17,000+ users
Click for interactive map
 File NameSizeNotes
 BOIN_V1.0.4.zip  55673 KB Zip file containing the installer for Windows version
Email Address: Organization: Occupation:
Your email will be used to send you occasional information about software updates. It will NOT be shared with third parties.

Bayesian Optimal Interval (BOIN) Design Desktop Program

The optimal interval design is a novel Bayesian phase I clinical trial design for finding the maximum tolerated dose (MTD) for single-agent and drug-combination trials. It strictly adheres to ethical considerations and optimizes the dose assignment for each patient enrolled into the trial.


The software (i.e., a Windows desktop program called “BOIN”) implements the Bayesian optimal interval (BOIN) designs of Liu and Yuan [1] for single-agent trials, of Lin and Yin [3] for drug-combination trials seeking a single MTD, and of Zhang and Yuan [4] for drug-combination trials seeking an MTD contour. The prominent advantage of the BOIN design is that it can be implemented in a simple way similar to the traditional 3+3 design, but yields excellent performance comparable to the more complicated model-based designs, such as the continual reassessment method (CRM). The BOIN design is motivated by the top priority and concern of clinicians, which is to effectively treat patients and minimize the chance of exposing them to subtherapeutic or overly toxic doses.

screen shot showing the Model Parameters tab page

System Requirements

  • Windows 7 SP1 (but may work on later versions of Windows as well, although this has not been tested)
  • Microsoft .NET Framework version 4.6.1 (x86 and x64)
  • Windows Installer 4.5
  • Minimum screen resolution 1310x724

If any required software component is absent from your system, the installation process will install it.


  • Self-contained Windows program
  • Very easy to use
  • Handles single-agent and drug-combination phase I trials
  • Enables the user to generate operating characteristics for protocol preparation
  • Automatically generates a protocol document template in either Word or HTML format if desired
  • An Estimate MTD calculator to estimate the MTD from trial data for single-agent trials
  • A Trial Conduct tab for conducting drug-combination trials and estimating the single MTD or MTD contour from trial data

Why use the BOIN Design?

  • Very simple to implement  Similar to the 3+3 design, the dose escalation/de-escalation rule is prespecified and can be tabulated, thus conducting a single-agent trial design by this method does not require additional software. During the trial conduct, clinicians can simply count the number of patients who experience toxicity and compare the observed toxicity rate with the prespecified dose escalation/de-escalation boundaries to determine dose assignment until the trial is completed. For drug-combination trials, the dose escalation/de-escalation rule is the same but the overall conduct of the trial is more complicated which is why we provide a trial conduct tab for those trials. Nevertheless each decision to escalate the dose, de-escalate the dose, or remain at the same dose is easy to understand, explain, and predict.
  • Superior performance  Both theoretically and numerically, it has been shown that the optimal interval design has desirable statistical properties and superior operating characteristics --- better than commonly used phase I designs.
  • Highly Ethical  The design optimizes the dose assignment for each patient in the sense that it minimizes the chance of assigning a patient to either a subtherapeutic dose or an overly toxic dose. This is consistent with the clinician’s point of view for conducting clinical trials, i.e., maximizing the patient’s treatment benefit.
  • Accommodates both single-agent and drug-combination trials  The BOIN design can be used to design both single-agent and drug-combination phase I trials. The resulting designs are easy to implement.
  • Solid statistical justifications  The design optimizes a sensible statistical criterion. It has a desirable finite-sample property (i.e., is coherent) and a desirable large-sample property (i.e., converges to the target dose). See the references for more details.


Extensive context-specific help is provided in the program itself for all aspects of the trial design and how to use the program. The program produces simulation output which can be incorporated in other documents and also can automatically generate a protocol document template in either Word or HTML format. An extensive statistical tutorial, written for the R package BOIN, is available separately. This tutorial covers the statistical basis of the method, and has guidelines for how to use the method.


Ying Yuan and Suyu Liu originally implemented the numerical algorithms using R. Liangcai Zhang contributed additional R code.

Richard C. Herrick and John Venier wrote this program.


Single Drug Study:
  [1] Suyu Liu and Ying Yuan (2015) Bayesian Optimal Interval Designs for Phase I Clinical Trials, Journal of the Royal Statistical Society: Series C, 64, 507-523.
  [2] Yuan Y., Hess K.R., Hilsenbecck S.G. and Gilbert M.R. (2016) Bayesian Optimal Interval Design: A Simple and Well-performing Design for Phase I Oncology Trials. , Clinical Cancer Research, 22, 4291-4301.

Combination Drug Study Seeking a Single MTD:
  [3] Lin R. and Yin, G. (2015) Bayesian Optimal Interval Design for Dose Finding in Drug-combination Trials, Statistical Methods in Medical Research Series, DOI: 10.1177/0962280215594494.

Combination Drug Study Seeking an MTD Contour:
  [4] Zhang L. and Y. Yuan. (2016) A practical Bayesian design to identify the maximum tolerated dose contour for drug combination trials. , Statistics in Medicine, 35, 4924-4936.