Open Vs. Closed Reduction in Type 2 Lateral Condyle Fractures

Abstract

Background:

Lateral condyle fractures are associated with high morbidity due to their risk of nonunion and avascular necrosis (AVN). This study aims to assess the outcomes between closed reduction and the more traditional open techniques for operative fractures.

Methods:

All lateral condyle fractures that required operative fixation (pins or screws) over a ten-year period were included. We compared open versus closed reduction for OR time, infection rate, AVN, nonunion, premature physeal closure, ulnohumeral angle, and interepicondylar width (IEW).

Results:

28 patients were identified in the closed reduction group while 41 were identified in the open reduction group. Average displacement at surgery for these two groups was significantly different at 3.95mm for the closed group and 9.47mm for the open group (p<0.0001). Operating room time was significantly greater for the open reduction group by an average of 45 minutes (p <0.0001). Additionally, the closed reduction group was significantly less likely to require postoperative admission compared to the open reduction group (p=0.0004). There were no significant differences between the two groups with regard to abnormal ulnohumeral angles, infection rates, avascular necrosis, nonunion, lateral spur formation, premature physeal closure, or reoperation rate.

Conclusions:

The significant differences in OR time and post-operative admission make closed reduction the preferred approach from a hospital system quality improvement standpoint.

Level of Evidence: IV

Introduction

Lateral condyle fractures are the second most common pediatric elbow fracture, comprising up to 20% of them.¹ Not only are they relatively common, but they are also associated with high rates of morbidity, including nonunion, avascular necrosis (AVN), and premature physeal closure.^2,3 Due to these risks, it has been recommended that they be managed operatively if they are displaced by more than 2mm based on the Jakob classification.^1,3-5

Surgical fixation may be completed in a number of ways, including with an open approach, a closed approach, with K-wire fixation, or with screw fixation.^6-8 There has been debate concerning the adequacy of open versus closed fixation.¹ Some have classically thought that an open approach was necessary in order to be certain that the articular surface was properly reduced; however, there have been some studies that show a closed reduction was not inferior in outcomes when performed on patients with 4mm or less of maximal displacement.^1,7,8 As closed approaches continue to gain popularity, we seek to evaluate their cost-effectiveness and re-examine the 4mm cut-off as an indication of their use.

Methods

After obtaining the appropriate Institutional Review Board approval, patients were identified for this retrospective chart review between the years 2003 and 2013 who were under the age of 18 and had the following ICD-9 codes: 812.42 for a closed lateral condyle fracture, 812.52 for an open lateral condyle fracture, and 812.44 for an unspecified condylar fracture. Patients were excluded if they received multiple fractures in the ipsilateral upper extremity. Information regarding patient demographics, surgical methods, and follow-up were collected from the chart. Fractures were classified based on the Jakob criteria.^4,9

Operative cases were selected based on an initial maximum fracture displacement greater than 2mm, progression of maximum fracture displacement to greater than 2mm, or at least 21 days of cast immobilization with resulting non-union. The decision to approach fixation with an open or closed approach was at the discretion of the operating orthopedic surgeon. The use of K wires or screws was also done at their discretion and the majority of them were unburied and removed in the office approximately one month after surgery.

Figure 1.

Measurement of ulnohumeral angle, which displays a normal value of 10˚.

Figure 2.

AP elbow demonstrating a lateral spur. The interepicondylar width (IEW) was determined by measuring the distance of the metaphyseal spur from the medial epicondyle to the normal metaphysis.

Figure 3.

Avascular necrosis. The capitellum demonstrates fragmentation and early closure of the lateral physis.

Radiographic measurements, such as maximum fracture displacement, ulnohumeral angle, and interepicondylar width (IEW), were redone at the time of the study. Ulnohumeral angle was measured from the most recent AP of the ipsilateral elbow and was used as a measure of varus or valgus alignment. The range for the normal valgus alignment of the elbow used was 5°-15°. Measurements less than 5° were considered varus while angles more than 15° were considered valgus.

