Closed Reduction and Percutaneous Pinning

Level III evidence has been instrumental in defining closed reduction and percutaneous pinning as the current treatment for displaced supracondylar fractures. The first such comparative study, in 1979 by Prietto,² retrospectively compared 27 patients treated with Dunlop’s traction (at one hospital) with 20 patients treated with closed reduction and percutaneous pin fixation. Clinical “gunstock” varus malunion occurred in 9 of 27 patients treated with traction (33%) and 1 of 20 patients with percutaneous pinning (5%). Mean hospital stay was 17 days for the traction group and 3 for the pinned group. Complications were low, and functional results were good in each group.

A much larger retrospective comparative study (Level III evidence) was published in 1988.¹ A total of 325 patients with displaced extension supracondy-lar fractures were treated by 1 of 4 methods, and 230 returned for late review. Treatments compared were closed reduction and casting (130), closed reduction and percutaneous K-wire fixation (78), traction (15), or open reduction and internal fixation (7). Among 130 patients initially treated with a flexion cast, 29 had the treatment discontinued because of problems with forearm circulation; 14 had varus malunion, of whom 8 had osteotomies, 6 had loss of motion, and 1 experienced development of a Volkmann’s ischemic contracture. Among 78 patients treated with K wires primarily plus 18 converted from casting to K-wire fixation, there were 3 varus malunions, of whom 1 had an osteotomy, 2 had loss of motion, and no patients had ischemia of muscle, although several had an initially absent radial pulse with an otherwise well-perfused hand. Pin placement was crossed medial and lateral pins in 47, and 2 lateral pins in 49 patients. No pin-related nerve complications with either placement were reported, and there were two superficial pin infections. About 78% of patients in the pinning group had both the carrying angle and range of motion within five degrees of the contralateral side at final follow-up; this was true of only 51% of the patients in the group treated with casts. The series was neither randomized nor prospective, and only 230 of 325 potential patients returned for a late review. Among those reviewed, the outcomes were much better and the complication rates lower in the closed reduction and percutaneous pinning group, and this remains the best available evidence supporting this practice. In summary, a large, well-done, retrospective cohort study (Level III evidence) supports closed reduction and percutaneous pinning of all displaced supracondylar fractures over closed reduction and plaster immobilization.

Crossed Pins or Lateral Entry Pins

A large retrospective series (Level III evidence) was published in 2001³ comparing the results of lateral pins versus crossed pins in 345 displaced extension-type supracondylar fractures. The configuration of the pins did not affect the maintenance of reduction. Fracture displacement was noted on 4% (4/103) of radiographs where lateral pins were used, 3% (4/120) where crossed pins were applied without hyperflexion, and 2% (1/58) where crossed pins were applied with hyperflexion. Ulnar nerve injury was not seen in the 125 patients in whom only lateral pins were used. The use of a medial pin was associated with ulnar nerve injury in 4% (6) of 149 patients in whom the pin was applied without hyperflexion of the elbow and in 15% (11) of 71 in whom the medial pin was applied with the elbow hyperflexed. Because the maintenance of reduction was equal with lateral or crossed pins, but the ulnar nerve injury rate was greater with crossed pins, the authors conclude that lateral pins alone provide adequate fixation. Two smaller Level III comparative studies had reached this conclusion previously.^4,5 A subsequent prospective Level IV study⁶ reports on 124 consecutive unselected displaced supracondylar fractures with no loss of reduction and no nerve injuries using lateral entry pins only, 2 pins for more stable, and 3 pins for less stable fracture patterns.

Two Level I studies (prospective, randomized trials) have since been performed addressing the question of lateral pins versus crossed pins.^7,8 Both trials were small. The first trial, with good methodologic quality, found no differences in minor loss of reduction (1/24 crossed, 6/28 lateral), ulnar nerve injury (none), or major loss of reduction (none).⁷ The second trial, with less complete description of the methodologic issues, had 5 ulnar nerve injuries among 34 patients with crossed pins, and 2 ulnar nerve and 1 radial nerve injuries among 32 patients with lateral entry pins.⁸ Although these are Level I studies, which remove selection bias, the small number of patients does not allow a precise estimation of the rate of rare events such as nerve injuries or loss of reduction. A systematic review pooling all data from previously published case and comparative series reports an iatrogenic nerve injury rate of 1.9% for lateral entry pins and 3.5% for crossed pins (among 1909 patients), and a loss of reduction rate of 0.7% for lateral entry pins and 0% for crossed pins (among 1455 patients).⁹ Evidence from the Level I studies alone is insufficient to decide about treatment because the series are small relative to the expected outcome rate, and because one of the studies had a high complication rate inconsistent with the rest of the literature.

