Bioequivalence analysis - two-period, two-treatment crossover trial

Should data be log-transformed prior to analysis?

Analysis on untransformed data

Log-transform for analysis

Paste data in the space below. When copying and pasting data from Excel, only the number of decimal places displayed in Excel will be copied. This may lead to small rounding errors if more decimal places are present in the original data.

Download example data
(Use the CVM35 or Theophylline sheets for log-transformed data and Chow_71 for untransformed data)

Submit

Inputs

At least five columns with labels as listed below:

• id: a number or text label to identify each subject;
• treat: treatment identifier. (Note: Coding of treatment and reference groups should be such that the Reference group code precedes the treatment group code alphabetically or numerically. For example "R" for reference and "T" for treatment (or 1 and 2);
• period: identifier for period (usually 1 or 2);
• sequence: identifier for treatment sequence for each subject;
• value: the raw (untransformed) data value to be used in the analysis;
• include a header of column names;
• column order is unimportant; and
• additional columns can be included and will be ignored.

Outputs

Outputs are a table and plot of the following values:

• analysis of variance table;
• summary of period and sequence data;
• summary table of mean differences;
• bioequivalence comparisons; and
• plot of IVP/RVP geometric means ratio.