The client, Retina Consultants of Houston, asked us to analyze a dataset of 40 patients from a prospective clinical trial where data were collected weekly on three treatment groups over two years.
We cleaned the data by replacing the missing values with the last non-missing observations. We showed differences in various clinical measurements within and between the treatment groups over time using box-plots and line-plots.
We tested these differences between the treatment groups and over time (weeks 0, 24 and 48) using 1-way repeated measure ANOVA with post hoc pairwise comparison, linear mixed models with multiple comparison, pairwise t-test, and chi-squared test respectively. We interpreted the statistical results and wrote detailed explanation of the methods.
Our analysis resulted publication of the client’s study.