One Level III retrospective study used a case–control design to analyze anatomic details of the medial pin placement comparing 18 patients with ulnar nerve injury with 72 patients without injury.¹⁰ The medial pin was inserted with much more anterior to posterior angulation (12.1 ± 11.7 degrees) in the group with ulnar nerve injuries compared with the group without (1.6 ± 11.0 degrees). Although this article gives some guidance on how to place a medial pin, the authors mention in the discussion that they have abandoned routine use of the medial pin in favor of lateral entry pinning, which was not discussed.

A final Level III retrospective study used a case–control design comparing 8 fractures that lost fixation with 271 that remained well reduced.¹¹ In each case with lost fixation, a technical error of initial pin placement was (retrospectively) identified: single pins in the distal fragment, lack of bicortical purchase, and lack of 2 mm separation of pins at the fracture site.

Although a larger Level I study could be suggested, a practical interpretation of the current literature is that lateral pinning is preferred (based on Level III and IV evidence and one good but small Level I study) because the loss of reduction is low in large, prospective, consecutive and retrospective, comparative series, and because we may value prevention of iatrogenic nerve injuries more highly than reduction of a low malunion rate.

Open Reduction versus Closed Reduction

Open reduction has been used selectively for closed supracondylar fractures that are “irreducible” and are mentioned in many series, but with no standard definition of “irreducible” or indeed of an acceptable reduction, it is difficult to judge what is being compared. One article¹² compares a policy of primary open reduction of 44 consecutive patients with a policy of closed reduction and pinning in 55 subsequent patients treated at the same institution (Level III evidence). Although the open reduction group was observed longer (35 vs. 21 months), more patients in the open reduction group were stiff at final follow-up examination (38%) than in the closed reduction group (20%), defining stiffness as 10 degrees or more loss of motion compared with the other side. No differences were reported in infections (three open, two closed) or ulnar nerve lesions (two open, two closed). The authors conclude that closed reduction with percutaneous pinning is preferable to a routine policy of open reduction.

Flexion Supracondylar Fractures

Flexion supracondylar fractures are much less common than extension-type injuries and have been thought to be less stable and more difficult to reduce because the posterior periosteum is torn at the fracture site with a flexion injury. A retrospective comparative cohort (Level III) from a single center found the rate of flexion injuries to be 3%, and reported 31% open reductions for flexion injuries compared with 10% for extension injuries and 19% preoperative ulnar nerve symptoms for flexion injuries compared with 3% for extension injuries.¹³

Pulseless Hand

Two case series from 1996 drew opposite conclusions from similar clinical material regarding the pulseless pink hand. The first was a case series of seven patients with a pulseless arm, all of whom had a “seemingly viable” hand after closed reduction and pinning, were treated by arterial exploration with vein grafting³ or microdissection vessel mobilization,⁴ and did well at follow-up.¹⁴ The authors of this first study recommend routine arterial exploration and repair for a pulseless but viable hand.

A larger case series of 22 children with a supracondylar fracture and an absent radial pulse on admission included 15 with a well-perfused hand after closed reduction and K-wire fixation, of whom 5 had no radial pulse. All patients with a well-perfused hand were treated with observation, including the five without a pulse, and none had any clinical circulatory problems at follow-up examination. Seven patients who had a cold white hand after closed reduction of the fracture underwent arterial exploration with or without repair. The authors conclude that routine exploration of the brachial artery is unnecessary in the circumstance of a pink, perfused hand with an absent radial pulse after closed treatment of a supracondylar fracture.¹⁵

One year later, a case series of 13 patients with pulseless, pink hands treated by closed with or without open¹⁰ reductions and various vascular treatments including vein patch angioplasty⁴ was published and included more information on the fate of the vessel using follow-up magnetic resonance angiography. Late magnetic resonance angiograms obtained on 10 patients showed occlusion or restenosis at the brachial artery in 5 patients, including 2 of 3 who had had a successful vein patch. Collateral flow was excellent in all arms with brachial artery compromise on magnetic resonance angiography, all patients had a palpable radial pulse on follow-up examination, and no patient had claudication or cold intolerance. Based on this experience, the authors recommend close observation instead of immediate invasive treatment for pink, viable hands without a radial pulse after reduction because regardless of repairing the vessel in the short term, arms were reliant on collateral circulation over the long term anyway, and none had poor clinically results.¹⁶