Can Technology Transform Behavioral Care? Healthcare Leaders Discuss the Promise and the Hype

Jan. 29, 2018
During a healthcare conference sponsored by MIT’s Sloan School of Management, Don Mordecai, M.D., Kaiser Permanente National Leader for Mental Health and Wellness, shared his perspective on the state of digital health innovation in mental health care.

One in five Americans experience a mental health condition in a given year, and one in 25 U.S. adults experiences a serious mental illness that substantially interferes with major life activities, according to the National Alliance on Mental Illness. Globally, the World Health Organization (WHO) reports that depression is the leading cause of disability worldwide.

These statistics indicate that mental health issues impact a significant number of patients in every health system and healthcare provider organization. With the evolution of digital health tools, can technology help healthcare providers solve these issues? And, if digital health can help with mild to moderate mental health issues, can these tools, such as cognitive behavioral therapy apps, help with severe mental health conditions?

Don Mordecai, M.D., National Leader for Mental Health and Wellness for the Oakland, Calif.-based Kaiser Permanente organization, addressed these issues and gave his perspective on the state of digital health innovation in mental health care while speaking at a health innovation conference Nov. 29 sponsored by the Massachusetts Institute of Technology (MIT). MIT’s Sloan School of Management’s Initiative for Health Systems Innovation (HSI) sponsored the “Innovating Health Systems: Digital Health Transformations” conference at the Cambridge, Mass.-based campus.

Kaiser Permanente, an integrated hospital and health plan organization based in Oakland, California, is a 39-hospital health system with 20,000 employed physicians serving 11.8 million members, with 4 million mental health visits a year, Mordecai said.

Citing statistics on mental health, Mordecai said individuals who enter treatment for a mental health disorder are two times as likely to also be living with a substance abuse disorder, and less than half of adults with mental health conditions received treatment in the past year.

At Kaiser, physicians and researchers are working to leverage data, digital tools and predictive analytics to better address mental health issues among the health system’s patients. “With our EMR (electronic medical record system), we are technologically advanced—we have labs, diagnoses, visit notes, progression of disease. We have incredibly rich data beyond just the claims database,” he said.

Mordecai pointed out that there are currently hundreds of mobile apps targeting mental health and behavioral health. “But, we find that the evidence base is pretty thin. And, where there is evidence that cognitive behavioral therapy through a computer works, the uptake is low. We’ve tested apps, but people don’t use them. That is the central problem—how do you design them so people use them?”

Mordecai said one key issue with designing behavioral health apps is to focus on facilitating human interaction rather than replicating it. “Texting facilitates human interaction. A standalone cognitive behavioral therapy app is trying to replicate it. And maybe that’s why it’s not working.”

He continued, “At this point, standalone digital solutions are not the answer for us yet. What is a winning solution? I think of a trifecta of digital health solutions in which you get uptake, which means people open it and are interested in checking it out, and then engagement, meaning they stay with it, and then, third, it affects their outcomes. From a large health care system perspective, I’m not particularly interested in an app without those three,” he said.

At Kaiser, physician leaders and researchers are focused on building an ecosystem of applications that meet those three standards. “With our EMR system, we have rich data, we can look at outcomes, such are the patients’ labs improving, and look at whether the patients are using the app. We’re working on that ecosystem now,” he said.

The process involves bringing together clinicians, patients and app developers to focus on the human-centered design process, Mordecai said. “How do we design this compelling human element into this product we want to create?” He added, “There is a sweet spot that combines the human element with the digital element to address mild to moderate mental health issues, and you have to grapple with the pitfalls and the challenges of big data, and bringing that to bear on these types of problems.”

Kaiser Permanente is currently an enterprise partner with a digital health company that developed a mobile application called Recovery Record, which is targeted to individuals with eating disorders. The app is an eating disorder management tool developed to empower patients and equip treatment teams with the complete clinical picture.

“The app connects people that have eating disorders with their treating provider, and takes episodic care and turns it into continuous care,” Mordecai said. “Patients love it, and, importantly, the clinicians really like it. One thing I’ve learned with app work, if the clinicians don’t support it, you’re not going anywhere.”

Mordecai said there are indications that the app is improving outcomes. “We’re seeing things like a decrease in missed appointments, and we’re looking at whether the patient’s weight stabilizes at a healthier level.”

