NormalGERValueDeterminationDiscussionMemo-07-30-02.doc

Hello, all Collaborators,The normal values for the 60 normal volunteers is below. This came from the statistics department at the Cleveland Clinic, which reproduced exactly the values Radu had extracted and provided for my presentations to the group at DDW 2002.Of the values that are presented below (I gave you parametric, and numerous non-parametric values to consider), I propose that these are the parameters we want to present in the tables which will be the most critical part of the normal volunteer publication:????? 1. MII parameters:???????? a) Four MII-GER types from 5 cm and 15 cm above LES: Total GER Acid GER Acid Rereflux Nonacid GER Includes both minor acid + nonacid values which would be presented in the results Note: only total is shown at 15 cm, but I would show the other values in the table)????????? b) % impedance bolus exposure 5 and 15 cm above the LES pH parameters:?????????? a) pH-only GER?????????? b) Percent acid exposure as traditionally defined The data is not normally distributed (i.e. skewed to the right). Therefore, non-parametric values are needed. Our statisticians at CCF suggest median and 25% / 75% quartiles, which would be displayed as a box plot. They also suggest abnormal be defined as the 97.5% value and since we have 60 normal volunteers this amounts to the 2nd highest value in the study. I think this should be displayed as a whisker or line extending to a bar at the 97.5% value (and 2.5% value when applicable) and individual dots representing the values above 97.5% to show the value of the outliers. A recent example of a similar display is in Gastroenterology 123:99. I disagree with our statisticians on one point and would like for this specifically addressed by the collaborators. I do agree with the box and whisker plot as described above, but feel we should use the 95% value. This is for two reasons: 1. If you look at he percent acid exposure values, the 97.5% value for total, upright, and supine are 7.6%, 12%, a