IP2: MEMORY AND PERCEPTION: THE ISSUE OF RECALL BIAS IN PATIENT-REPORTED OUTCOMES

Moderator: Nancy Kline Leidy PhD, President and CEO, The MEDTAP Institute at UBC, Bethesda, MD, USA

Panelists: Donald E. Stull PhD, Research Scientist, The MEDTAP Institute at UBC, Bethesda, MD, USA; Olivier Chassany, Département de la Recherche Clinique et du Développement Assistance Publique, Hopitaux de Paris Hôpital Saint-Louis, Paris, France

ISSUE: Recall bias in patient-reported outcomes: Fact or fiction? OVERVIEW: Patient-reported outcomes (PROs) provide essential data for understanding and communicating the value of new pharmaceuticals from the patient's perspective. As their role and importance increase, the validity and reliability of PRO measures and data come under greater scrutiny. Regulatory agencies, such as the U.S. Food and Drug Administration (FDA) or the European Medicines Agency (EMEA), rely, to a large extent, on the scientific community for information on the reliability and validity of outcome measures. Concern has been raised about the “best” or “most appropriate” recall period for capturing PROs in pharmaceutical and device trials. This issue has been examined extensively in the survey research field for several decades; only recently has it become a focus in the health outcomes arena. Early findings from this nascent health outcomes research area suggest that there is no single correct answer; magnitude of recall bias varies across PROs, with divergencies particularly apparent in global assessments, health-related quality of life, symptoms (particularly pain), and satisfaction. The complexity of the task, meaning of the event, duration of the recall period, intervening events and patient demographics can each play a role in affecting recall bias. A key question, therefore, is the magnitude of bias present and its impact on trial results and subsequent conclusions. This panel will juxtapose concerns about and implications of recall bias from industry and regulatory perspectives with scientific evidence from the survey and health outcomes literature to inform the burgeoning debate on recall bias in PROs and encourage further empirical work on this important issue.