Comparison of complications and results of early versus delayed surgery for Gartland type III supracondylar humeral fractures in pediatric patients

Abstract

Objective:  To investigate the effect of timing of surgery on clinical results and perioperative complications in pediatric patients with Gartland III type supracondylar humeral fractures without neurovascular compromise.

Methods:  Eighty‐six consecutive children treated surgically at our hospital from April 2005 to June 2007 for displaced supracondylar humeral fractures were reviewed. All these patients were treated by the same group of doctors. The children were divided into two groups: early if treated within 12 hours after injury and delayed if treated later than that. Perioperative complications and clinical results, especially for open surgery, were compared between the two groups.

Results:  Forty pediatric patients underwent surgery in the early group and 46 in the delayed group. There were no significant differences between the two groups in perioperative complications such as pin tract infection, iatrogenic nerve injury, compartment syndrome and conversion to open surgery. For open surgery, both the clinical results and perioperative complications were not affected by delaying for more than 12 hours after injury. However, blood loss and operation time were greater in the early than in the delayed group, possibly due to relatively more edema.

Conclusion:  Delay in surgery, regardless of whether it is closed or open, for more than 12 hours after injury does not influence the perioperative complications and clinical results for displaced supracondylar humeral fractures in children. However early open reduction and pinning may increase intra‐operative blood loss and take longer.

Introduction

Supracondylar humeral fracture is the commonest fracture around the elbow joint in pediatric patients under the age of seven ¹ . Because complications such as neurovascular damage, compartment syndrome, elbow deformity and loss of joint motion are not rare in supracondylar humeral fractures, especially in Gartland type III fractures, their treatment is challenging. It has been accepted that surgery is the best treatment for these displaced fractures, however the timing of that surgery is still a controversial topic. Traditionally, emergency treatment has been thought to be helpful in reducing perioperative complications, although there has been no convincing proof of this. Recently some authors have suggested that there are no significant differences between early and delayed surgery in regard to complications and functional outcomes ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ . However, it has not been clear whether surgical timing affects the results of open surgery, such surgery not being routinely undertaken. To explore this issue we retrospectively studied clinical results and complications in pediatric patients who had sustained displaced supracondylar humeral fractures and undergone surgery with different timing.

Materials and methods

All records of children with surgically treated supracondylar humeral fractures between April 2005 and June 2007 were reviewed. Inclusion criteria were age range from 3 to 15 years and Gartland type III supracondylar humeral fractures. Those who sustained open fractures, associated neurovascular compromise and ipsilateral forearm factures were excluded. Eighty‐six consecutive cases were included in this study series. The patients' data, including age, sex, side of fracture, time and date of injury, operation details, perioperative complications and functional outcomes at final follow‐up, were documented. The operation details included operation time and date, blood loss, operation length and surgical approach. The patient data are presented in Table 1. Perioperative complications such as pin tract infection, iatrogenic nerve damage, compartment syndrome and conversion to open surgery were also documented. Furthermore, the details of open surgeries were studied.

Table 1.

Patients' details in early group and delayed group

Items	Early group	Delayed group	P value
Number	40	46	—
Mean age () (years)	7.2 ± 2.6	7.6 ± 3.5	P > 0.05
Male/female	24/16	30/16	P > 0.05
Fracture side (L/R)	22/18	29/17	P > 0.05

All patients presented to the Emergency Department with elbow swelling, with or without varus or valgus deformity, movement limitation and pain. The diagnoses were made on the basis of the clinical and radiograph findings. Ultrasonic examination was performed if considered necessary for exclusion of arterial damage. All fractures were immobilized with a cast and the arm elevated while awaiting further treatment.

All patients underwent surgical intervention. The same group of doctors performed all the operations. The timing of the surgical interventions was determined by availability of operating theater resources. The early group was defined as surgery undertaken within 12 hours of trauma and the delayed group as surgery more than 12 hours after trauma. No patient underwent surgery more than 72 hours after injury. All patients initially underwent manipulative reduction under brachial plexus block combined with basal anesthesia, after which the reduction results were checked by fluoroscopy television. If the reduction results were satisfactory, internal fixation was performed with two or three crossing Kirschner wires inserted percutaneously under fluoroscopy. The pins were inserted such that they penetrated the contralateral cortex and crossed at an angle of 30°–40°. If the closed attempt at reduction had failed, open reduction and internal fixation were performed. A posterior triceps sparing approach was used to expose the fracture, and transposition of the ulnar nerve was performed in all cases. Once internal fixation had been achieved, a cast was used for additional stabilization with the elbow flexed at 60°–90° and the forearm in a pronated position.

