The following are software packages, apps and websites dedicated specifically to prognosis research.

Many generic (e.g. regression model or machine learning) packages can also be used to undertake prognosis research

​

The list is in development, and certainly not exhaustive.

Please get in touch to provide more recommendations to add. 

​​

• rms: suite for regression and prediction modelling produced for R by Prof Frank Harrell Jr.

• An excellent introduction to the rms and hsmisc packages is here

​

• PM suite of modules produced by Dr Joie Ensor:

  • pmsampsize: R and Stata packages for sample size calculations for prediction model development studies

  • pmcalplot: Stata package for producing calibration plots 

​​

• stcoxcal: a Stata package for examining the calibration performance of a survival model at particular time-point(s)

• metamisc:  an R package for conducting meta-analysis of prognosis studies by Dr Thomas Debray

• bsvalidation: a Stata package for bootstrap internal validation command for predictive logistic regression models

​

• A web app by Mike Kattan for calculating Riley et al.'s sample size required for developing clinical prediction models

​​

• Visit decisioncurveanalysis.org for software, data sets, tutorials etc. on net benefit & decision curves

​

• BMJ series (forthcoming) on validation by Riley, Collins et al .... software code as follows:

- Paper 1: R code for validation of the CRASH model is here

- Paper 2: Stata example code for the external validation of the binary and time-to-event models

- Paper 3: Stata code for the sample size for external validation of the continuous and binary outcomes is given

in the paper: more generally, see pmvalsampsize)

​

• Various researchers have GitHub pages offering code for prognosis and prediction research, including:

- Gary Collins: see here

- Glen Martin: see here

- Hisashi Noma: see here

- Valentijn de Jong: see here

- Karandeep Singh & others: see here

- Daniele Giardiello: see here