[Long-term results following fracture of the femoral neck in children].

PURPOSE OF THE STUDY: This retrospective study was designed to evaluate the severity and nature of long-term sequelae of femoral neck fractures in children in relation to the strategy and technique of therapy.

MATERIAL: The study included 15 patients with a fracture of the femoral neck who, at the time of injury, had an opened proximal physis. The average age at the time of injury was 11.5 years (range, 4 to 16.3 years). There were eight boys and seven girls. Twelve children suffered injury due to a fall varying in gravity, two were knocked down by a vehicle, and one was injured as a co-driver in a car accident. The group involved no type I fracture, six type II, seven type III and two type IV fractures, as classified by the Delbet and Colonna system.

METHODS: All children were operated on within an average of 4.1 days after injury. The delay was caused by a late referral from an outside hospital or was due to associated complications.However, the majority of fractures were treated within 24 hours of injury. Surgery was carried out from the anterolateral approach. Miniarthrotomy was performed in 11 patients to remove hematoma and decompress the intra-articular space. The methods of stabilization included Kirschner's wires in four children, compressive osteosynthesis using lag screws inserted extraphyseally in 10 children and the combination of both methods in one child. No plaster of Paris spica or traction was applied after surgery.

RESULTS: The long-term results were evaluated at a minimum of 5 years after injury. The average follow-up was 9 years and 11 months, with the range from 5 years and 1 month to 15 years and 5 months, and 12 patients were involved. Six had a type II and six had a type III fracture. Ten patients were treated by arthrotomy. Stabilization of the fracture was performed with lag screws in nine children and Kirschner's wires in three patients. Both subjective and objective findings were evaluated by the modified Rattlif criteria. Concerning pain, eight (67 %) children had excellent outcomes, three (25 %) reported good outcomes and only one (8 %) had a poor outcome. The activity following injury was subjectively evaluated as excellent by 11 (92 %) patients and as good by one (8 %) patient. None of the patients was noticeably limited in their activity, as compared with the pre-injury state. Objective findings were based on X-ray images and the range of hip motion. No or minimal radiographic changes were found in five (42 %) patients and were assessed as excellent outcomes. A good outcome, i. e., a spherical head with a moderate neck deformity, was achieved in five (42 %) patients. A poor outcome, i. e., avascular necrosis free of revitalization, but with collapse, on X-ray images was recorded in two (16 %) patients. The values for the range of motion and limb-length discrepancy were excellent in eight (67 %), good in two (16.5 %) and poor in two (16.5 %) patients. Poor objective and subjective findings were recorded in the patients who had not undergone miniarthrotomy.

DISCUSSION: A comparison with the literature data showed that the occurrence of each fracture type was in agreement with the reports of other authors. The finding of a higher frequency of type I fractures can be explained by a pathological slip due to hormonal changes. The opinions on treatment of these fractures have developed to the view that surgery is necessary in the early post-injury period, preferably with the use of compressive osteosynthesis or Kirschner's wires. However, subsequent immobilization in a plaster cast spica is not necessary. The subjective and objective findings were not exactly correlated due to differences in patients' age and different intervals between injury and treatment.

CONCLUSIONS: The authors recommend early surgery and stabilization by compressive osteosynthesis or Kirschner's wires, together with miniarthrotomy in order to decompress the articular space.