PSI Webinar: Patient preference studies
Date: Tuesday 13th October 2020
Time: 14:00 - 15:30 (BST)
Speakers: Sheila Dickinson, Rachael DiSantostefano, and Gaëlle Saint-Hilary.
You can now register for this event. Registration will close at 12:00 on 12th October 2020.
Free to attend
To register your place, please click here
Patient preference studies are becoming more frequently used in drug development. In this webinar you will hear an introduction to what a patient preference study is as well as an overview of where this type of study can inform regulatory decision making. This will be followed by 2 examples looking at potential approaches to eliciting patient preference demonstrating how such studies can be designed and analysed.
Sheila Dickinson is a Senior Expert Quantitative Safety Scientist at Novartis Pharma AG. She is on the management board of the IMI PREFER project, which is working on developing guidelines about when and how to perform patient preference studies to support medical product decision-making. Her other responsibilities include supporting patient preference study activities within Novartis, and promoting and supporting the use of a structured benefit-risk approach by project teams.
Sheila holds a degree in mathematics from Imperial College, London and an MSc in Medical Statistics from the London School of Hygiene and Tropical Medicine.
After joining Novartis in 1997, she worked as a statistician supporting projects in the various disease areas including both diabetes and malaria, before moving to the Quantitative Safety group in 2013.
Prior to joining Novartis Pharma, Sheila worked as a statistician for Pfizer.
How patient preference studies can inform decisions during drug development.
Patient associations, industry, regulators and HTA bodies agree that patients’ views should inform drug development activities. Patient preference studies are one way to assess patients’ views, and can be particularly helpful in 2 types of scenario: where there is a need to understand which issues in a disease matter to patients, and where there is a need to understand the acceptability to patients of a benefit-risk trade-off (i.e. it’s not clear that a particular drug is going to be the obvious choice for all patients e.g. because this particular drug offers strong benefits at the expense of important side-effects; because patients need to make a choice between very different alternatives such as surgery vs. drug therapy etc.). This presentation describes how patient preference studies can inform routine regulatory decisions, and discusses an example patient preference study assessing which endpoints matter to patients with COPD. It also introduces a framework that offers a structure to guide a preference study sponsor through key issues when design, conducting and analysing a patient preference study, and that supports the discussion between industry, regulators and HTA bodies about preference studies intended to inform medical product decision-making.
Rachael L. DiSantostefano, MS PhD, is a Senior Director of Benefit-Risk in the Epidemiology Department within Janssen Pharmaceuticals, R&D, LLC. She has more than 25 years of pharmaceutical research experience across the quantitative disciplines of epidemiology, biostatistics, and health outcomes. Currently, she focuses on benefit-risk assessment and quantitative patient preference research. Dr. DiSantostefano is also an active member of PREFER, a 5-year public-private partnership that examines how and when to perform and include patient preference studies in decision making during the medical product life cycle. Her research interests also include drug safety, observational studies, and innovation in observational studies.
Parent Preferences for Delaying Insulin Dependence in Children: A Discrete Choice Experiment.
Authors: Rachael L DiSantostefano1, Jessie Sutphin2, Joseph A Hedrick3, Kathleen Klein2, Carol Mansfield2
1Janssen Research & Development, LLC, Titusville, New Jersey.
2RTI Health Solutions, Research Triangle Park, North Carolina.
3Janssen Research & Development, LLC, Raritan, New Jersey.
Background: Screening for auto-antibodies can identify children at increased risk of progression to type-1 diabetes (T1D) that requires insulin.
Objectives: We investigated parents' preferences for treatments to delay the onset of insulin dependence in children who are at high risk.
Methods: A web-based survey (n = 1501) was administered to a stratified sample of parents (children <18 years) in the United States from an online panel. Parents were told to hypothetically assume that their youngest child would become insulin dependent within 6 months or 2 years and were offered a series of choices between no treatment and two hypothetical treatments that would delay insulin dependence. Random-parameters logit analysis and latent class analysis were used to evaluate the relative importance of treatment benefits and risks overall and by groups of parents with unique preference sets.
Results: Most parents chose at least one active treatment. For parents of children without T1D (n = 901), delaying insulin dependence and reducing the risk of long-term health complications and serious infection were the most important treatment attributes. There was some heterogeneity of preferences.
Conclusions: When told to assume their child would develop T1D, most parents considered active treatments to delay progression. For medicines under development to delay insulin dependence in T1D, the preferences expressed in this survey provide guidance on acceptable benefit-risk trade-offs.
Gaëlle Saint-Hilary is Statistical Methodologist at Servier (France) since 2018. She started in 2006 as statistician on clinical projects, first at Servier and then at Novartis Oncology, where she was responsible for the clinical development and the licensing of medicinal products in neuropsychiatry and leukemia. Passionate statistician, she decided to go back to university, and she obtained in 2018 a PhD on “Quantitative Decision-Making in Drug Development” at the Polytechnic University of Turin (Italy), where she continues to conduct research projects. Her main scientific interests are benefit-risk assessment, knowledge and preference elicitation, historical data and quantitative decision-making.
Graphical Elicitation Framework for Trade-Offs and Preferences (GEF-TOP).
Authors - Gaëlle Saint-Hilary (Servier, Polytechnic University of Turin); Pavel Mozgunov (Lancaster University)
Multi-criteria decision analyses (MCDA) have been proposed to perform drug benefit-risk assessments, incorporating preferences of the decision-makers regarding the relative importance of the criteria. These approaches require upstream work to capture the trade-offs the stakeholders make between multiple benefits and risks. Discrete Choice Experiment (DCE) and Swing-Weighting (SW) are the most popular methods for eliciting criterion weights from contributors (patients, experts...) in benefit-risk analyses. While DCE requires a large sample size and might not be appropriate in situations where the number of stakeholders is limited, SW permits to collect preferences from a panel of few participants. However, SW usually necessitates noticeable cognitive requirements, as the contributors are asked to quantify the relative importance of changes on the criteria, which may be difficult to provide. Moreover, many implementations of SW require a consensus to be reached by the stakeholders. Therefore, SW may be demanding to be applied in practice. To address these concerns, we propose a Graphical Elicitation Framework for Trade-Offs and Preferences (GEF-TOP) as an alternative to SW for preference elicitation in the setting with a small number of stakeholders. Via the visual representation of the benefits and risks criteria and few simple questions phrased in terms of treatment performances, the approach permits to limit the participant effort while ensuring accurate capture of the preference information. The questions are constructed to maximize the precision of the weight estimates. We present the application of the approach to linear and non-linear benefit-risk aggregation methods, and study its patterns of behavior compared to SW in a comprehensive simulation study.