Sixteen cases in the early group and 24 in the delayed group underwent open reduction and internal fixation (ORIF) treatment, respectively. X‐ray films were taken 3 weeks after surgery and the cast removed if there was callus bridging the fracture gap. The casts were removed briefly twice a day for passive motion exercises during the first 3 weeks after surgery. Active motion exercises were recommended after removal of external fixation. Once radiographs had confirmed bony healing of the fractures, the crossing pins were removed.

All children were followed up for at least one year. Functional results were evaluated by Flynn's criteria and radiographs were taken at each follow‐up visit. The final clinical results were evaluated according to Flynn's criteria: the result was classified as excellent if the loss of carry angle and extension‐flexion motion range loss was less than 5°, as good if between 5° and 10°, as fine if between 10° and 15°, and as bad if greater than 15°.

Statistical analysis

SPSS 13.0 software was used for statistical analysis. Statistical methods used in this study included Student's t‐test and the χ² test. A P‐value of 0.05 or less was considered to be statistically significant.

Results

Perioperative complications

In the early group closed reduction results were unsatisfactory in 16 fractures; immediate open reductions were performed. In the delayed group, closed reduction was unsuccessful in 24 fractures; these were treated by immediate open reduction and pinning. The incidence of open reduction was 40% in the early group and 52% in the delayed group, this difference not being statistically significant (Table 2). These results indicate that delays of more than 12 hours after injury do not increase the open surgery rate.

Table 2.

Perioperative complications in early group and delayed group

Items	Early group	Delayed group	P value
OS rate (case (%))	16 (40)	24 (52)	P > 0.05
IND (case (%))	3 (7.5)	2 (4.3)	P > 0.05
Infection (case (%))	1 (2.5)	3 (6.5)	P > 0.05
ORT of CS (mins, )	53 ± 17	48 ± 13	P > 0.05
ORT of OS (mins, )	86 ± 23	71 ± 19	P < 0.05
BL in OS (ml,)	280 ± 20	185 ± 35	P < 0.05

BL, blood loss; CS, closed surgery; IND, iatrogenic nerve damage; ORT, operation time; OS, open surgery.

The mean duration of closed surgery was 53 minutes (53 ± 17 minutes) in the 24 fractures in the early group and 48 minutes (48 ± 13 minutes) in the 22 fractures in the delayed group, this difference was not significantly different. However, the differences in duration and blood loss of open surgery were statistically significant. The mean duration of open surgery was 86 minutes (86 ± 23 minutes) in the early group and 71 minutes (71 ± 19 minutes) in the delayed group, and blood loss was greater in the early group (185 ± 35 ml) than in the delayed group (280 ± 20 ml). Thus open surgery within 12 hours of injury took longer and resulted in greater blood loss. The relatively more severe swelling and hyperemia around the fracture area may have contributed to that difference by making the surgery more difficult.

Three patients in the early group and two in the delayed group had evidence of ulnar nerve palsy after closed reduction and pinning. All five recovered completely without special treatment. No iatrogenic nerve injury occurred with open reduction and pinning in either the early or the delayed group. One child in the early group and three in the delayed group presented redness and mild swelling around their pin tracts without skin necrosis or purulence. These minor infections were cured by changing of dressings and oral antibiotics before the patients were discharged. No case required pin removal for treatment of infection. There were no significant differences between the two groups in iatrogenic nerve damage and pin tract infection. It is notable that no nerve damage occurred in any open surgery; this may have been because the ulnar nerve was routinely isolated before reduction and pinning was performed.

Functional outcomes

All cases achieved bony healing. There was no significant difference between the two groups in healing time (data not shown). For open surgery, the mean time for bony healing was 4.8 months (2–8 months) in the early group and 5.1 months (2.5–7.5 months) in the delayed group, this difference not being statistically significant. The range of motion of the elbow at 12 months follow‐up was not significantly different between the two groups, neither was the general function of the elbow according to Flynn's clinical evaluation criteria (Table 3).

Table 3.

Clinical results of open reduction and pinning in the two groups

Items	Early group	Delayed group	P value
Union time (months),	4.8 ± 1.8	5.1 ± 2.3	P > 0.05
Range of motion (°)	124.5 ± 3.9	126.1 ± 4.8	P > 0.05
Flynn scores
Excellent (case)	12	16	—
Good (case)	2	6	—
Fine (case)	1	1	—
Poor (case)	1	1	—
Rate of E & G (%)	87.5	91.7	P > 0.05

Notation: E & G, excellent and good.

