Who is this event intended for?&nbs
p\;Statisticians with an interest in understanding causal inference method
s or in applying these to practical situations in drug development
\n
What is the benefit of attending? \;To understand the
potential of the practical application of causal inference methods in dru
g development and be able to apply these to real world problems or clinica
l trials.
This webinar is free of charge to both Me mbers and Non-Members of PSI.
\nTo register f or this event\, please click here.
\nThe event wi ll be structured as two webinars in consecutive weeks\, each of 2.5 hours.
\nThe first webinar will provide an introduction to causal inferenc e ideas and methods and how these relate to the estimand framework in both the setting of RCTs or real world data. Graphical methods for communicati ng causal networks will also be outlined.
\nThe second webinar is a imed at illustrating real practical applications in drug development using case studies of how such ideas can provide valuable understanding of the effects of treatments in the presence of intercurrent events or where effe cts may be mediated by intermediate factors.
\n\;
\n\n Speak er \n | \n Biography \n | \n
Abstract \n |
\n
| \n
Dr. Kelly Van Lancker is a postdoctoral researcher in biosta tistics at Ghent University\, Belgium\, where she also obtained her PhD. P reviously\, Kelly was a postdoctoral researcher at the Johns Hopkins Bloom berg School of Public Health. Her goal is to develop innovative designs an d analytical techniques for drawing causal inferences in health sciences. A big part of her research focuses on more accurate and faster decision-ma king in randomized clinical trials by making optimal use of the available data. She thereby mainly focuses on covariate adjustment\, data-adaptive m ethods\, complex designs\, estimands and especially a combination of these topics. Her recent research is primarily aimed at learning about the oppo rtunities and challenges in running pragmatic trials within clinical pract ice\, and developing better prediction tools for personalized medicine. \n |
\n The role of causal inference in clinical trials: an introduction \nIn light of the recently publ ished ICH E9(R1) guideline on estimands and sensitivity analysis (2019) an d the FDA draft guideline on covariate adjustment (2023)\, causal inferenc e is progressively taking a more prominent role in the landscape of global drug development. In this talk we will try to bridge the gap between comm unities by elaborating on how this field provides a convenient\, unifying framework\, language and relevant tools to formally establish causal relat ionships. We will hereby illustrate how causal thinking\, combined with im portant tools such as potential outcomes\, can facilitate defining\, ident ifying and estimating treatment effects. Building on this\, we discuss the role of causal inference in different trial settings\, including targetin g intention-to-treat effects with covariate adjustment\, handling intercur rent events and the incorporation of external control data. \n |
\n
| \n Ily a Lipkovich is a Sr. Research Advisor at Eli Lilly and Company. Ilya recei ved his Ph.D. in Statistics from Virginia Tech in 2002 and has >\; 20 ye ars of statistical consulting experience in pharmaceutical industry. He is an ASA Fellow and published on subgroup identification in clinical data\, analysis with missing data\, and causal inference. \n | \n This presentation revisits recent ICH E9 (R1) Addendum on Estima nds and Sensitivity Analysis in Clinical Trials and discusses various stra tegies for handling intercurrent events (ICEs) using the causal inference framework. The language of potential outcomes (PO) is widely accepted in t he causal inference literature but is not yet recognized in the clinical t rial community and was not used in defining causal estimands in ICH E9(R1) . I will try to bridge the gap between the causal inference community and clinical trialists by advancing the use of causal estimands in clinical tr ial settings and illustrate how concepts from causal literature\, such as POs and dynamic treatment regimens\, can facilitate defining and implement ing causal estimands for different types of outcomes providing a unifying language for both observational and randomized clinical trials. I emphasiz e the need for a mix of strategies in handling different types of ICEs\, r ather than taking one-strategy-fit-all approach and suggest that hypotheti cal strategies should be used more broadly and provide examples of differe nt hypothetical strategies for different types of ICEs. \n < /td>\n |
\n