Moving forward, as Kaiser physicians and researchers look to build that ecosystem of apps, they are looking at the layers of data that can be incorporated, Mordecai said, such as patient reported outcomes, EMR data, genetics and even passive monitoring, such as social media data.

“There are fascinating streams of data that can come out of your cell phone, and you can feed this all into a data stream to identify who needs care and for what and move it way upstream. If you can pull this data together effectively, care could be delivered months or years in advance, and the provider then can become an interpreter of this data to help you understand what this data means. I don’t think the human gets pulled out of this,” he said.

The goal is to use data to be predictive rather than reactive, Mordecai said, and to this point, he shared how researchers are using machine learning to build and evaluate prediction models to identify which patients might be at increased risk of suicide and when that risk is reduced or elevated.

As previously reported by Healthcare Informatics Contributing Editor David Raths, several Kaiser hospitals are contributing data to the project on about 350 predictors ranging from socio-demographic characteristics to psychiatric diagnoses and co-occurring substance use disorder. The researchers have found that the machine learning models for suicide risk seem to offer statistically significant improvements over previous methods of identifying suicide risk. In the next few months, Kaiser Permanente in Washington, D.C., is planning to upload the results of the risk scores generated by the algorithm into its EHR at the point of care.

“The real advantage is to get it into clinicians’ hands so they can use it at the point of care,” Mordecai said. He also said he is interested in researching whether other data streams, such as genetics and passive monitoring, can provide key information to identify a patient with a significant suicide risk.

“Suicide risk prevention is one example, and heart disease is another, and it’s not just about knowing who is at highest risk for a given outcome, but also who is likely to respond best to a given treatment. So, you can sit down with the patient and have an EMR system that says ‘This person is similar to 250,000 other people in this stage of illness and the next best step is X.’ That is exciting to me as a provider,” he said.

Through the use of sophisticated data analytics, healthcare organizations can potentially develop powerful predictions to reduce health risk, help patients advert negative health outcomes and help providers pick the best care pathways, Mordecai said, which will help to improve population health and reduce the cost of care.

Can Digital Health Tools Help the Most Vulnerable Patients?

During a separate panel on state models and population health, three senior leaders representing Medicaid programs for New York, California and Massachusetts addressed the promise of digital health innovation and whether the hype around mobile apps can apply to the needs of Medicaid patients, who are often the older, poorer and sicker in the population.

Michael Wilkening, undersecretary, California Health and Human Services Agency, noted the range of digital health apps that are available, and agreed with Mordecai’s assessment: “If it doesn’t change behavior and isn’t engaging, it’s just going to be more technology and you’re not getting any real benefit.”

For the Medicaid population in California, there is promise in the use of telehealth to improve rural Californians’ access to healthcare services as well as health information exchange, Wilkening said. “I think there is some promise there, but I do think there is some overhype as to what you can expect out of technology just from technology. Unless it’s engaging or fundamentally changing something that is broken in the system, then it’s just adding more layers and more costs,” he said.

Daniel Tsai, assistant secretary, MassHealth, and the Medicaid director for the commonwealth of Massachusetts, said, “On the consumer side, with the Medicaid population, we’re thinking about the most complex, highest-cost individuals. The things that we’re worried about are less the kinds of things that an app can help to address. We need to make sure someone has housing and food. I would be excited to see consumer engagement digital health tools that could help on those things, but I don’t know what that would look like.”

Jason Helgerson, Medicaid director, State of New York Department of Health, said the healthcare industry was “lights years behind” other industries in terms of transforming through technology. “If we ultimately want the system to be more cost effective, we have to find ways to harness technology. I would agree, it hasn’t been a panacea to date and I’m not saying it’s going to be a pure panacea, but it has to be part of the solution,” he said.

He noted that after-hours care is still considered innovative in the healthcare industry. “If you compare retail and healthcare, in retail, can you imagine if Amazon shut down its website at 5:01 on Friday, and a message flashed up on its website that said to call 911 to receive services, which is our competitor, and pay three or four times as much or wait for us until 8:01 on Monday morning? That is our healthcare system. At the end of the day, we have to find ways to harness the power of technology to redesign the models of care. Technology is not an end in and of itself, but it is a means to transform the delivery system,” Helgerson said.

Moving forward, there needs to be more innovation in integrating healthcare and social determinants of health and coordinating “whole person” care, and technology can play a role in that, Wilkening said. “It’s really driving changes in those sorts of interactions and the innovations that are associated, both culturally and technologically, with integrating those different aspects together,” he said.