Discussion

Supracondylar humeral fractures are the commonest fractures around the elbow in pediatric patients, the extension type accounting for two thirds of hospitalized children with elbow trauma ⁹ . Although the results of treatment of most children with supracondylar humeral fractures are satisfactory, perioperative complications are not rare, especially in Gartland III type fractures. These fractures always involve serious swelling and complete displacement and have a high rate of associated neurovascular injury, which make their treatment challenging. The usual treatment is closed reduction and percutaneous pinning, or open reduction and pinning if the closed attempt fails. Turhan et al. found no significant difference in radiographically assessed reduction between patients treated by closed versus open reduction ⁸ . However, some authors have expressed concern about the cosmetic and functional outcome of open surgery ¹⁰ . Some state that closed reduction is associated with fewer complications, such as infection and loss of motion. ¹¹ , ¹² However, in those reports open reduction followed failed attempts at closed reduction, indicating that these were more severe fractures that were more difficult to manipulate successfully. Kazimoglu et al. compared primary open surgery with primary closed surgical treatment of Gartland III fractures ¹³ . They found no significant difference between the two groups in outcomes according to Flynn's criteria, but the hospital stays were shorter in the closed reduction group. In our study all children underwent initial closed reduction, which was successful in 53.5% of patients. We do not recommend repeated manipulation because this increases swelling and exposes the patients to more X‐ray irradiation. If the reduction results are not satisfactory after two or three closed attempts, our policy is to switch to open reduction. We prefer a posterior triceps sparing approach because this adequately exposes the fracture and makes it easy for us to judge the reduction results. Furthermore, this approach decreases the possible damage caused by exposure to X‐rays in children.

Traditionally, pediatric patients with supracondylar humeral fractures have been treated as emergencies because of the belief that delay results in increased risk of perioperative complications, such as compartment syndrome, nerve injuries and infection, and increases the likelihood of conversion to open reduction ¹³ , ¹⁴ , ¹⁵ . Recently that concept has been challenged by some authors. They found no significant difference in regard to perioperative complications or conversion to open reduction between patients treated early and after a delay ² , ⁴ , ⁵ , ¹⁶ . Furthermore, some authors have proposed that delayed surgery allows time for better preoperative preparation ⁷ .

In previous reports the time criteria for early versus delayed surgery were 8 or 12 hours after injury ⁴ , ¹⁶ . In our study the children were treated from 3 to 72 hours after injury. We set a time criterion of 12 hours after injury. The results of our study confirm that perioperative complications, risk of conversion to open reduction, and functional outcomes are not significantly different when surgery is delayed more than 12 hours after injury. These findings are consistent with previous reports ² , ³ , ⁴ , ¹⁶ .

In particular, we studied the effect of surgical timing on the incidence of iatrogenic nerve damage, operation length, blood loss during open surgery and Flynn's outcomes at the last follow‐up. We found no statistically significant difference between patients operated on within 12 hours and more than 12 hours after injury in regard to iatrogenic nerve damage and Flynn's outcomes, while blood loss was greater and operation time longer in the early treatment than in the delayed group. This may be explained by the more severe swelling within 12 hours of injury. However, our samples were relatively small, which may have lead to bias in statistics. Larger samples should be studied to strengthen these conclusions.

Although arrested epiphyseal growth was not found at last follow up, because of the relatively short follow‐up time it cannot be concluded that trauma or surgery did not damage the development of epiphyses in our study patients.

Postoperative management is not extensively discussed in previous published reports. In our study of pediatric patients, all injured elbows were immobilized in flexion of 60–90° with a plaster cast for 3 weeks. From the second day after surgery, the casts were temporally removed twice a day for passive motion exercises. These functional treatments require close cooperation from the parents and are of key importance in decreasing motion restriction.

Conclusion

We found no significant differences between early and delayed surgery groups in pediatric patients with supracondylar humeral fractures. Notably, with open reduction and pinning the incidence of perioperative complications and functional results according to Flynn's criteria were not affected by the surgical timing, but blood loss and operation time were greater for open surgery within 12 hours than for that beyond 12 hours of injury. Our results suggest that it is better to perform surgery more than 12 hours after injury than within 12 hours of injury in pediatric patients with Gartland III type supracondylar fractures of the humerus.

Disclosure

The authors did not receive any outside funding or grants in support of this research for, or preparation of, this work. Neither they nor a member of their immediate families received payments or other benefits, or a commitment or agreement to provide such benefits, from a commercial entity.