Accelerating Research with AI and Data Interoperability

One of the most exciting and innovative developments in healthcare today is the emergence of new medicines that show promising results in treating diseases that affect very large numbers of patients around the world. The most notable of these are new drugs for Alzheimer’s disease and obesity.

While therapies for Alzheimer’s are still in their early stage, the pace of discovery in this disease state is accelerating, promising hope for the millions of older people who suffer from this terrible illness and their families. These new therapies also hold the potential for reducing the expense of treating patients with Alzheimer’s, which cost the U.S. economy alone more than $300B in 2022. [1 ]

The emergence of the GLP-1s (Glucagon-like peptide-1 agonists) has been even more transformative. Obesity is the root cause of many diseases, including diabetes as well as some types of heart disease and cancer, and the ability to address that root cause directly could completely change how clinicians approach the treatment of numerous current illnesses.

The GLP-1s are expensive and Alzheimer’s treatments will be even more costly. Given the size of the patient populations which could benefit from these drugs and others that will follow, there are serious concerns the U.S. health system could be overwhelmed by the cost of drugs in just a few disease categories. [2 ] This is a challenge society will have to wrestle with as bio-pharmaceutical innovation advances.

While drug pricing is a complicated, multifaced issue, certainly the R&D costs of bringing new drugs to market contributes to the price of new medicines. A recent article in the prestigious journal JAMA estimated the cost of conducting clinical trials at more than $800M if the costs of failed compounds and capital were included. [3 ]

AI for automating manual data entry tasks

AI offers significant advantages in accelerating research and reducing costs by streamlining critical processes such as molecular targeting, identifying the right target populations, and automating routine manual tasks like data entry and analysis. These capabilities not only improve efficiency but also enhance the accuracy and speed of decision-making throughout the research lifecycle.

A major contributor to inefficiencies in clinical trials is data management, including how data is captured, recorded, and analyzed. Because Oracle owns both life sciences and healthcare technologies, including EDC and EHR, we are uniquely positioned to design processes for automating the entry of research data and sharing it with clinicians to help improve patient care.

While automating data transfer from EHR to EDC is an obvious and impactful use case, automation can also play a critical role in drug safety solutions, such as in systems like Oracle Argus and Empirica. AI and data automation can streamline adverse event reporting, improve signal detection, enable accuracy, and faster, real-time insights.

Manual data entry across disconnected systems in clinical trials is not only time-consuming and burdensome but also increases the likelihood of errors that can lead to inconsistencies and compromised data integrity. It also reduces time that can be spent on actual research. Researchers often find themselves bogged down by the manual task of literally transferring data between two computer screens on their desks, one for research participants and one for clinical patients, a frustration known as the 'swivel chair problem.'

The average oncology trial has three million data points. It takes an estimated three minutes to find, input, and verify each data point as it is transferred from an EHR to an EDC. For one average five-year oncology trial, the math adds up to 15 person-years of work, per year, purely spent on copying data! Considering there are close to 4,000 oncology trials started every year, that’s close to 60,000 person-years spent on data input rather than on taking care of real research-related tasks. [4 ]

AI technology can perform real-time quality checks, to help ensure that the data transferred from EHR to EDC is accurate, complete, and formatted correctly. That’s why Oracle is building AI-embedded clinical research solutions to help reduce and eventually eliminate the need for manual data entry. Our EHR-agnostic approach will make AI-driven automation available on any EHR platform, extending the benefits of EHR/EDC integration across the entire research ecosystem.

Building an intelligent, interoperable data ecosystem

The EDC was created and designed to be a unique system to serve the needs of clinical trial researchers before the era of interoperability. But given new federal regulations around interoperability, as well as the speed of technological change, it is possible to rethink the very need for a separate research data collection system.

Today, with embedded AI technology, we can automate the data transfer from EHR to EDC with an EHR-agnostic solution. This is one example of how AI can help bridge the gap between clinical research and clinical care. We are excited to be working with customers to advance this important initiative.

In the future, the EHR could become the “source of truth” for both research and clinical data. Providers will be able to treat a patient in a clinical trial or a regular clinic using the same information system. The need for a separate EDC could well be eliminated.

Oracle is working towards making this a reality. With our unique strengths in AI, cloud, healthcare tech, cybersecurity, and data management, Oracle is poised to lead the revolution of clinical research through our AI-embedded healthcare and life sciences platforms.

Bridging the gap between clinical research and clinical care

Our vision is to bring the worlds of clinical trials and clinical care together to increase the number of clinical trials and participation in such trials, while also decreasing their length and administrative costs, all with the ultimate goal of accelerating medical innovation and the availability of new and presumably less expensive treatments for patients worldwide. This is one way the growth of total healthcare costs can be slowed significantly while accelerating drug discovery.

We are working towards a future where clinical research and insights are integrated directly into patient care, accelerating our ability to move information from the bench to the bedside and back again.

[1] https://www.jec.senate.gov/public/index.cfm/democrats/2022/7/the-economic-costs-of-alzheimer-s-disease

[2] The GLPs have an estimated annual cost before discounts of $12,000, while Eli Lilly’s new drug for Alzheimer’s, called Kisunla, has an annual price tag of $32,000. One estimate suggests that the cost of the GLPs could reach $1 trillion annually if all patients who qualified for the drug received it. See here: https://www.nytimes.com/2024/03/04/opinion/ozempic-wegovy-medicare-federal-budget.html

[3] https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2820562

[4] https://bmccancer.biomedcentral.com/articles/10.1186/s12885-023-11690-9