Regulators are evaluating the use of noninterventional real-world evidence (RWE) studies to assess the effectiveness of medical products. The RCT DUPLICATE initiative (Randomized, Controlled Trials Duplicated Using Prospective Longitudinal Insurance Claims: Applying Techniques of Epidemiology) uses a structured process to design RWE studies emulating randomized, controlled trials (RCTs) and compare results.

ECAs have a “long history” in regulatory decision-making and are used when randomization is unethical or unfeasible. They can also serve as a benchmark in health technology assessments. Advances in data access, quality, methodologies, and societal influence have broaden the use cases for ECAs. If data confidence is relatively high, an ECA could become an option when a randomized controlled trial (RCT) is impractical to do. 

The assessment of patients' performance status is important for toxicity monitoring, treatment selection, and clinical trial eligibility. By mining data of >120,000 cancer patients from Flatiron Health, we derived the ROPRO score for overall survival in cancer patients. We are using baseline ROPRO for patient selection as a quantitative, unbiased decision support to determine 12-week life expectancy. Longitudinal delta ROPRO analyses may help clinicians detect treatment benefits or progression early.

The potential role of real-world evidence to advance pharmaceutical research is well recognized. This is especially critical in the development of medicines for patients with unmet needs. While examples abound reflecting the use of such evidence in regulatory decision making, there are challenging methodological and operational issues that require careful attention. We discuss some of the commonly arising issues, along with the pros and cons of routinely used mitigation measures. Recent developments in the use of modern analytics to address confounding and high dimensionality are reviewed, and illustrations are provided from the pertinent literature.    

This talk presents the fast-emerging landscape of RWD/RWE in innovating traditional clinical trials, with a focus to communicate the vision, strategies, the best practice, the areas of success, and the challenges yet to conquer.

Time and motion study design is used in clinical and real world settings to generate evidence to improve efficiencies in healthcare processes and workflow. This presentation will discuss the evolution of this study design from its original industrial application to a valued tool for clinical research today. Topics include how TM studies may differ in design depending on the intended clinical setting, real-world examples, regulatory and ethical procedures, and case studies.

Optum partnered with three provider groups to streamline and automate clinical trials through their digitized solution, the Optum Digital Research Network. This approach to recruitment, data entry and PROs resulted in a more cost-effective, less burdensome process that could benefit practices interested in clinical trial participation and expand diversity of patient populations available for pragmatic trials in a real-world setting. This session will discuss the study methodology, results and participant perspectives.

We developed a framework and tools to conduct regular RWD studies. The objective is to improve operational efficiency and ensure reproducibility/repeatability/transparency. The framework provides best practice recommendations from data standardization to data visualization, and the tools include a suite of R packages to provide a user-friendly implementation.

Radical shifts are happening in the healthcare service delivery system. “Episode Centric” approach is changing to a “continuum of care” approach, with the consumer needs at the very center. Dedalus aims to enable such a revolutionary transformation with its innovative framework bridging the gap between various systems of record while generating contextually relevant insights, improving the clinician experience, and progressing study design and recruitment of more precise sub-populations of patient cohorts.

The widespread adoption of electronic health records (EHRs) alongside the advent of scalable clinical molecular profiling technologies has created enormous opportunities for deepening our understanding of health and disease. However, in most diseases, disease-relevant markers are spread across multiple biological contexts that are observed independently with different measurement technologies and at various time schedules, and their manual interpretation is therefore in many cases complex. Machine learning holds promise for integrating comprehensive, deep phenotypic patient profiles across time for (i) predicting outcomes, (ii) identifying patient subtypes and (iii) associated biomarkers. In this talk, we will outline opportunities and challenges for clinical prediction models built from deep phenotypic patient profiles in clinical research and beyond.

Learn which AI-based technologies are in production for which ICSR process steps. Understand various considerations for planning, implementation, and validation. Understand key learnings from early adopters of AI-based technologies within the ICSR process.