4 questions for a data science innovator

There’s exponentially more data available now than ever before in history. The whole healthcare system is being digitized, which means that there are tremendous amounts of data within reach. From this information we can learn about how healthcare is being delivered to patients, what unmet needs these patients have and how our medications are being used in the marketplace. It’s a big effort to prepare for the launch of our innovative medicines and then to ensure that patients and physicians are aware of them and that patients can access them once they are on the market. Data science can help in multiple ways here.

People may understandably be concerned about data privacy, which is something we take very seriously. There are many safeguards in place that ensure that patient data is completely deidentified so that a patient’s identity is protected. For example, we use appropriate security systems to process personal information, and our expert teams are trained on privacy principles and appropriate ways to gather, store, use and ultimately govern this information. Across our company we are committed to ethical artificial intelligence (AI) principles, which include shaping policy, ensuring appropriate data and AI use and avoiding and monitoring for bias in our data sets and insights.

Beyond the amount of data out there right now, there’s so much technology available, like cloud-based computing, to manage and store all of this information, plus really powerful analytical tools to help us gather insights very rapidly.

If you go back 10 years, most of this information was on paper: physician notes, patient charts, diagnostic reports. Sure, back then researchers could review patient charts under appropriate privacy protections, but it’s very labor intensive and the information is limited. Now, a lot of that information has been digitized so we can use our massive storage capabilities and computing power to learn from hundreds, thousands or even millions of patients in ways that we never could before.

One exciting benefit? Patients who may have struggled finding treatments for uncommon or rare illnesses may be able to get the help they need because data analysis more quickly connects the dots between new medications and the patients and doctors who want to learn more. Targeting these individuals with the right information at the right time and place will enable more efficient and effective communication strategies for our innovative medicines.

Real World Evidence (RWE) is produced from Real World Data (RWD), which includes data from the healthcare system about our products and about patients as they navigate their healthcare journeys. It’s what’s actually happening every day in healthcare, and it serves to substantially amplify insights and evidence.

For example, RWD is produced from people going to the doctor, filling a prescription at the pharmacy, having a procedure or using digital tools, like health apps. If patients are in the hospital, how long are they staying? What treatments did they receive? What are their outcomes? The answers to all these questions can be provided through RWE generation, and healthcare companies gain access to this information to varying degrees around the world depending on data privacy laws. And when we do gain access to this information, it’s always in a deidentified manner. That means we don’t know who the individual is and we can’t ever find out.

By contrast, data generated from clinical trials happens under highly controlled situations and therefore does not reflect the real world—the study design selects people for the trial and controls certain conditions of the trial including patient visits and what data is collected.

In 2014, I had the fortune of launching the first RWE Department at Janssen in the U.S., and since then we have seen a very robust growth of RWE across our entire organization. Around that time, we published one of the first FDA post-marketing safety commitments using a RWE study design where we looked at the safety of a treatment in over 27,000 patients with atrial fibrillation (AFib). This was a very unique approach at the time and set the stage for additional innovative post-marketing safety-related RWE study designs used today.

Last year, in the context of rapid growth of RWE and data science across all of Johnson & Johnson, I helped form the Global Commercial Data Science team in the Janssen Global Commercial Strategy Organization. This is the fourth RWE and data science team that I have had the privilege to launch over the past eight years at Johnson & Johnson. The goal of this new team is to rapidly accelerate our growth in the fields of commercial data, data science, RWE, precision medicine and digital health.

I’m very excited about the possibilities when you combine these areas with other experts in marketing, medicine, market access, epidemiology, strategy, digital health, legal—and much more—to advance human health by bringing novel therapeutics to patients.

The Heartline Study is a nationwide study that aims to determine whether an app-based heart health engagement program used in combination with Apple Watch’s irregular heart rhythm notification and electrocardiogram (ECG) screening technology can potentially help lead to the earlier detection and diagnosis of AFib—and allow for patients’ physicians to consider treatments that might ultimately reduce the likelihood of stroke.

It’s such an important area of research: I learned a lot about the unmet medical need of patients with AFib while I was leading the R&D team for a Janssen medication in that area, as well as the devastation that stroke can cause for patients and their families.

Around 3 million people in the U.S. have AFib, and up to 30% of these cases go undiagnosed until potentially deadly complications occur. So it’s really critical to diagnose these patients as early as possible, so they can discuss treatment options with their physician.

In order to try to find more patients with AFib in the US, we launched the mSToPs study, a randomized, pragmatic, direct-to-participant digital health clinical trial—and a precursor study to the Heartline Study. We found that a wearable single-lead ECG patch can significantly increase AFib detection in an asymptomatic and undiagnosed at-risk patient population. It also found decreased hospitalization and emergency room visits in those who wore a patch versus those who didn’t. This type of research helped inspire Apple and Johnson & Johnson to collaborate on Heartline. There are a number of interesting learnings that have come out of the Heartline Study and we expect much more to come.

One big area of opportunity is medication adherence. It’s still true that, on average, by one year after starting a medication 50% of patients have stopped taking it. There are dozens of reasons for this drop-off, such as people saying that they feel fine and no longer want to take a drug. It is estimated that between 25% to 50% of patients around the world don’t take their medications as prescribed. Each year in the U.S., medication non-adherence is associated with 125,000 deaths and 10% of hospitalizations.

Digital tools can play an important role in the future by coaching people and encouraging them to stay on their important—and sometimes even life-saving—medication regimens. One interesting example of this is technologies that send alerts to patients’ doctors when they swallow a pill—or miss a dose. There is a lot of promise for such digital technologies in the future to improve outcomes for patients.

Precision medicine is also a big priority for Johnson & Johnson. Some of our targeted oncology drugs, like ones for lung, prostate and bladder cancer, require a diagnostic test to see if a patient is a candidate for our medications or not. This type of precision medicine is not routine care—it’s new and complex, and data science can help us ensure that we understand the marketplace and the complexity of getting tests done and the results back to physicians quickly so they can help patients get access to these targeted medicines.

Another area of interest is disease awareness and screening tools. We have a major opportunity to help patients understand their own health better, and to educate healthcare providers about how to help them. This includes remote monitoring, which is especially important as care shifts to going to the patient rather than the patient going to care, as with telehealth. We want to allow people to recognize signs and symptoms on their own and encourage them to seek medical care—and direct them to the right type of care when they do.

Just like Apple Watches can help detect AFib, there are other digital health tools being developed that may be able to help screen for abnormalities that need medical attention. For example, there may be data science-driven ways to screen patients using remote monitoring devices for common or rare cardiovascular, oncologic, immunologic, neurologic and psychiatric conditions, as well as infectious diseases.

Being successful in this field really takes a village and all groups across Johnson & Johnson have a part to play. We have some amazing talent in this area in Johnson & Johnson and are rapidly growing our teams. We seek to grow awareness of data science and digital capabilities across the company so that we all contribute to our mission of transforming the trajectory of health for humanity. It’s an exciting time to work in medical innovation, and I can’t wait to see how else we can use data science-derived insights to drive progress that will transform the health of people everywhere.