Next Generation Oncology Research
Explore Study Concepts
AI-enriched, population-scale Real-World Data lets you explore and compare patient populations against different eligibility criteria in order to find underserved areas where a therapy would bring the greatest benefit.
Real-World Data linked with other data sources brings a complete understanding of past patient treatment histories, comorbidities, and disease burden for robust, high-confidence clinical study designs and External Control Arms.
Optimize Site & Investigator Selection
The EurekaHealth cloud ecosystem integrates our Real-World Data with your internal and third-party datasets to identify those sites with appropriate patient populations and higher performing investigators for more successful study deployment.
External Control Arms
As regulators encourage novel uses of Real-World Data, a major innovation is External Control Arms as a replacement for traditional randomized clinical trial controls for hard to study diseases.
Our work in External Control Arms involves rigorous, research-grade Real-World Data based on multiple EMRs representative of patients from community and regional hospitals and academic centers and is suitable for regulatory submissions.
Our data avoid the bias of past RCT controls and a single EMR by providing access to patients across care settings and geographies in sufficient numbers to meet the study design and statistical analysis plan, for even the rarest cancers.
Our data scientists and epidemiologists collaborate with you on the study design, from creating the patient control cohort appropriate for regulatory submissions to full analysis and representation of results to regulatory authorities.
Our teams work within an infrastructure that is fully compliant under Good Clinical Practice (GCP) and 21CFRPart11 (effective 9/30/2019).
We are bringing the leading Artificial Intelligence and Machine Learning approaches to Real-World Data management, analysis, and applications.
Our AI models increase confidence in outcomes analyses. One class of AI models accounts for the ‘messiness and missingness’ in Real-World Data to assure highly accurate and reliable data. Another class of AI models helps predict rate of progression, treatment durability and mortality to enable robust study protocol development and inclusion and exclusion of specific subpopulations.
Unsupervised and semi-supervised learning approaches reflect previously unknown patterns and relationships in data to reveal areas ripe for potentially higher value treatment approaches or sources of outcomes variability and unpredictability.
Our AI models have a high degree of mathematical and clinical validity and are designed to:
- Work with our data, your data or third-party Real-World Data
- Be deployed anywhere in the workstream to run optimizations and support next best actions (NBA).
Two classes of our AI models were published at ISPOR 2019 and ASCO 2019:
Using Artificial Intelligence to Improve Capture of Metastatic Breast Cancer Status in Electronic Health Records
Development of an artificial intelligence model to predict survival at specific time intervals for lung cancer patients