Medical testing
Different tests on the same medical samples
You are diagnosed with some disease. However, you have some doubts and you ask for a second opinion. Chances are that a different testing method will be applied. Here we show results from real human samples.
Visualisation of test outcomes
The figure shows a method with only two outcomes (Whiff) against a more refined method (BV) with three outcomes. This type of figure is called a parallel coordinate plot.
Medical tests do not diagnose a disease unambiguously. The more common Whiff-test on 289 human samples shows a positive diagnosis in 102 cases. Most likely this resulted in a treatment for these 102 patients, and none for the remaining 187 patients.
The less common BV test separates the same samples into three categories, with three possible treatment scenarios.
The figure illustrates that diagnosing a disease sometimes is not trivial. Positive outcomes from one test can show up as a negative result in another test. And vice versa. In such cases your trust in the expertise of your medical practitioner becomes an important factor.
Data Source
LCPL, Leiden, The Netherlands.
Copyright
© 2019 Bontekoe Research, Amsterdam. Partner in NARMA42.