The data used for Capire360 models have been both internally and externally validated.
Internal validation involves splitting the data up into a training set and a separate test set. Once the model has been trained, it is validated against the test set.
Statistical accuracy has been estimated using the model’s concordance and external validation of the methods used.
The MyCancerJourney analytic platform was developed and back-tested over the past decade. Experts have validated the models in healthcare analytics at the University of Rotterdam, Erasmus. Ewout Steyerberg is a Professor of Medical Decision Making, particularly prognostic modeling, at Erasmus MC–University Medical Center Rotterdam, the Netherlands. Professor Steyerberg is one of the world’s leading experts on validating prognostic modes in medicine. Various research grants stimulated his work on prediction models, including a fellowship from the Royal Netherlands Academy of Arts and Sciences. He has published over 500 peer-reviewed articles collaborating with many clinical researchers in methodological and medical journals.
The Erasmus team performed internal and external validation of the models for the four most prevalent cancer sites (breast, prostate, colorectal, and lung). Internal validation, based on 100 bootstrap samples, demonstrated that the models provided an accurate survival prediction with an average optimism of 0.01 around the c-indices with values ranging from 0.005 to 0.01.
The model’s clinical significance was validated through reviews by oncologists, and it has been compared to the NCI SEER and American Cancer Society outcomes data.
Statistical accuracy has been estimated using the model’s concordance. These models have been validated by experts at Erasmus Medical Center, Rotterdam, Netherlands. The model’s clinical significance was validated through reviews by oncologists.