D J Fagan1, S Davies. 1. Department of Radiology, Royal Glamorgan Hospital, Ynys Maerdy, UK.
Abstract
OBJECTIVES: to assess the key symptoms and physical signs in the acutely injured knee that are significantly associated with fracture and utilise these to develop an algorithm for the use of radiography in acute knee trauma. DESIGN: retrospective review of all acute knee injuries presenting to the accident and emergency unit at a District General Hospital over a 6-month period in 1998 (over 500 cases). After exclusions, a study group of 370 eligible adults was obtained. These patients were reviewed for the key symptoms and physical signs, and radiological evidence of fracture. OUTCOME MEASURES: the X-ray films were interpreted by a consultant radiologist. Of those patients on whom radiographs were not performed, follow up was assessed by a review of any re-attendance to the A&E department. RESULTS: there was a statistically significant correlation, using chi-square (Mantel-Haentzel) and linear regression analysis, between the presence of a fracture and each of the following variables: age, non-weightbearing in the department (p<0.05), haemarthrosis, point bony tenderness, effusion (p<0.001). CONCLUSION: these key symptoms and physical signs may form the basis of an algorithm for examining knees radiologically, with acceptable positive and negative predictive values. (This is currently being tested prospectively).
OBJECTIVES: to assess the key symptoms and physical signs in the acutely injured knee that are significantly associated with fracture and utilise these to develop an algorithm for the use of radiography in acute knee trauma. DESIGN: retrospective review of all acute knee injuries presenting to the accident and emergency unit at a District General Hospital over a 6-month period in 1998 (over 500 cases). After exclusions, a study group of 370 eligible adults was obtained. These patients were reviewed for the key symptoms and physical signs, and radiological evidence of fracture. OUTCOME MEASURES: the X-ray films were interpreted by a consultant radiologist. Of those patients on whom radiographs were not performed, follow up was assessed by a review of any re-attendance to the A&E department. RESULTS: there was a statistically significant correlation, using chi-square (Mantel-Haentzel) and linear regression analysis, between the presence of a fracture and each of the following variables: age, non-weightbearing in the department (p<0.05), haemarthrosis, point bony tenderness, effusion (p<0.001). CONCLUSION: these key symptoms and physical signs may form the basis of an algorithm for examining knees radiologically, with acceptable positive and negative predictive values. (This is currently being tested prospectively).