Using AI and a Little Common Sense to Restore Trust in Healthcare
In June, after 30 years as a private practicing General Internist, as a Chief of an academic division of General Internal Medicine and as a Chief Population Health Officer among other roles, I left the provider side of healthcare.
Why? Because I wanted to have a bigger impact, an impact at scale, and I wanted to work with smart and optimistic people. Also, I was tired of feeling like Sisyphus, pushing a rock up the hill every day only to have it roll back down just as I got to the top. And even if I managed to get a rock over the top, I found more rocks waiting for me when I got back down to the bottom.
I also spent a decade trying to tackle the Triple Aim, then they made it the Quadruple Aim…and a moon shot became an intergalactic journey. But the triple aim was never going to happen. It was brilliantly aspirational. It provided focus, an agenda and abundant discussion, but operationally it was quite problematic. Its strength was its seeming simplicity. It distilled a complex problem into three (or four) easily identifiable tranches of work. However, the achievement of that work would require an alignment of interests that had a track record of resisting change. In retrospect, the simplicity of the message obscured the complexity of the task required to change the behavior of many.
The status quo was and is too entrenched, there are just too many interested parties, including ourselves…and if you doubt me, just look at your retirement portfolio and tell me that you don’t see a health insurance company, pharma company, or for-profit healthcare company.
I believe that if we are going to make a difference with innovation, we need to rethink the agenda. Great thoughts and grand plans are wonderful if you want to become famous, but if you want to achieve, you need to think differently. The currency of fame is likes and tweets, they are ephemeral and have no integrity. The currency of achievement, however, is execution. Only in execution do we create opportunities to learn, and only in execution do we get the ability to iterate and improve.
Recently, I found a video of Denzel Washington on YouTube. He was speaking to a group of young actors, and he said something that really resonated with me: “dreams without goals are just dreams.”
In my view, there is an analogy to this for innovators in health care: ideation without execution, is just ideation. Innovation requires robust ideation, but that alone is NOT sufficient. Health care’s troubles are so large that we need innovation that drives execution. And our patients, our families, our friends and ourselves need us to execute.
So, moving forward, I would suggest that as innovators we reframe our notion of innovation, from the demonstration of aspirational brilliance to the PURSUIT of what my friend and CEO of PatientWisdom, Greg Makoul, has called radical common sense.
By radical common sense I mean use the tools at hand to solve existing problems, create business models that are good for everyone, and work to make healthcare more just and equitable. Essentially, do the things that would make your mother proud of you!
Now let me demonstrate this concept of radical common sense as it applies to health care.
In talking to hundreds of patients and surveying hundreds of others, my colleagues and I at Syllable have come to recognize that there are at least 4 archetypes that describe a doctor:patient or health system:patient relationship. These archetypes are based on the length and emotional depth of the relationships (See Figure 1).
We describe long, deep relationships as “intimate.” These relationships are built on implicit trust. Examples of intimate relationships include the relationship with a spouse or significant other, a clergy-person and, I would argue, the historic notion of a primary care relationship.
Long, shallow relationships are what I would term business or service relationships. We have many of these types of relationships. They are characterized by warmth and respect or “trust but verify” and are exemplified by the relationships we have with our dental hygienist who we see twice a year, by the people who do our hair at the salon or barbershop and even our work relationships. In healthcare, I believe that this represents the majority of health system-patient relationships.
Fewer of us have short, deep relationships because these are specialized in nature. They are called “swift trust” relationships because they usually develop in situations where individuals who are little known or unknown to one another are thrust together for a specific, usually urgent or high-risk task. These relationships are based on explicit trust, reputation or when a standard operating procedure is present. Basically, I trust you to do your job because I know you have been well trained to do it, even though I don’t know you and haven’t worked with you before. These relationships are seen in airline crews, disaster relief agency workers and trauma teams.
Short, shallow relationships are ubiquitous. They include the preferred way we now do our banking and achieve multiple weekly tasks. The successful ones are fast and accurate. In healthcare, I would argue that there are opportunities to improve the transactional experience for things like procedures in low risk populations, helping patients choose a physician and simple administrative tasks.
Historically, intimate relationships were defined by time, not necessarily because intimacy takes time, but because there were no other elements that needed to be accomplished. There was not a cascade of transactions that needed to be successfully executed. You saw your doctor and it was either good or not. As long as that interaction was positive, the relationship grew. And in talking to patients there is a yearn to return to intimacy. Patients are craving recognition as an individual rather than a health insurance policy holder. But today there are just so many transactions that occur, from insurance verification, to refill requests, to call back requests, to making appointments, to getting directions. Each of these transactions, if done well, reinforces a positive relationship, any ONE done poorly can lead to a dissatisfied patient and a negative review. The same analogy also applies to health systems and employees. There are dozens of transactions that occur during a day; make them easy and you have happy employees, make them hard and you have unhappy, less productive, and less ambassadorial colleagues.
Intimacy used to be the foundational relationship between patients and their caregivers, but that is no longer the case! Based on what we have seen from our research and heard from employees at health systems, we believe that transactions are now the foundation upon which deeper and more satisfying relationships in healthcare depend.
In 2020, with so much of our doctor-patient relationship tied to successful transactions, we can’t restore intimacy without first solving the transactional conundrum. Refilling a prescription, answering a call, scheduling a test, delivery of hot food to a hospitalized patient…it is the ease, accuracy and speed at achieving these tasks upon which we are evaluated …we don’t get credit for the care we give because no one goes to a doctor or to the hospital expecting to come out feeling worse or sicker than when they went in.
As healthcare innovators as well as current and future patients, we need to focus on creating a stable transactional foundation for healthcare, without which intimate, longitudinal trust is impossible. Focusing AI on pedestrian, common problems can result in improved experiences for patients, providers and employees. Once we solve these problems we can progress to more grand applications. I know this doesn’t sound sexy, but healthcare needs to socialize AI and AI needs to learn more about healthcare. AI, machine learning and NLP offer us tools to create a digital assistant that can not only create accurate, fast, reliable solutions for healthcare transactions, but they can also confer a bit of intimacy by expressing some knowledge of the user.
Consequently, in 2020, we at Syllable are focused on the transaction.
Using this approach, we will deploy AI solutions that, among other things will:
Make it easy to find a doctor
Make it easy for a healthcare provider to perform a simple administrative task without talking to a human, like changing their password, and
Make it easy for a patient with an undifferentiated condition to find the right provider and site of care.
In doing so we are going to help our partners develop a stronger foundation on which to build an intimate brand, and a roadmap to deliver better care with an experience that inspires and delights patients and colleagues, at a lower cost.
