Clinical use of the knee height measuring device to detect growth deficiency.

Abstract

AbstractAssessment of growth deficiency cannot usually be carried out until after 6 months of observation because errors associated with height measurement preclude use of shorter periods. In an effort to shorten this time period, the Knee Height Measuring Device (KHMD), a highly precise and reliable instrument and method for measuring child growth, was compared with height in detecting and clasifying growth in a group of healthy school children, and in a small group of children with Crohn disease known to be experiencing growth failure. Children were measured at 28 day intervals. Measurement error (me) (the standard deviation of the replicates for knee height, and technical error of measurement \documentclass{article}\pagestyle{empty}\begin{document}$$ \sqrt {\sum {{\rm d}^{\rm 2} {\rm /2n,}}} $$\end{document} for height) was used as the basis for detection and classification of growth. To estimate an error that might be typical in a clinical setting, the me from this study was doubled. Then to account for errors associated with both the first and second measurements used to compute growth rates, this value was doubled again. The me for knee height was 0.33 mm (SD = 0.22), yielding a minimum detectable growth of 1.32 mm, and the me for height was 2.7 mm, yielding a minimum detectable growth of 10.8 mm. Using 56 day knee height growth (mean 3.77 mm SD 1.43) more than 95% of growth rates for healthy children exceeded the minimum criterion, whereas for 56 day height growth (mean 8.7 mm, SD = 5.3), only about 28% of the growth rates for healthy children were greater than the criterion. Sensitivity and specificity for 56 day growth knee height and height growth were compared for classifying growth rates of the healthy children and children with Crohn disease using the me criteria. For knee height, a sensitivity of 100% and specificity of 96% were achieved, whereas the sensitivity and specificity for 56 day height growth were 75% and 30% respectively. Parallel analyses comparing 56 day knee height with 6 month height velocity could not be carried including the children with Crohn disease. However, using the 5th centile from age and sex specific height velocity from Baumgartner et al. [Am. J. Clin. Nutr. 43:711–722, 1986], the expected number of healthy children in this sample (5.7%) were less than 5th centile. This would yield a specificity for 6 month height growth (94.2%) which is similar to that for 56 day knee height. It is concluded that measurement of knee height over 56 days may potentially be useful in shortening the time for clinical detection of growth deficiency. © 1993 Wiley‐Liss, Inc.