OBJECTIVES: We investigated the agreement between the Blumensaat method and patellar height ratios and evaluated the reliability of the method in patellar height measurements. METHODS: We retrospectively evaluated 105 digital knee x-rays of 77 patients (23 males, 54 females; mean age 32 years; range 21 to 62 years), whose radiographs were obtained at 30 degrees of knee flexion. Patellar heights were measured with the use of the Blumensaat, Insall-Salvati, modified Insall-Salvati, and Blackburne-Peel methods. Patellar height was also evaluated after standardization of the angle between the femoral shaft and the Blumensaat's line (corrected Blumensaat method). Agreement and correlations were sought between the qualitative and quantitative results of the methods, respectively. RESULTS: When the qualitative results were considered, agreement of the Blumensaat method was very weak with the Insall-Salvati, modified Insall-Salvati, and Blackburne-Peel methods (kappa coefficients: 0.21, -0.14, and 0.12, respectively). The corrected Blumensaat method also had a very weak agreement with the other indices (kappa coefficients: 0.27, -0.11, and 0.13, respectively). Correlation analyses of the quantitative results showed very weak to moderate correlations between the methods used. The Blumensaat method was relatively well correlated with the modified Insall-Salvati method (r=0.514, p<0.05). Correlations were weak to moderate between the patellar height indices, the highest correlation being between the modified Insall-Salvati and Blackburne-Peel methods (r=0.557, p<0.05). CONCLUSION: The results of the patellar height measurements obtained by the Blumensaat method do not agree with widely used patellar height indices.
OBJECTIVES: We investigated the agreement between the Blumensaat method and patellar height ratios and evaluated the reliability of the method in patellar height measurements. METHODS: We retrospectively evaluated 105 digital knee x-rays of 77 patients (23 males, 54 females; mean age 32 years; range 21 to 62 years), whose radiographs were obtained at 30 degrees of knee flexion. Patellar heights were measured with the use of the Blumensaat, Insall-Salvati, modified Insall-Salvati, and Blackburne-Peel methods. Patellar height was also evaluated after standardization of the angle between the femoral shaft and the Blumensaat's line (corrected Blumensaat method). Agreement and correlations were sought between the qualitative and quantitative results of the methods, respectively. RESULTS: When the qualitative results were considered, agreement of the Blumensaat method was very weak with the Insall-Salvati, modified Insall-Salvati, and Blackburne-Peel methods (kappa coefficients: 0.21, -0.14, and 0.12, respectively). The corrected Blumensaat method also had a very weak agreement with the other indices (kappa coefficients: 0.27, -0.11, and 0.13, respectively). Correlation analyses of the quantitative results showed very weak to moderate correlations between the methods used. The Blumensaat method was relatively well correlated with the modified Insall-Salvati method (r=0.514, p<0.05). Correlations were weak to moderate between the patellar height indices, the highest correlation being between the modified Insall-Salvati and Blackburne-Peel methods (r=0.557, p<0.05). CONCLUSION: The results of the patellar height measurements obtained by the Blumensaat method do not agree with widely used patellar height indices.