Digital health is transforming medicine at a breathtaking pace: wearable biosensors, mobile health applications, and electronic health records have revolutionized data collection and processing, allowing low-cost, quick delivery of data from patients directly to healthcare providers, administrators, and analysts. Modern technologies have the potential to dramatically transform health care for the better. It may be tempting to rush in and invest or develop new products in the medtech space, but you can’t just walk into digital health technology and expect it to work on the front lines of clinical care. The mistake many entrepreneurs make is thinking that their tools and experience will apply to healthcare as well, which is rarely the case.
Health care is unlike any other industry in the world. It requires rigorous research and a level of care that no other industry demands. As Director of Health Services Research for Cedars-Sinai Health System, my focus is on applying digital health science to strengthen the patient-doctor bond, improve outcomes, and save money. Currently our team has partnered with Applied VR and Samsung Healthcare to test the use of virtual reality (VR) technology to manage pain in hospital settings. We spend a lot of time placing digital devices on patients and we’ve seen what happens next – it’s not always what we expect.
The very first patient I approached for our study was a 45-year-old woman who had metastatic lung cancer. We approached her with a set of VR goggles, explained how the technology might help with pain management, and described how the experience might offer a temporary “escape” from the hospital. She looked at us silently and unblinkingly, then politely turned down the VR. She died days later. In this case the technology was not appropriate to the situation. There is a very human element to health care, and what may sound like a fantastical voyage to a healthy person may be an unwelcome intrusion to someone with advanced illness.
There are many practical examples of times technology is inappropriate. In our study, 83% of patients with seizures, nausea, stroke, motion sickness, and dementia were medically ineligible to even attempt to use VR. Some patients were in isolation for infection control. We have to be extremely careful about not spreading infection in vulnerable patients.
Sometimes the device itself becomes a barrier. Activity monitors like FitBit may be comfortable or even fashionable in normal situations, but for a patient with joint pain or swelling, wearables can cause additional discomfort. Switching the location of wear may not work either: for example, activity monitors designed to be worn on the ankle may make patients feel like they are wearing parole monitors. These findings may be obvious in retrospect, but our presuppositions about how “easy” or unobtrusive a device might be are easily challenged in practice.
Sometimes the data collected doesn’t align with predicted results. Take for example data collected from a 72-year-old woman who tested an activity monitor as she underwent treatment for rheumatoid arthritis. As her pain decreased, theoretically her activity should have increased. But the sensors reported the opposite: the better she felt, the less she moved. She explained to us that her true passion was writing – a stationary activity that she had been prevented from pursuing because her joints would become stiff and painful, forcing her to get up and move. As the pain receded, she was now able to sit for long stretches at a time and write. Only by understanding the relationship between her pain and lifestyle were we able to put the results into context.
“Applying technology in a medical setting comes with unique challenges. You must work closely with all the stakeholders, patients, providers, and families, then meticulously study and test these interventions.”
Negative data can be clinically useful in other ways as well. I recently evaluated a young woman with severe abdominal pain. She had been extensively tested and all results were negative. Doctors suspected she was suffering from psychogenic pain so we decided to try VR: nothing happened. Her pain broke straight through the illusion. The failure was so absolute that it caused us to re-think her diagnosis. If it were psychogenic we would have expected at least a minor response. We dug deeper and ultimately found a rare but treatable condition in which the major blood vessels in her abdomen were being crushed by a ligament on her diaphragm. Her pain had a striking physical cause, and the failed VR response forced us to look harder and find it.
Applying technology in a medical setting comes with unique challenges. You must work closely with all the stakeholders, patients, providers, and families, then meticulously study and test these interventions. It is a worthy pursuit and has enormous potential to improve treatment outcomes, relieve pain, educate patients, offer alternative therapies, and reduce the length of hospitals stays. What we’re doing here at Cedars-Sinai can be a model for others: partner with medical institutions, spend time with patients, interview patients about their experiences, improve devices based on patient feedback, and strive to understand and interpret the data accurately.