PRESENTING THE RESULTS OF DIAGNOSTIC TESTS ROUTINELY ORDERED TO DETERMINE GENERAL HEALTH STATUS

waterfall

Typically for clinical trials and non-interventional studies reams of output tables are generated presenting biochemistry and haemotology test results, perhaps summarised over time. How useful it is to present a multitude of output tables, of test result summary statistics, is open to debate. Firstly, the presentation is extremely dry. There is the issue too, that it is quite possible a different data set can give rise to similar summary statistics. Spotting errors in the data is also not easy and tracking individual patient profiles not possible. I would argue that these parameters should always be plotted. R Shiny is an ideal medium to communicate this information in a more useful and exciting way. We will focus on biochemistry tests that are routinely ordered to determine a patient’s general health status. I will make an attempt using simulated data.

Avatar
Uber stats nerd
Statistician

Interests include Bayesian statistics, graphics, modelling, prediction particulary in the clinical realm.

Related