Talk:Rhenus:Validation 2008 (Rhenus-A-0.11-GER)

Sorry, but there must be a major mistake in the validation of Rhenus A versus the Rosso study regarding mortality. An absolute 15.5% difference of the same patient cohort in only 7 years is not plausible. I think this isn't a mistake in the model. Most likely there has been a mistake in performing the calculation. Please check! Wendelin Schramm 11:39, 23 May 2008 (CEST)

Plausibility of Rhenus mortality

The Rhenus model was checked again, but there was no mistake found in the model. The calculation was executed correctly. I added the baseline mortality (mortality table of "Statistisches Bundesamt") in the graph. The Rhenus model assumes a high excess mortality factor, which is based on clinical trials. --Twillert 23:40, 9 June 2008 (CEST)

Mortality still too high?

Thanks for this precious work. As a conclusion, I think we will have to look more closely on the excess mortality in the Rhenus model. Still I have the impression that it is rather high. I might be wrong as well...

On the other hand, we can see that the ROSSO data are almost equal with the standard mortality. Does the ROSSO study underestimate the diabetes related excess mortality? This is IMHO an indicator to look at least for another externals study for the external validation and it would be intersting to have data available form the coross-testing with other comparable models (similar model structure)! Wendelin Schramm 10:26, 11 June 2008 (CEST)