Primary aims: To determine if diagnostic ultrasound is an appropriate screening test (high sensitivity and at least moderate specificity) and non-inferior to x-ray for diagnosis of acute stress fracture of the lower extremity.
Secondary Aims: To determine what grade(s) of acute stress fracture that x-ray and ultrasound are capable of detecting reliably and accurately.
Population: Subjects will be recruited from the Sports Medicine clinic at the University of Virginia Health System in the Department of Physical Medicine and Rehabilitation. Subjects will have suspected acute stress fracture of the lower extremity and will be age 14 years and older. Anatomic locations included in this study are as follows: all bones of the foot and ankle, tibia, fibula, and femoral shaft. Patients with suspected stress fracture of the femoral neck will be excluded due to the difficulty of appropriate ultrasound evaluation secondary to depth as well as the potential for significant morbidity if not diagnosed properly. We estimate that the Sports Medicine clinic diagnoses approximately 12 acute stress fractures per month per physician amongst the three physician's clinics involved in this study. Therefore, our anticipated number of subjects over a 6 month period will be at least 216 patients.
Design: In this double-blind , prospective clinical study, subjects suspected to have an acute stress fracture of the lower extremity will be recruited from Sports Medicine Clinic (conducted by two physicians boarded in Physical Medicine and Rehabilitation and one physician board in Family Medicine and all boarded in Sports Medicine). When patients present to one of three physicians in Sports Medicine Clinic with a suspected acute stress fracture of the lower extremity, they will be offered the opportunity to enroll. Subjects will be required to have already undergone the initial step of the traditional diagnostic algorithm with x-ray of the suspicious region on the day of enrollment. All subjects will then be scheduled to undergo diagnostic ultrasound at a separate appointment with ultrasound procedure performed by a blinded clinical physician boarded in Physical Medicine and Rehabilitation, competent in performing and reading diagnostic ultrasound. All patients will then undergo a confirmatory MRI of the region of concern. Subject recruitment and data collection are expected to be completed over a six-month period. A statistician in the Department of Public Health at the University of Virginia will be performing statistical analysis. We plan to analyze findings using a McNemar chi-square test to evaluate for significant differences between the sensitivities of ultrasound and x-ray in diagnosing acute stress fracture of the lower extremity. All x-ray and MRI obtained for this study will be read by a Radiologist board certified in Musculoskeletal Radiology.
Diagnostic ultrasound performance: The physician performing the ultrasound will be blinded to the results of the x-ray and MRI. The physician performing the ultrasound will be informed of the bone in question (ex. Tibia) and will be able to perform a focused history and exam. The bone in which acute stress fracture is suspected will then be scanned throughout its length in two orthogonal views. The criteria for diagnosing acute stress fracture on ultrasound include displaying 2 out of 3 of the following: hypoechoic periosteal elevation of cortical bone, visible cortical disruption, hyperemia surrounding the periosteal lesion on power Doppler. The ultrasound physician will then be asked to make a determination whether the ultrasound study is positive for acute stress fracture or negative/indeterminate.
Statistical Analysis/Power Analysis: Sensitivity and specificity of diagnostic ultrasound and x-ray were calculated. MRI was used as the gold standard confirmatory test. Positive predictive value and Negative predictive value were also calculated for comparison. A priori power analysis determined a sample size of 186 was needed to obtain a power of 80%. The primary aim will be analyzed using a bootstrap test to compare the x-ray sensitivity to the ultrasound sensitivity. In order to account for the paired nature of the data, the bootstrap sampling will be at the subject level and will draw pairs with the x-ray and ultrasound outcomes associated with individual subjects. A statistician from the University of Virginia was hired for all statistical calculations.
Subject Recruitment: Subjects presenting to the Sports Medicine clinic with suspected acute stress fracture of the lower extremity on initial presentation will be offered participation. Participation includes an additional clinic visit with a separate physician to perform a diagnostic ultrasound of the area of concern in addition to the standard work up they would receive for their presenting complaint. Informed consent was obtained at the time of study enrollment. Subjective will be offered ultrasound evaluation as well as follow up confirmatory MRI. The cost of follow up MRI is included in the study to prevent selection bias based on insurance coverage for MRI. The costs associated with the additional clinic visit for diagnostic ultrasound is also included in the study budget.