| \n Martin Ho is an ASA Fellow\, and a Senior Director of RWE Rare Disease\, Evidence Generation Platform at P fizer\, leading all RWE activities for Sickle Cell Disease assets. Prior\, he was the head of biostatistics at Google\, LLC for 2 years and served t he public at the U.S. FDA for 13 years\, with the last 3 years as Associat e Director of the Office of Biostatistics and Epidemiology at the Center f or Biologics Evaluation and Research. Before joining the U.S. FDA\, he wor ked as senior biostatisticians in contract research organizations for clin ical studies. He co-led the RWE Scientific Working Group of the ASA Biopha rmaceutical Section and the Working Group published 5 papers in landscape assessments as well as state of the science prospects of RWE for regulator y considerations. One of the papers introduces a casual inference roadmap for RWE design and analysis in regulatory context. \n | \n As real-world data (RWD) become more readily available\, the regulatory agencies\, medi cal product developers\, and other key stakeholders has increasing interes ts in exploring the use of real-world evidence (RWE) to support regulatory decisions alternative to traditional clinical trials. To facilitate and p romote statistical research in design\, analysis\, and interpretation of R WE studies for regulatory decision making\, the ASA Biopharmaceutical Sect ion established the RWE Scientific Working Group to address challenges and identify opportunities in the statistical research of this area. In a Wor king Group publication in 2022\, Ho and colleagues (DOI 10.1080/19466315.2 021.1883475) have proposed a causal inference roadmap for study design and analysis that generates RWE for regulatory considerations. In this talk\, Martin will briefly review the steps of the roadmap before using an examp le to illustrate how to apply the roadmap to generate RWE for regulatory c onsideration. \n |
\n
| \n I am currently a statistician with Novartis based i n Basel\, Switzerland. I obtained my Bachelor&rsquo\;s degree in Statistic s from the University of Michigan in my hometown of Ann Arbor and then a P h.D. in Biostatistics from Harvard University in 2020. My doctoral dissert ation focused on statistical methods for dealing with missing data when th e &ldquo\;Missing at Random&rdquo\; assumption does not hold. My current w ork at Novartis focuses on leveraging causal reasoning in the pharmaceutic al industry. \n | \n Single-World Intervention Graphs f or Defining\, Identifying\, and Communicating Estimands in Clinical Trials \nConfusion often arises when attempting to ar ticulate target estimand(s) of a clinical trial in plain language. We aim to rectify this confusion by using a type of causal graph called the Singl e-World Intervention Graph (SWIG) to provide a visual representation of th e estimand that can be effectively communicated to interdisciplinary stake holders. These graphs not only display estimands\, but also illustrate the assumptions under which a causal estimand is identifiable by presenting t he graphical relationships between the treatment\, intercurrent events\, a nd clinical outcomes. To demonstrate its usefulness in pharmaceutical rese arch\, we present examples of SWIGs for various intercurrent event strateg ies specified in the ICH E9(R1) addendum\, as well as an example from a re al-world clinical trial for chronic pain. LaTeX code to generate all the S WIGs shown is this paper is made available. We advocate clinical trialists adopt the use of SWIGs in their estimand discussions during the planning stages of their studies. \n |
\n <
em> | \n Stephen Ruberg\, PhD spent 38 years in the pharmaceutical industry and is currently the founder and President of Analytix Thinking\, a consulting company dedicated to advancing the us e of statistics in the design of clinical trials\, the conduct of analysis \, and the interpretation of scientific data. His present interests are in estimands\, subgroup identification\, Bayesian statistics and digital med icine. He is a Fellow of the American Statistical Association\, the Intern ational Statistics Institute and was a member of an advisory committee to the Secretary of Health and Human Services in the US federal government.\n | \n Estimating Treatment Effects in Patients Who Adh ere to Treatment (Part 1) \nThe estimation of t reatment effects has traditionally been based on the value of randomizatio n and the causal inference it confers. However\, causal inference from ran domized controlled trials requires that all patients be analyzed as random ized AND\, importantly\, that all patients be followed for the duration of the trial and the primary outcome measured. Since many large or long-term trials involve patients who discontinue the study or discontinue their st udy treatment\, this approach &ndash\; often called intent-to-treat (ITT) &ndash\; actually becomes an estimate of the effect of initiating (or bein g assigned) a treatment and NOT the effect of actually taking the treatmen t\, which we call the direct treatment effect. An alternative approach is to censor the data from the time of treatment deviation and impute the res ulting missing values (e.g.\, a hypothetical strategy). This approach uses all randomized patients but requires strong assumption on the potential o utcome after the deviation away from the randomized treatment. While ICH-E 9 recommended the ITT approach in general (or at least the use of all rand omized patients in the analysis)\, ICH-E9(R1) has opened the door to other possible estimands and strategies for estimating a treatment effect. One such alternative is the direct treatment effect in patients (principal str atum) who actually would take/adhere to a treatment (Adherers Average Caus al Effect &ndash\; AdACE). This lecture will be divided into two parts: th e first will motivate why such an estimand is of major importance\, and th e second will provide technical details on its estimation using causal inf erence methods. Examples will be given to highlight the methods\, the code needed\, and the interpretation of such the AdACE estimate. \n |
\n
| \n Yongming Qu is currently a Vice President at Eli Lilly and Company. He received his PhD in Statisti cs from Iowa State University in 2002. He has made significant contributio ns in all phases of clinical development at Lilly and has been active in r esearch for using novel analytics and statistical methods in drug developm ent. He is a Fellow of American Statistical Association. \n | \n <
p>Estimating Treatment Effects in Patients Who Adhere to Treatment
(Part 2) \n \n (Abstract as above)\n |
\n
| \n Sean is a Principal Statistical Scientist at Roche\, UK. Prior to this\, he was a Research Associate at the MRC Biostatistics Unit\, University of Cambri dge\, which is also where he obtained his PhD in biostatistics. At Roche\, Sean works on establishing new endpoints in neuroscience\, understanding the operating characteristics of treatments on clinical outcomes through c haracterizing drug concentration-response relationships and disentangling distinct pathways from treatment to outcomes via potential mediators\, as well as on real world observational studies for making indirect treatment comparisons and estimating long-term treatment effects. He also supports t he conduct of ongoing Phase 3 and post marketing requirement studies in mu ltiple sclerosis as a study statistician\, and collaborates with academics \, external companies and multiple sclerosis registries in his projects. H is research interests include longitudinal data\, causal inference\, missi ng data and composite endpoints. \n | \n Comparative sa fety analysis of time-varying exposures in post marketing observational st udies \nHealth authorities often mandate licens e holders of approved treatments to conduct post marketing observational s tudies to sufficiently assess long-term risk of important safety events\, e.g. malignancies\, since randomized clinical trials are typically too sho rt and underpowered to detect treatment effects on such events. Furthermor e\, comparative safety analysis of newly approved versus other already app roved treatments may be requested as part of the post marketing requiremen t. However\, performing comparative safety analysis of long-term observati onal studies where treatment assignment is based on clinical practice is c hallenging and not well established in the regulatory setting\, particular ly when treatment switching (from control to active and vice versa) is ant icipated to be frequent and often occurs prior to safety events of interes t. Using a case study for OCREVUS\, which is an approved treatment for adu lt patients with relapsing or primary progressive forms of multiple sclero sis\, I will describe one specific post marketing requirement from the FDA on comparative safety analysis\, the challenges of performing such analys es in the presence of multiple treatment switching\, and highlight severe limitations of conventional methods based on time fixed treatments. I will then describe how established methodology for drawing causal inferences f or the effects of time-varying exposures in the presence of time-dependent confounding\, e.g. marginal structural Cox models\, can address limitatio ns of the conventional methods\, and provide feedback from the FDA on the use of causal inference methodology in this observational setting. \n |
\n
| \n Marti n is a statistician with close to 15 years experience in the pharmaceutica l industry and currently employed by Novo Nordisk A/S. He combines regular activities in clinical trials with an interest in statistical methodology \, including the use of causal inference. \n | \n Media tion analysis for a cardiovascular outcome trial \nThere is a growing interest in statistical analyses that can answer qu estions concerning how a drug may affect an outcome via intermediate varia bles (mediators). The LEADER trial is an example. The trial showed a benef icial effect of the drug liraglutide on cardiovascular outcome in people w ith type 2 diabetes and high cardiovascular risk. Key opinion leaders as w ell as regulatory agencies asked whether the effect on cardiovascular outc ome could be explained by previously known effects of liraglutide on blood glucose levels or body weight. The question is best answered within the f ramework of causal inference which provides methods for statistical analys is but also clarifies the assumptions necessary for a meaningful interpret ation of the results. \nIn this presentation\, we will c onsider some selected methods for causal mediation analysis that will be a pplied to the LEADER data. The methods include an approach developed joint ly with experts from academia which specifically handles the case where th e outcome is a time-to-event variable and the mediator is repeatedly measu red. \n |