11. Interval estimation for dose-finding studies

Team: Prof Werner Brannath (University Bremen), Prof Luc Bijnens (Janssen)
Clinical Advisor: Dr John Warren
, director of Medicines Assessment Ltd and ample experience in pharmacology and medicines regulation. He previously worked as a medical assessor at the MHRA and is a former member of the European Scientific Advice Working Party.
Location: University Bremen (Bremen, Germany)

Finding the correct dose is crucial during the early phases of most drug developments16 . Usually, a maximum tolerated dose is estimated during a phase I study. Hereafter, we will mostly be concerned with phase II studies, where statistical considerations are used to help in the selection of a dose to use in a subsequent confirmatory phase III study. Statistical objectives in such a phase II study are the estimation of the minimal effective dose (the smallest dose level whose expected response is larger than the placebo response by some pre-specified margin) or estimation of the median effective dose (ED50).

Two approaches exist for the analysis of phase II dose-response studies. First, in a modelling approach, a specific form of the relationship between the dose level and the mean response is assumed. The data from the study are then used to estimate the parameters of the dose-response shape within a certain class of parametric nonlinear models. Techniques such as modelling averaging, that combine predictions from multiple models, are frequently used to avoid the dependency on modelling assumptions from a single model. Second, in a multiple-comparison procedure, pairwise comparisons between the dose levels are performed. The estimation of the quantities of interest, such as the minimal effective dose, are then based on which of the pairwise comparisons achieve statistical significance. Depending on the size of the study, multiplicity adjustment may be applied to the comparisons. Such adjustments either increase the critical levels of the comparisons or they constrain the ordering of the comparisons and requiring that certain comparisons must give significant results before considering comparisons of lower ordering (for example, performing sequential t-test of decreasing dose levels against a placebo zero dose). An interest combination of these two adjustment strategies is to allow the ordering of the comparisons to be partially data-dependent. Contingent on a parameter that governs the degree to which the ordering can depend on the data, this strategy includes the sequential t-test as well as the Dunnett test as special cases17.

While the calculation of confidence intervals is mandatory in many areas of clinical research, confidence intervals for dose-finding studies have not yet been thoroughly investigated. The goal of this project will be the development and analysis of confidence intervals for dose-finding studies based modelling techniques on the one hand, and on multiple comparison procedures on the other hand. First, for the modelling approach, it has been suggested to use bootstrap to construct confidence intervals18 . However, it may also be possible to obtain exact confidence limits using ideas in the spirit of Hotelling19 .

Second, for confidence intervals based on multiple comparison procedures, Wolfsegger et al25 proof control of the coverage probability for the minimal effective dose. We believe that their procedure can be used for other quantities of interest as well.

Meet our Early Stage Researcher: Saswati Saha, University Bremen

Saswati New

I previously worked in Risk Analytic in a renowned MNC in Bangalore, India. I completed my graduation and post-graduation in statistics from Indian Statistical Institute, Kolkata. I had always wanted to work in projects where I can provide analytical insight through extensive data management and statistical analysis. After working in financial risk analytic domain for 2 years, I have decided to change my domain to medical statistics and explore better opportunities of analytical work in the new domain. IDEAS gave me the golden opportunity to work as an early stage researcher for developing statistical methods for early drug development and start my PHD program in University of Bremen from January 2016.

Apart from my academic interest I am also passionate about travelling, exploring various parts of the world and cooking or rather learning to cook various types of dishes and cuisine. I am also passionate about hiking and I am willing to take part in climbing / hiking adventures in the future.


16 Bretz F, Hsu J, Pinheiro J and Liu Y (2008). Dose finding – a challenge in statistics. Biom J, 50:480-504.

17 Wolfsegger M, Gutjahr G, Engl W, Jaki T (2014) Biometrics, ;70(1):103-09.
18 Bornkamp B, Pinheiro J, Bretz F. (2009) MCPMod: An R package for the design and analysis of dose-finding studies. J Stat Softw, 29, 1-23.
19 Knowles M, Siegmund D (1989). On Hotelling’s Approach to Testing for a Nonlinear Parameter in Regression. Int Stat Rev, 57:205-220.


Project News 1 of 10 pages

Meet the IDEAS Network partners

The IDEAS network brings together leading institutions in the training of medical statisticians, large and small sized pharmaceutical companies and is supplemented by clinical experts.