[Observer agreement in the measurement of leg length]

The lower limb length measurement is an important element for the diagnosis of mechanical or structural lumbar pain. Also it has been used for referral pain associated with hip or knee osteoarthritis or the groin and suprapubic areas. The aims of the present study were: 1) to measure the intra and inter observers variation; 2) to measure the intra-method variation using two different techniques for lower limb length measurement, one called the "apparent measure" (9) and comparing both with the radiological measurement technique. Two medical doctors, training on the techniques for lower limb measurement, performed the measurements. The exclusion criteria were flexion deformity of the hip or an overweight greater than 20% over the mean weight expected according to age and sex. A correlation coefficient and its 95% confidence interval (CI) were estimated, one tail test (Ho: r = 0.75). Seventeen patients fulfilled the inclusion criteria, 15 females and two males. The mean age was 35.8 years +/- 13.0 (SD). The correlation coefficient for the inter-observers variation using the "apparent measure" was 0.99 (CI = 0.985) and for the difference between legs it was 0.88 (CI = 0.10). The inter-observers variation for lower limb length measurement using the technique of "real measure" showed a correlation coefficient of 0.77 (CI = 0.95) and for the difference in length between legs it was 0.99 (CI = 0.85). The intra-observer correlation coefficient was 0.95 (CI = 0.85). The correlation coefficient for the inter-observer using the X-ray pictures was 0.98 (CI = 0.92).(ABSTRACT TRUNCATED AT 250 WORDS)