IEW is a measurement used to assess the severity of lateral spurring, which is a common radiographic abnormality after lateral condyle fracture.¹⁰ IEW indicated mild spurring if it was between 100% and 110%, moderate spurring if it was between 110% and 120%, and severe spurring if it was greater than 120%.10

Additionally, the presence of avascular necrosis or premature physeal closure was identified on the most recent follow-up radiographs. AVN was defined as either the presence of a fishtail deformity or the loss of a previously developing capitellum. Premature growth arrest was defined as bony bridging across a physis.

Infection was monitored and was analyzed in total and after segmental analysis looking at superficial wound infections and septic arthritis.

In order to assess statistical significance between the open and closed groups, several tests were employed. The Student’s T test was done when comparing mean values between the two groups. This was calculated with the use of StatPlus in Microsoft Excel.¹¹ In addition, the Mann-Whitney U test was done to compare values in which the median was thought to be a better comparison of the average than the mean.¹² Finally, Fisher’s Exact Test was computed for categorical data.¹³ All p-values used were two sided and significance was noted when p<0.05. For cases where patients needed a reoperation, intent to treat analysis was used.

Results

164 patients with lateral condyle fractures were identified in our search. Of those, 79 required surgical fixation. A total of 69 patients met our research criteria and were included in the study. 28 patients were identified in the closed reduction group while 41 were identified in the open reduction group. Median follow-up for the closed group was 9.5 months (1, 76), which was significantly longer than the open group, which was followed for 4 months (1.38, 43) post-operatively (p=0.02). Patient demographics are displayed in Table 1. The most common mechanism of injury was a direct fall onto the elbow with 43.6% of patients in the closed group and 41.4% of the open group reporting this. No patients in either group presented with an open fracture. Age at injury, gender, side of injury, and BMI were not significantly different between the two groups. The median maximum displacement at the time of presentation to our hospital system for the closed group was 2.5mm (range 0,6.32), which was significantly different compared to the open group, which had a median of 7.8mm (range 1.17, 24.3)(p <0.0001).

Table 1.

Patient Demographics and Original Injury Characteristics

	Closed (n = 28)	Open (n = 41)	P value
Age	5.16	6.0	0.08969
Gender	57.1% M	65.9% M	0.6135
Injured Side	71.4% L	63.4% L	0.6054
BMI	18.30	19.06	0.56277
Maximum Displacement	2.5 (range 0,6.32)	7.8 (range 1.17,24.3)	<0.0001
Jakob Criteria:			<0.0001
Jakob I	46.4% (n = 13)	7.3% (n = 3)
Jakob II	32.1% (n = 9)	17.1% (n = 7)
Jakob III	21.4% (n = 6)	75.6% (n = 31)

Comparisons at the time of surgery for the two groups are demonstrated in Table 2. Patients who underwent closed reduction of their fractures waited longer before surgery compared to the open reduction group with an average time of 9 days and 2 days, respectively (p <0.0001). At the time of surgery, the closed group had a median maximum displacement of 3.95mm, which was significantly different than the open group, which had a median maximum displacement of 9.47mm (p <0.0001). Although 100% of patients received prophylactic antibiotics at the time of surgery, cefuroxime was the preferred antibiotic used for both groups with 96.2% of the closed group receiving it and 85% of the open group receiving it (p=0.23). Surgical fixation was completed with either pins or screws. 78.6% of closed reductions used screw fixation while only 17.1% of open reductions used screws (p <0.0001). Additionally, the OR time required to complete the two surgical approaches was significantly different as the closed group needed just 78.6 minutes while the open group required 123.3 minutes (p <0.0001). Moreover, patients who had an open reduction were more likely to require overnight observation compared to those who underwent closed reduction (47.5% vs. 7.1%, p = 0.0004).

Table 2.

Comparisons at the Time of Surgery

	Closed (n = 28)	Open (n = 41)	P value
Time to surgery (days)	9 (range 1,29)	2 (range 0,82)	<0.0001
OR Time (min)	78.64	123.29	<0.0001
Maximum Displacement	3.95 (SD = 1.61)	9.47 (SD = 5.71)	<0.0001
Fixation Type	78.6% Screws	17.1% Screws	<0.0001
Admitted to Hospital	7.1%	47.5%	0.0004

Figure 4.

AP radiograph with pin fixation just prior to pin removal.

