Intramedullary Nailing of the Femur:Trochanteric vs Piriformis Starting Portals

This project is designed as a prospective, randomized, comparative study between the Trigen Trochanteric Femoral Nail (Smith \& Nephew, Memphis) and a standard femoral intramedullary nail that utilizes a piriformis fossa portal in the treatment of fractures of the subtrochanteric and diaphyseal shaft regions of the femur. The hypothesis is that the trochanteric insertion portal will have no difference when compared with the piriformis fossa portal in terms of pain and strength of the hip abductor muscles, while allowing faster surgical fixation of the femur fracture.

Any patient who has sustained a fracture of the femur, is at least 19 years of age, and meets the study's inclusion and exclusion criteria will be asked to participate in the study. After informed consent has been obtained, the patient will be taken to the Operating Theater for surgical fixation of the femur fracture. No fracture will be entered into the study unless antegrade femoral nailing is the treatment of choice of the attending surgeon.

The outcome variables consist of the intraoperative and postoperative adverse device effects recorded in the medical record and on the study data collection form, intraoperative and postoperative complications recorded in medical record and on the study data collection form, hip functional outcome using the WOMAC score recorded for each patient, SF-36 score recorded for each patient, results of clinical examination at 6 weeks, 3, 6, and 12 months recorded in medical record and on the study data collection form, radiographic examination evaluated by two surgeons for the presence of callus and radiographic healing recorded in the medical record and on the study data collection form. Radiographic healing at 6 weeks, 3, 6, and 12 months recorded in the medical record and on the study data collection form. (The injury severity score for all multiple trauma patients recorded on the study data collection form, and soft tissue score according to Gustillo and Anderson for open injuries and according to Tscherne for all closed injuries recorded on the study data collection form. This data will be taken into account relative to the outcome variables). The effects will be measured by clinical examination, radiographic examination, and subjective patient's outcome survey. Additionally, hip abductor muscle strength will be measured and ability to perform one leg stance and hop will be documented at 6 and 12 months.

Data will be collected and entered into EXCEL spreadsheets with double keying for quality control purposes and managed by the Department of Orthopaedic Surgery at the University of Alabama at Birmingham (UAB). After data entry is complete, the file(s) will be copied to diskette and delivered to the UAB Biostatistics statistician for analysis and report generation. The analysis will involve simple descriptive statistics (means, standard deviations, proportions) to assure balance of the treatment groups with respect to relevant variables. T-test will be used to compare the mean times to callus formation and radiographic healing between groups. Analysis of covariance will be performed if there are variables that require adjustment. Growth curve analysis (survival models) and Cox regression procedures will be used to compare the rates of callus formation and/or healing between the two groups. Assumption of proportional hazards in the model will be verified before attempting Cox regression procedures. P-values will be computed using a Mann-Whitney U test for non-parametric data.