Hello. This Shiny App uses the Continual Reassessment Method (CRM) to assign the next dose or determine the maximum tolerated dose (MTD) in a phase I trial.

The app either uses Maximum likelihood estimation or Bayesian inference to estimate the probability of toxicity at each dose. When the maximum sample size is not yet reached, the app gives the dose assignment for the next cohort. When the maximum sample size is reached, the app gives the MTD. The MTD is the potentially effective and safe enough dose to be used in the next trial (phase II).

Before conducting the CRM, many design parameters must be specified such as the target toxicity level, the dose-toxicity function, and the total sample size. There is a separate app called ‘CRM Design’ that one can use to determine the necessary parameters. Alternatively, one can download the parameter template in the ‘Data Description’ tab and fill in the necessary parameters.

The app is organized through the following tabs:

  • In the Data Description tab, one will find the required format for the data and statistician file
  • In the Data Upload tab, one will upload trial data similar to the format from the sample data set, shown in the `Data Description` tab, and upload the parameters for the CRM, referenced as the ‘Statistician file’
  • In the Next Dose tab, the app gives the recommended next dose or MTD using the CRM based on the data and statistical parameters provided. The predicted toxicity at each dose is also given in a figure
  • In the Visualization the app offers visualization of the number of DLTs at each dose by patient ID and date. See the Phase I Visualization App for more

References:

Ken Cheung (2019). dfcrm: Dose-Finding by the Continual Reassessment Method. R package version 0.2-2.1. https://CRAN.R-project.org/package=dfcrm

DISCLAIMER: Users are solely responsible and liable for use of content throughout the app.

Developed by
Rebecca B. Silva

In the Data Upload tab, the user will upload trial data similar to the format from the sample data set on the right. This data can be downloaded and used as a sample for CRM trial conduct in the 'Dose Assignment' tab. The data should be a CSV file and have the following format:

  • Patient ID variable: distinct number or character for each patient.
  • Dose level variable: numeric or character variable indicating the dose level each patient has been assigned.
  • Toxicity Variable: binary variable indicating whether or not a patient experienced a dose-limiting toxicity (DLT). 0 or 1 (1 = DLT) for each entry.
  • Evaluable Variable: optional variable indicating whether or not the patient is evaluable in the trial. 0 or 1 (1 = Evaluable) for each entry.
  • Start and End Date Variables: optional variables for visualization in the 'Visualization' tab. They can be date or character variables in any of the following formats: MM/DD/YYYY, MM/DD/YY, YYYY/MM/DD, YY/MM/DD, MM-DD-YYYY, MM-DD-YY, YYYY-MM-DD, YY-MM-DD.

Sample Trial Data Set

Download sample data


If using a statistician parameter file not obtained through the CRM Design App, download the template file on the right.

Note: the options for the 'Dose-Toxicity Model' are 'empiric' or 'logistic' and the options for the 'Method' are 'bayes' and 'mle'.

Sample Statistician Parameter File

Download parameter file

Next dose or MTD:

Maximum sample size of trial:

Number of patients in trial:

Number of evaluable patients: