Biases that Generate Immortal Time: A Structural Diagnosis and Practical Solutions for HTA Submissions

Moderator

Miguel Hernán, Harvard T.H. Chan School of Public Health, Boston, MA, United States

Speakers

Xabier Garcia de Albeniz, Adigens Health Limited, Sant Cugat del Valles, Spain; Thibaut Sanglier, PharmD, MPH, PhD, Basel, Switzerland; Julia Vila, PhD, RTI Health Solutions, Spain

Purpose Artificial introduction of immortal-time due to study design choices remains one of the most consequential and under-recognized threats to the validity of real-world evidence (RWE) used in market access and health technology assessment (HTA). When eligibility criteria are applied after treatment assignment, or when patients are misclassified into treatment strategies, estimated treatment effects can be biased — with direct implications for reimbursement decisions, comparative effectiveness claims, and external control arm acceptability. This workshop will equip attendees with a structural framework to diagnose design decisions that introduce immortal-time and provide concrete, implementable solutions grounded in the target trial framework. Participants will learn to identify the two principal mechanisms generating immortal time, apply design fixes, and critically appraise RWE submitted in HTA dossiers. Description Attendees will gain a working knowledge of how to classify, according to their structure, biases that generate immortal time and some design choices that prevent it. The workshop will cover a) the formal classification of biases that generate immortal time, b) resources to eliminate them at the study design stage, and c) case studies illustrating such applications. Xabier Garcia de Albeniz will introduce, using causal diagrams, the structure of biases that introduce immortal time (15 min.). Thibaut Sanglier will present cloning-censoring-weighting as a tool to address these biases in a case study for breast cancer (15 min.). Julia Vila, will illustrate analytical resources consisting of sequential emulation in a real-world colorectal cancer study and show the consequences of wrong design choices on effect estimates (15 min.). The workshop will close with an exercise to prompt audience participation and reaction to alternative study designs for hypothetical studies susceptible to incurring in immortal time generation. A structured discussion of remediation strategies will be offered (15 min.). This workshop will be especially valuable to RWE generators, HTA assessors, and market access professionals.

Topic

Methodological & Statistical Research, Real World Data & Information Systems, Study Approaches

Your browser is out-of-date

ISPOR recommends that you update your browser for more security, speed and the best experience on ispor.org. Update my browser now

×