I would like to analyse the additive predictive value of a biomarker over an old model in development of stroke using NRI.
I was wondering:
1. Is it possible to calculate NRI in SPSS?
2. I know that NRI tries to identify how a new model can reclassify subjects (in my example as having stroke or not). Is it possible to calculate NRI for a model with no binary outcomes (e.g. low, medium and high risks)?
Thanks in advance.
Hi Simin
I employ SPSS but it has some limits...as in your instance.
IBM suggest: "IDI and NRI are not directly available in an SPSS Statistics procedure. An enhancement request has been filed with SPSS Development. An R package named survIDINRI ( see link) exists that provides these measures, so it may be possible to produce such measures using the R plugin (se 2nd link).
Good Luck and.....have a naice job
https://cran.r-project.org/web/packages/survIDINRI/index.html
https://www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/We70df3195ec8_4f95_9773_42e448fa9029/page/Downloads%20for%20IBM%C2%AE%20SPSS%C2%AE%20Statistics
Hi Davide,
Thank you for your reply. I was trying to avoid using R since I have no experience with it. But i think for the NRI definitely R would be the best choice.
Thanks again,
This is the script for R to calculate the NRI and ROC curve,
first make the linear regression model with R rms package, you also require the Hmisc and PROC packages:
Best wishes
###################################################################
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS