My PCP Should Be a BOT

I like my PCP. She does her best, swimming against the tide of a dysfunctional healthcare system. And let’s be clear: if I had a serious medical condition, I’d want to be treated in the U.S. and nowhere else. 

So I have no time for yet another tear jerking expose of American healthcare! I mean, if that’s how you feel, don’t use it, and next time a loved one needs a complex medical procedure, feel free to head off to Canada, or the UK, or Sweden - wherever you think you’ll fare better; but as you travel, be careful not to bump into ‘medical tourists’ from those countries ‘sneaking’ across our borders to get their care in the U.S.

Not your cup of tea after all? Then shut it! (Full disclosure: I worked 40 years for a company I co-founded to provide health benefits to employees of mid-sized U.S. businesses. So caveat lector.) 

That said, enough is enough! Prior to 2019, my interaction with the healthcare system consisted of an annual physical and a monthly prescription for a beta blocker. In the most recent 5 years, I have been admitted (overnight) to the hospital a half dozen times - incarcerated for a total of 40 days overall. Not fun! However, I was mainly well treated and the care I received was helpful, albeit minimal.

Then I stubbed my left big toe. I’ve been stubbing my toes since the age of 3. Painful but no big deal! This time, however, I damaged the nail and after a week or so it was clear that the nail was coming off, whether I liked it or not.

So I called my podiatrist…who could not see me for a month. No problem. I called my PCP and made an appointment for the next day. Her PA took one look at the toe and announced, “We can’t treat this here. You’ll need to go to Urgent Care (UC).”

Not a problem either! I’m a strong believer in UC; it has served my family well in the past. This time, however, the two closest UC facilities were ‘booked solid’ and ‘not seeing walk-ins’ that day; so much for ‘urgent’. Plus I needed an X-ray and neither had X-ray capability on site.

So my PCP sent me off to the ER…for a stubbed toe! They took me in right away (glad for the revenue perhaps), but 5 hours later, I was still there. Following an X-ray, my ER Doc proudly announced, “Your right toe has a small fracture…which BTW we don’t need to treat.” Ok, but it was the left toe I stubbed! Hmm…

In any event, I was sent on my way with a tetanus booster and an Rx for an antibiotic and a referral back to my Podiatrist who suddenly found he could ‘fit me in’ after all. So, all is well, right?

Not even a little bit! In a country where some folks have virtually no access to healthcare, I unintentionally and unnecessarily consumed thousands of dollars’ worth…for a stubbed toe. Plus, during my week-long ordeal, I noticed that my docs were more interested in ‘ruling out what isn’t’ than they were in ‘treating what is’. They were running through a check list; they were practicing defensive medicine. They weren’t really looking at my toe…or listening to me!

So how can the system be made better?

Begin by transitioning healthcare generalists (PCPs, Pediatricians, Family Doctors) from their current ‘front line’ positions to an ‘oversight’ role. Most current PCP functions could be the responsibility of a new uber-doc, AI Bot, MD, first in its class at Harvard Med. 

Dr. Bot would handle patient in-take, conduct the initial interview, download my electronic health record, build my medical history, order appropriate tests, and offer my PCP a preliminary diagnosis along with any alternatives that ‘we can’t rule out’.

Each diagnosis would come with a full complement of treatment options, including the ‘null option’, and prognoses for each. Only now, at this stage, would a human healthcare professional (PCP) review the data, evaluate Dr. Bot’s conclusions, and make a recommendation to the patient.

This sounds good…but can it work IRL? Turns out, it can and it does. On February 10, 2026, the University of Cambridge (UK) reported the following:

“Researchers…analyzed heart sounds from nearly 1,800 patients using an AI algorithm trained to recognize valve disease, a condition that often goes undiagnosed until it becomes life-threatening.

“Valvular heart disease affects more than half of people over the age of 65, with around one in ten having significant disease. In its early stages, it is often symptom-free. ‘By the time advanced symptoms develop, the risk of death can be as high as 80% within two years if untreated’, said co-author Professor Rick Steeds from University Hospitals Birmingham.

“The AI correctly identified 98% of patients with severe aortic stenosis, the most common form of valve disease requiring surgery, and 94% of those with severe mitral regurgitation. When tested against 14 GPs who listened to the same recordings, the algorithm outperformed every single one and did so consistently. 

“The technology is not intended to replace doctors, but could be a useful screening tool, helping doctors decide which patients should be referred for further investigation and treatment.”

So what does this change accomplish?

• We have ensured that each patient and PCP get to see the full playing field of potential diagnoses and treatment options.

• We have massively boosted the productivity of every PCP, allowing them to spend more quality time with each patient and allowing us to increase their compensation. 

• We have empowered doctors to do more MD-level work.

• We will save money by more often getting the right diagnosis the first time, by reducing the number of ‘false positives’, and by cutting back on the ineffective care wasted treating misdiagnosed conditions. (Note: 20 – 30% of all the care delivered in the US today is not appropriate for the patient’s actual condition.)

• And last but by no means least, we will deliver better patient outcomes…much better!  

As you can see, we are not talking about incremental improvement here. We are talking about revolutionizing the practice of medicine. Results will include a better professional experience, better patient outcomes, and dramatically lower costs. 

But this future depends on our willingness to integrate Dr. Bot into our practices. Once upon a time only white men were doctors; we fixed that! Now we need to add AI docs to our treatment teams…and we need to do so now!

***

Image: "The Doctor," oil on canvas by Luke Fildes, 1891. Located at the Tate Gallery in London.