I02.1
  Type II020.7
  Type III10077.1
Attending specialty, %.55
  General26.119.9
  Pediatrics39.131.1
  Sports21.717.9
  Trauma4.49.9
  Other8.721.2
Surgeon specialty, %.22
  General17.419.2
  Pediatrics56.539.7
  Sports17.410.6
  Trauma0.08.6
  Other8.721.9
Median time (IQR), h
  Injury to presentation8.1 (7.0, 14.9)6.5 (4.0, 17.5).15
  Injury to surgery15.5 (11.4, 19.0)14.7 (9.1, 24.4).91
  Presentation to surgery4.1 (1.5, 6.9)6.3 (2.9, 14.8).049a
High-energy fractures,b %87.068.7.08
Mean maximum displacement,c mm17.5±10.711.6±8.0.03a
Mean maximum displacement,c normalized by cortex width, mm/mm0.75±0.580.61±0.43.35
Direction of displacement,c %.15
  Lateral5073
  Anteroposterior5027

10.3928/01477447-20120919-18

Sign up to receive

Journal E-contents