Delve Health: Literature Reviews Meets Big Data

24 May, 2016, by ClinCaptureTeam

Wessam Sonbol is an entrepreneur focused on innovative life science solutions. He has 17 years of experience in the life science industry, working with variousElectronic Data Capture (EDC) and Clinical Trial Management Systems (CTMS) solutions, as well as providing analytics and implementation services. He is the Founder and President at DELVE HEALTH, a Minnesota-based technology company leveraging big data to streamline literature reviews, pre-qualify sites/investigators for clinical trials and provide industry insights. Wessam has deep product management, systems implementation, integration, and analytics experience. Prior to DELVE HEALTH, he developed and launched www.tctmd.com, and developed the first Open Source Clinical Trial Management System (CTMS) www.voxce.com.

ClinCapture : Why did you decide to start DELVE HEALTH?
Wessam Sonbol: Wessam Sonbol: The starting point was the realization that we needed a shift in mentality and in technology to streamline clinical research, and literature research in particular. I realized how difficult it was to gather existing data from literatures and clinical trials to identify better treatment options. There is so much data out there, but it takes so long to get what we need out of it and purpose it for our needs. At the same time, we are supposed to be living in the golden era of cloud-based technologies, big data, and advanced analytics. Therefore, I started DELVE HEALTH to help narrow this gap and empower users through Natural Language Processing and Big Data Analytic technologies to provide faster and more accurate literature reviews.

“The starting point was the realization that it is difficult to gather existing data in the R&D world. Data points are often disparate and difficult to access.”

ClinCapture : What are the DELVE HEALTH system intended use cases and who are your users?
DELVE HEALTH gathers data from publications, journals, clinical trials and adverse events to provide a single source of truth, helping researchers and scientists identify better treatment options using Natural Language Processing and Big Data analytics. Researchers, scientists and doctors are always searching journals, publications and clinical trials for disease specific information, to identify prior treatments, understand prior adverse events and learning about new techniques through published documents and clinical trials. Initially, DELVE was a single project for a medical device customer, which then introduced us to multiple companies with similar needs. One of the main use case today revolves around literature reviews, which are used for various reasons such as; create a Clinical Evaluation Report (required by European regulators for clinical trials), Systematic Review to identify enrollment criteria for similar studies or identify trends for certain drugs or devices. We realized that we are able to identify more accurate results than PubMed and save users 15 to 20% of time spent doing the usual manual processes. The system reviews and compares data from multiple literature sources such as the PubMed, Embase, FDA, clinicaltrrials.gov and more. For example, if a user search heart attack with abnormalities, the system will provide a set of reports identifying the number of publications, doctors, devices, manufacturers, clinical trials, KOLs and investigators associated with that search.

“With over 30 million literatures available across nearly 50,000 disease conditions, identifying the right set of documents to review can be quite a daunting task”

ClinCapture : What did the process look like before DELVE HEALTH?
Researchers and Scientists would browse databases manually, download the data in excel, document their findings, merge data across various excel sheets, use sticky notes, analyze the data in excel and identify a way to house and share the data with colleagues. This process is very manual, time-consuming, and potentially error-prone. Conducting literature reviews is like a jigsaw game, where users have to put together the pieces from multiple sources and publications, then describe the resulting picture to the reader, pointing out what parts of the picture are clear, what parts are fuzzy, and what parts are missing altogether. As you can tell, it is often exhausting, manual and often frustrating. With DELVE, the data is safe and sharable from one central repository, one location. Rather than importing data into excel spreadsheets, users can see all their data organized into one comprehensive page, making it easy to review and identify what they need. We have access to 30 millions literature publications across 50,000 conditions. The technology behind DELVE leverages natural language processing algorithms in conjunction with big data techniques.

“DELVE captures, mines and brings data to users’ fingertips in an integrated dashboard.”

ClinCapture : Is DELVE a Clinical Research Management System (CRMS)?
In a sense. But we prefer to call it a clinical literature intelligence platform because it’s focused on literature reviews. We don’t stop at literature though, we also help researchers connect with investigators , hospitals or clinics and doctors that would be a good fit for their next clinical trial. Using the literature, the system can also identify KOLs in specific studies, which is essential for clinical trials to activate sites.

Screenshot: Integrated clinical literature intelligence dashboard in DELVE Health

ClinCapture : Have you integrated DELVE with other eClinical systems?
We haven’t integrated DELVE with EDC or eClinical systems, but we’re in talks with technology vendors and systems providers. We can definitely see a lot of value from integrating DELVE to a CTMS or an EDC. The data we mine out of a literature would then be integrated with CTMS, allowing users to quickly identify the latest information on their Investigators and Sites. When it comes to investigators in particular, they absolutely need to know about recent publications in the field, current clinical trials, new doctors, new conditions that are being picked up, etc.

ClinCapture : How can we sign up to the system?
DELVE is subscription-based, users pay as they go. We have a pricing model by user and by organization. We currently have 23 users on the system and over 2,000 searches performed to date.

“Systems are not smarter than humans but systems can identify things than the human can’t see.”

ClinCapture : How has the industry changed since you entered it nearly 30 years ago, and where do you see it going?
In the last 15 years, the number of EDC and CTMS systems has skyrocketed. Adoption rates are following. People understand they need to utilize technology and cloud-based solutions to save time and money to deliver drugs to market faster. We live in the world of data, where we have a new publication published every 20 minutes and thousands of data points available every day. In order to make better sense of what we have, we need to easily connect the dots and help save more lives sooner. That’s where we believe DELVE will help.