The apparent discrepancies noted between the figures are related to the method of data presentation. Each patient within a group had a series of sequential observations. To display all of the individual patients as lines on the graphs would be unwieldly, so the data were averaged at each time point, and the means were displayed. The problem noted comes from the handling of missing observations. In Fig 1, a few missing observations have a strong effect on the mean of some groups, eg, the 18-month point for group 4, where a number of the patients with the highest general values were not observed at that time. The analysis of variance method chosen requires a complete design, and missing values were supplied by interpolation, as mentioned in the "Statistical Methods" section. The means that resulted from this procedure were then displayed in Figs 2 through 4 and represent the
Elias AN, Gwinup G, Ascher M. Exophthalmos and Treatment of Graves' Disease-Reply. JAMA. 1982;248(18):2239–2240. doi:10.1001/jama.1982.03330180013015
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: