Seth's Blog : 16 minutes

16 minutes

Some facts and assertions about healthcare (particularly in the US) and then an outline of a change agent that could improve health, perhaps dramatically.

We spend an enormous amount of time and money on healthcare.
The typical visit to a healthcare clinic lasts an hour and a half.
The typical time with a doctor is sixteen minutes.
Improving the timing and accuracy of a diagnosis has enormous leverage in changing outcomes and reducing costs.
There is a significant amount of anxiety and dissatisfaction with the system. Patients are upset at the cost, society is paying more than ever, and chronic diseases take an ongoing toll.
Women and non-white men are often underserved, misdiagnosed and dismissed. The same is true for humans who don’t fit standard types of symptoms and backgrounds.
The healthcare system makes treatments, not health. It’s treatments that are measured, marketed and sold.
The rigorously maintained status roles assigned to each player in the system can make communication more difficult. Patients hesitate to share enough information, and communication between and among the professionals in the system is intermittent and often poorly facilitated.
Doctors and other health professionals are hard-working and care a great deal, but often face burnout because the demands of the system are obstacles on the path to what they actually signed up to do.

And so, a system that’s organized around treatments and status, that misallocates time and effort, causing stress for practitioners and patients. Historical bias in training leaves more than half of the population underserved and unseen, and, as a result, stress is high, many people don’t get the right treatment or hesitate to get any treatment at all, and costs continue to rise.

Systems change is difficult, because persistent systems are good at sticking around. They create cultural barriers that make their practices appear normal, and there are functional barriers as well.

When a change agent (often an external technology or event) arrives, the system must respond, often leading to change. All around us, we see systems changing, and often, that change agent is the smart phone. 91% of adults in the US have a smartphone, and it’s even higher among people under 65.

The ubiquity of the connected supercomputer in our pockets has overhauled the taxi industry, the hotel business, restaurants and most of all, pop culture. But it hasn’t transformed the healthcare system. Add AI to the mix, and it’s possible that change is about to happen.

Imagine an app that does the following:

The user narrates their symptoms, over time, to a thoughtful AI that provides prompts, encouragement and reassurance.
The app keeps a date-stamped log of what’s bothering the patient and refines the data over time by prompting with useful questions.
The user goes to the doctor and the app offers several things:
- It gives the doctor a high-level overview of what the patient is dealing with.
- It records the conversation with the doctor for later analysis and referral.
- It offers the doctor prompts on questions to consider or studies to refer to.
- It asks for any test results or diagnostics to be shared with the patient.
After the appointment, the app compares the doctor’s interactions with all the doctors dealing with all the patients in a similar situation. This gives the patient insight around standards, and more agency in finding a doctor who might be more effective.
It maintains a record of healthcare encounters over time, and correlates them with patient health.
It follows up with the patient to make sure prescriptions are filled and actually taken.

Some patients aren’t comfortable with this sort of interaction, they won’t use it. Some doctors might be skeptical, or concerned about the shift in their status, and they might not work with patients who use it. These behaviors are typical of status quo systems, and both often fade in the face of better results combined with user satisfaction.

Like any tool that shifts an industry of this scale, the business models often take care of themselves.

The biggest information shift here is the more accurate collection and correlation of symptoms and treatments. The secondary (but ultimately longer-term) shift is finding threads of common interest and comparing doctors in their responses to symptoms. (And the side effect of giving patients agency and the solace that comes from insight can’t be ignored). Because both of these data shifts will lead to better patient outcomes (usually at much lower cost, with less trauma) the healthcare professionals who signed up for precisely this outcome will also thrive.

It’s not a panacea. But shifting information flows, improving peace of mind and the quality and timing of diagnosis are problems we can work to solve.

Systems change is never without real dislocations and regret, but this one might be worth building.