Interactive demo

This interactive demo runs BJM’s dynamic prediction for a single patient in the browser, with no R needed. Pick a patient from the pbc3 data, or upload one patient’s follow-up visits as a CSV file. Then choose a prediction time and a window and press Predict. The interface is available in English and Chinese.

If the frame does not load, open the demo in a new tab. On the free hosting plan the app goes to sleep when nobody is using it, so the first visit can take up to a minute to start.

What the demo shows

The app fits the two sub-models once, when it starts, on pbc3: a Cox model for death (status3) with age and sex, and linear mixed models for three biomarkers (log serum bilirubin, albumin and log prothrombin time). For the selected patient it then uses only the visits up to the prediction time $t$ and shows:

Panel	BJM function
Cumulative risk of the event in $(t, t + h]$ for $h$ up to the window	predictRisk()
Predicted mean and 95% interval of each biomarker at the end of the window	predictLongitudinal()
Observed history and predicted future trajectory of each biomarker	predictLongitudinal()

The biomarker predictions are conditional on the patient still being event-free at each future time. This is the quantity the backward joint model predicts jointly with the event risk.

Uploading a patient

The CSV needs one row per visit and the columns year, age, sex, serBilir, albumin and prothrombin, on the same scales as pbc3. The app has a template you can download. The event time is not needed: BJM integrates over it.

Running the app locally

The app’s source is in the shiny/bjm-demo folder of the repository. To run it on your own computer, clone the repository and run:

install.packages(c("shiny", "bslib"))
shiny::runApp("shiny/bjm-demo")

The demo is for illustrating the method only. Its models are fit on the PBC data set and must not be used for clinical decisions.