Bayesian Statistics: Model to predict recruitment and inform supply chain

Duration: 60 minutes


**For more information, please see PharmaLex Privacy Policy.

If you do not wish to receive any communication from us, you may unsubscribe at any time.

Why watch


The smooth running of patient enrolment is a key determinant of success for clinical trials. Yet many trials fail to complete on time due to delays in patient recruitment. Indeed, more than 80% of clinical trials do not reach recruitment targets on schedule (Huang et al., 2018). Despite efforts over multiple decades to identify and address barriers, recruitment challenges persist. This causes delays in drug submissions, then shorten the duration of the licence of exploitation of the product, and so a delay generates a significant loss of income. These delays may also end up with a premature termination of the trial due to lack of recruitment, leading to major financial issues but also to ethical issues when patients have been followed in a trial that cannot give robust results due to a restricted sample size. Thus, accurate predictions of the duration to complete recruitment is therefore of crucial importance for conducting of clinical trials

This webinar presents a statistical model which addresses this issue by providing crucial information regarding two practical questions. First, given the current status of an ongoing trial, how long will it take to recruit the remaining patients needed? Second, how many additional centres should be opened in order to ensure completion within timelines? More practically, the proposed methodology is carried out under the Bayesian framework and aims at predicting the randomisation dates of future patients in the context of ongoing multicentre clinical trials. At any time during the conduct of a trial, informative and quantitative metrics for decision-making are provided: First, it gives predictive probabilities of completion at a given date in the future; Second, it provides, for a required level of credibility, the recruitment period estimated to complete the study.

Additionally, the approach can be adapted to follow additional event occurrences, e.g. Progression Free Survival (PFS) or Overall Survival (OS) in oncology trials. Also, the predictive results can be used to feed the supply chain with valuable information such as time when each centre can be short of supply and need shipping of material, prediction of clinical material needed over time and therefore optimal time to initiate production and shipping to minimize the probability of short supply. All that valuable information can be derived under various scenarios to minimize the loss in clinical material and therefore considerably reduce the costs.


Key Learnings

  •   Time-to-event approach within a Bayesian modeling framework allows to adapt for multiple recruitment scenarios
  •   Bayesian approach proposed has been shown in our cases studies to be effective to support commissioning decisions
  •   The approach can be adapted to more complex design e.g. trials with additional event occurrences as in oncology
  •   The approach can be extended to address other fields of clinical trials such as supply chain

A full hour of inspirational speaker(s)

Bruno Boulanger 

Senior Director, Global Head Statistics and Data Science 

Bruno Boulanger has 30 years of experience in several areas of pharmaceutical research and industry including discovery, toxicology, CMC and early clinical phases. He holds various positions at Eli Lilly in Europe and in USA. Bruno joined UCB Pharma in 2007 as Director of Exploratory Statistics. Bruno is also since 2000 Lecturer at the Université of Liège, in the School of Pharmacy, teaching Design of Experiments and Statistics. He is also a USP Expert, member of the Committee of Experts in Statistics since 2010. Bruno has authored or co-authored more than 100 publications in applied statistics and co-edited one book in Bayesian statistics for pharmaceutical research.


Clément Laloux 

Specialist Statistics & Data Sciences

Specialist Statistics & Data Sciences at Pharmalex Belgium, I have a background in business engineering (bachelor and master) where I specialized myself in supply chain management. During my master, I had the opportunity to do an internship at Pharmalex Belgium where I discovered statistics in its practical aspects as support to various business activities. This experience gave me the motivation to complete my formation with a master in Statistics. 
During both my internship and this second master, I worked and specialized on models to predict the recruitment in clinical trials. For my internship, I worked on developing a tool to use information from the predictions of the model to inform the supply chain (expected demand, production information…). For my master thesis in Statistics, I worked on the influence of prior on the predictions of ongoing clinical trials.
As a Specialist Statistics & Data Sciences, I’m providing statistical support on various fields of business, such as clinical, pharmaceutical, or cosmetical activities. The projects I worked on are notably related to method validation, study of treatment effects or reporting and exploratory studies in clinical trials.