OBJECTIVE: To determine the medial metaphyseal beak (MMB) cut-off angle predicting Langenskiöld stage II of Blount's disease and to study the intra-observer and inter-observer reliabilities of angle measurements and the influence of the experience level of observers. METHODS: A retrospective study was conducted on children aged 2-4 years from January 2000 to December 2017. Children were identified through a computer-based search. Children with Langenskiöld stage II of Blount's disease who had been initially evaluated at our institution were categorized into Blount group and children who were diagnosed with physiologic bowing were categorized into control group. Data on the patients' ages, genders, and affected sides were collected. The MMB angles were measured on standing anteroposterior radiographs of the knees. The angle was formed between one line drawn parallel to the medial cortex of the proximal tibia, and a second line running from the intersection of the first line with the proximal tibial metaphysis through to the most distal point of the MMB. Measurements were independently performed by six observers. All observers repeated the measurements 2 weeks after they were first done. RESULTS: There were 148 legs from 79 children (48 males and 31 females) with an average age of 28.6 months. The average MMB angle of the Blount group was 128.52° ± 5.38° (P-value <0.001) and of the control group was 114.45° ± 4.89°. The average femorotibial angle of the Blount group was 15.48° ± 6.81° (P-value <0.001) and of the control group was 7.71° ± 7.94°. The receiver operating characteristic curve showed that an MMB angle >122° (sensitivity 92.7%; specificity 97.0%) was associated with Langenskiöld stage II. The intraclass correlation coefficient of the intra-observer reliability ranged from 0.93-0.97, and the inter-observer reliability was 0.93. CONCLUSIONS: By using anteroposterior (AP) radiographs of the knee, the MMB angle is a potential radiographic parameter to distinguish between Langenskiöld stage II of Blount's disease and physiologic bowed legs, with an MMB angle >122° predicting Langenskiöld stage II.
OBJECTIVE: To determine the medial metaphyseal beak (MMB) cut-off angle predicting Langenskiöld stage II of Blount's disease and to study the intra-observer and inter-observer reliabilities of angle measurements and the influence of the experience level of observers. METHODS: A retrospective study was conducted on children aged 2-4 years from January 2000 to December 2017. Children were identified through a computer-based search. Children with Langenskiöld stage II of Blount's disease who had been initially evaluated at our institution were categorized into Blount group and children who were diagnosed with physiologic bowing were categorized into control group. Data on the patients' ages, genders, and affected sides were collected. The MMB angles were measured on standing anteroposterior radiographs of the knees. The angle was formed between one line drawn parallel to the medial cortex of the proximal tibia, and a second line running from the intersection of the first line with the proximal tibial metaphysis through to the most distal point of the MMB. Measurements were independently performed by six observers. All observers repeated the measurements 2 weeks after they were first done. RESULTS: There were 148 legs from 79 children (48 males and 31 females) with an average age of 28.6 months. The average MMB angle of the Blount group was 128.52° ± 5.38° (P-value <0.001) and of the control group was 114.45° ± 4.89°. The average femorotibial angle of the Blount group was 15.48° ± 6.81° (P-value <0.001) and of the control group was 7.71° ± 7.94°. The receiver operating characteristic curve showed that an MMB angle >122° (sensitivity 92.7%; specificity 97.0%) was associated with Langenskiöld stage II. The intraclass correlation coefficient of the intra-observer reliability ranged from 0.93-0.97, and the inter-observer reliability was 0.93. CONCLUSIONS: By using anteroposterior (AP) radiographs of the knee, the MMB angle is a potential radiographic parameter to distinguish between Langenskiöld stage II of Blount's disease and physiologic bowed legs, with an MMB angle >122° predicting Langenskiöld stage II.