Follow-up and outcomes data are illustrated in Table 3. Median follow-up time was 4 months for the open group and 9.5 months for the closed group (p= 0.02). There were no significant differences between the two groups with regard to abnormal ulnohumeral angles, infection rates, avascular necrosis, nonunion, lateral spur formation, interepicondylar width or premature physeal closure.

Table 3.

Patient Follow-Up and Outcomes

	Closed (n = 28)	Open (n = 41)	P value
Follow-up Length (months)	9.5 (range, 1, 76)	4.0 (range, 1.38, 43)	0.02
Skin Infection	0%	10%	0.14
Septic Elbow	0%	5%	0.51
Nonunion	7.7%	0%	0.19
Avascular Necrosis	10.7%	12.2%	1.00
Premature Physeal Closure	7%	0%	0.164
Lateral Spurring	85.7%	78%	0.54
IEW:	n=23	n=32	0.42
Mild	21.74% (n=5)	34.38% (n=11)
Moderate	69.56% (n=16)	50% (n=16)
Severe	8.7% (n=2)	15.62% (n=5)
Abnormal Ulnohumeral Angle	32.1%	39.5%	0.6106

Discussion

Both the relatively high volume and morbidity of lateral condyle fractures make it an injury that should be examined for quality improvement. Currently, there are different methods to fix them, including through an open approach or through a closed approach.^7,8

Figure 5.

AP radiograph and last follow-up demonstrating a healed lateral condyle with retained implant. Apparently normal distal humerus growth is present despite persistent screw fixation.

When looking at the open and closed groups, their outcomes were quite similar, which has been described in other studies.^7,8 One difference encountered was the significant difference in the time to surgery with the closed group waiting 7 days longer than the open group (p<0.0001). In one systematic review, pediatric lateral condyle fractures displaced at a rate of 14%.¹⁴ The difference in time to surgery is therefore likely due to non-operative fractures displacing into operative fractures.

Despite the difference in displacement, the long-term outcomes assessed during this study were not significant. This includes septic arthritis, avascular necrosis, lateral spurring, and the percent with abnormal ulnohumeral angles.

Although both groups ultimately had similar outcomes, there were some significant differences at the time of surgery. The closed group was significantly more likely to have screw fixation with 78.6% of them being treated with screws while only 17.1% of the open group received screw fixation (p<0.0001). The type of fixation used may significantly affect the outcomes, but after rearranging our sub-group analysis, there were no significant differences in long-term outcomes when our patient population was divided into screws vs pins.

The most important differences at the time of surgery were OR time and the rate of post-operative hospital admission. The closed group was in the OR for an average of 44.7 minutes less than the open group (p<0.0001). This most likely represents the time necessary to perform the open approach and perform a layered closure. It may also represent a failed attempt at closed reduction. A shorter operative time decreases unnecessary exposure to anesthesia and dramatically decreases the cost of care as OR time and hospital charges are directly correlated.¹⁵ Aside from bringing down the cost of care for lateral condyle fractures, it would open up the OR, an important and limited resource, to be used for other patients. Additionally, patients who underwent closed reduction were significantly less likely to be admitted to the hospital after surgery. Only 7.1% of closed reduction patients were admitted after surgery while 47.5% of open reduction patients were admitted post-operatively (p=0.0004) for pain control. This difference likely represents a significant cost-savings from a healthcare perspective.

There were several limitations to this study. First, it was a retrospective chart review, so randomization, which would have increased the exchangeability of the patients in our groups of interest, was not conducted. Second, we used ICD-9 inpatient diagnostic codes to identify our study population. Errors with coding may have inadvertently limited the size of our study population. Lastly, the size of our study population was limited. While the significant values highlighted are likely true, there may have been true differences that were not significant due to lower power.

In conclusion, closed reduction in pediatric lateral condyle fractures has similar outcomes to open reduction. This, coupled with the significantly decreased OR times and decreased rate of post-operative hospital admission, make it the preferred approach from a hospital system quality improvement standpoint. Also, our paper highlights that the 4mm of maximum displacement is not a hard cut-off as approximately 1/3 of our patients treated with the closed approach had displacement greater than 4mm. With standard reduction techniques, their postoperative fracture gap was less than 2mm. Establishing a less stringent cutoff for attempting closed reduction maneuvers may be the subject of future research.