For devices, data from product registries can be augmented by direct collection from a wearable/implant, claims/EHR or via patient/physician facing mobile apps. The constantly evolving landscape of medical devices is fueled by both incremental changes in products and large shifts in surgical practice, including digital health solutions. While the languages of drug and device surveillance are similar, the data sources, tools, and application of methodologies differ, and are often uniquely defined by the specific clinical area. Product surveillance and signal management: Now is the time for methodologist engagement! Mary E Ritchey 1 Danica Marinac-Dabic 2 Theodore Lystig 3 Paul Coplan 4 1Med Tech Epi 2FDA/CDRH 3Medtronic 4Johnson & Johnsonīackground: As pharmacoepidemiologists are realizing that many drugs rely on devices (e.g., for delivery, pre-use testing, follow-up evaluation), the time is ripe to marry the methodologic expertise of drug epidemiology with the deep knowledge of unique aspects of device epidemiology and surveillance. We will conclude with a panel discussion and interactive Q&A with participants (25 min). Format: Using virtual interactive polls, we propose a dynamic interactive 90-min workshop: 1) Introduction (5 min, Dr Brodovicz) 2) Regulatory perspective on RWE to support decisions for medical devices (15 min, Dr Chang) 3) Quality aspects in RWE research (15 min, Dr Leclerc) 4) Study designs and analytic features that minimize bias (15 min, Dr Assimon) 5) How effect size, pre-specification, study design and data quality drive credibility and believability, and steps to triangulate effect estimates (15 min, Dr Girman). Finally, our panel will interact with the audience on various protocol influences of believability and how larger effect sizes, fit-for-purpose data for the specific research question, and transparent analytic techniques minimize the risk of various biases, thereby maximizing believability and robustness of results. Pre-specified replicability approaches may also enhance believability. Special note will be made on using confidence intervals over p-values for interpretation, as the latter are driven by larger sample sizes typical in RWE. A pre-specified statistical analysis plan, including statistical assumptions verification, sensitivity analyses, and bias quantification techniques, are essential to ensure believability. Techniques to prevent or minimize information bias will be discussed, such as ensuring that data are complete, reliable and relevant for the specific research objective (i.e., fit-for-purpose). Designs and methods that minimize confounding are preferred (e.g., new-user active comparator design, propensity score approaches).

For the RWE & Regulatory Decisions workgroup of ISPE RWE Task force, we will review approaches to improve believability of RWE to regulators and health authorities.ĭescription: Foundational to believability are well-articulated pre-specified objectives, study designs and analysis plans detailed in registered protocols. Attendees should expect to learn more about what health authorities need to evaluate robustness and believability of RWE results. Objectives: To discuss various study designs and methods fostering believability of RWE to guide informed regulatory decisions.

However, what makes RWE results believable to health authorities is not widely understood among pharmacoepidemiologists.

The FDA recently issued a data quality framework and regulators globally are working on guidelines to use RWE for regulatory purposes. SYMPOSIA Believability of real-world evidence (RWE) for informed regulatory decisions Jacinthe Leclerc 1 Kimberly G Brodovicz 2 Stephen Chang 3 Magdalene M Assimon 4 Hisashi Urushihara 5 Cynthia J Girman 6 1Universite du Quebec a Trois-Rivieres 2Boehringer Ingelheim 3FDA 4University of North Carolina School of Medicine 5Keio University 6CERobs Consulting, LLCīackground: Regulators are increasingly considering RWE of medical products to support new indications or expand labeling.

Mike karch floyd total security