A Must-Have Recipe for Real-World Evidence Generation

Real-world evidence (RWE) is crucial for life sciences companies to better understand the patient journey and therapies being used in the real world. RWE can assist clinical development, health economics and outcomes research, and commercial teams across the entire drug development lifecycle. To unlock the full potential of RWE, there is a must-have recipe for success, and Verana Health has just the cookbook for you to view.

In this blog, we use a “cake” analogy to put it all into perspective. Imagine a cake made of the highest-quality ingredients and prepared by professional bakers. Now imagine this cake being capable of accelerating clinical research and helping transform patient care. The “cake” represents RWE, and the ingredients represent real-world data.

Baking the RWE cake requires the very best RWD ingredients, which Verana Health possesses in the form of high-quality, research-ready, exclusive datasets. The baking process entails utilizing secure and advanced artificial intelligence (AI)-driven technology, complemented by a team of expert bakers – industry veterans who provide clinical oversight and data science insight – to ensure the cake is beautiful and delicious.

Using Quality Ingredients

Verana Health’s RWD ingredients consist of de-identified data on patients that are collected from real-life interactions, including structured and unstructured data from electronic health records (EHRs) linked with other real-world sources (e.g., medical claims and images). Historically, structured data have primarily been used for RWE generation because of the ease of use, but it only answers the “what” question (e.g., what medications were used) and not the “why” question (e.g., why the clinician selected that medication). This is like baking a cake with only flour, butter, and eggs. It might produce something that looks like a cake, but what about the sugar, vanilla, cocoa, etc.? That’s where unstructured data comes into play.

Verana Health’s Qdata^Ⓡ modules, which are fit-for-purpose, quality data modules that are designed to confidently drive business insights and inform research outcomes, are high-quality because they meet the following criteria:

• Depth – Our recipe requires EHR data that includes a high-degree of detail and comprehensiveness that are captured at the point of clinical care.
• Validity – Our ingredients come straight from the source through exclusive partnerships with leading medical societies, making them the most accurate and reliable data for the purpose of research. 
• Speed – Qdata are refreshed on a monthly basis. The more recent information that can be placed into the hands of life sciences companies, the more decisions can be made in a timely manner.

The Baking Process

These data ingredients come together in the baking process utilizing Verana Health’s VeraQ^® population health data engine that connects the dots across unfiltered healthcare data. VeraQ is clinician-directed and enhanced by AI to securely power a data integrity feedback loop of nearly a half-billion raw, point-of-care health encounters. VeraQ ingests, curates, and links the high-fidelity data to enable trusted observational research:

• Ingestion – Structured and unstructured raw-source EHR data are ingested from EHR systems directly into VeraQ. By ingesting data from the point of clinical origin, Verana Health is able to retrieve complete, usable data from practice EHRs including unstructured notes. Data curation and attribution of raw source data mapping farther upstream not only involves clinicians earlier in the process, but enables Verana Health to further minimize errors and improve control and positive impact to AI algorithms.
• Curation – The curation process involves harmonizing, verifying, and organizing data to improve the accuracy and usability of information housed in a database for multiple applications. Verana Health applies AI-powered large language models (e.g., machine learning and natural language processing) to analyze patterns of language in unstructured clinical notes that signal key milestones and clinical insights that occur during the patient journey. Verana Health’s team of physicians and data scientists, with deep expertise in data-driven research, continually train and establish rules for how the unstructured data is cataloged and categorized to make it useful in the real world. Additionally, all patient data is de-identified so that any unique identifiers on an individual are removed. 
• Linkage – Once the data is de-identified and curated, Verana Health partners with Datavant, which creates irreversible, site-specific encrypted tokens for each patient record. These tokens preserve patient privacy, while ensuring datasets can be linked. Data from EHRs can be linked with other sources, such as medical and pharmacy claims and images to connect the “what” and “why” in a single study. These data can be used to generate new questions and hypotheses that may inform the development of medical products designed to improve patient outcomes and care. 

Our Expert Bakers 

This RWE cake doesn’t make itself. As mentioned, a team of clinical and data experts come together to ensure the data and the resulting analyses are clinically relevant. To preserve the integrity of the original clinical context from a healthcare clinician’s point-of-care notes and inputs, ML and NLP are coupled with the human touch of our experts. This team is involved throughout the entire data curation process to ensure that each step has clinical validation and that the utilization of the model isn’t occurring in a silo. These experts not only ensure Qdata are research-ready through rigorous oversight and testing, they also support teams in extracting insights from the data.

A Slice of the Cake

A compelling demonstration of Verana Health’s RWE recipe is Qdata Prostate Cancer. Pulled from EHR data, this real-world dataset includes more than 364,000 de-identified patients and an average of more than 2.6 years of follow-up for patients with linked claims, which provides invaluable insights to better understand the patient journey. The ability to capture this data in near-real-time enhances its relevance and utility for understanding individual patient experiences throughout their healthcare journeys. To prepare this massive amount of data as ingredients for an RWE cake, Verana Health employs AI-powered large language models to analyze patterns of language in unstructured clinical notes, enabling the identification of key variables and clinical insights. A team of clinical experts, including experienced urologists, continually train, test and establish rules for cataloging and categorizing these unstructured data. This approach ensures that the dataset is curated in a manner that maintains high-quality and fidelity, making it useful for real-world applications in prostate cancer research. 

If you’re ready to place your RWE cake order, or would like to schedule a consultation with a Qdata expert, visit veranahealth.com.