Who's Afraid of an HMO?

1850 words, 8 minute read

Smug and self-satisfied, those of us in the Healthcare game were once fond of asking, “When was it that medicine first began to do more good than harm for its patients?” The consensus: c. 1900. 

On what planet? Certainly, 1900 was a major turning point in Intellectual History. It was as if a starter’s gun had been fired. Everything from Relativity and Quantum Mechanics to Cubism and Jazz quickly followed. But not affordable access to quality healthcare.

So if not 1900, when? The correct answer is… wait for it… drum roll please… 2035, if we’re lucky… and then only if we follow the advice offered here! Did I mention ‘smug and self-satisfied’?

***

Next to National Health Insurance (UK: NHS), Health Maintenance Organizations (US: HMO) were the worst healthcare policy idea ever implemented on a wide scale. Intended to facilitate access and improve quality, they did the exact opposite. 

HMOs dumbed down quality and erected formidable barriers to care. They made us learn a whole new, and may I say diabolical, vocabulary: PCP, gatekeeper, referral, out-of-network, etc. They pioneered the 12 minute office visit and the 40 patient work day. They transitioned care from MDs to Nurse Practitioners… before there were enough trained NPs to deliver it.  

The results were so disastrous that they played a major role in catalyzing the (Newt Gingrich) Revolution of 1994, which remains for most of us the defining political event (US) of our adult lifetimes. Four years later, a polemical anti-HMO movie, As Good As It Gets, won Best Actor & Best Actress at the Academy Awards. And so, the game was up: HMO RIP!

Too bad, because in concept HMOs had more potential to improve healthcare quality and access than any competing policy initiative since LBJ’s Great Society!

Full Disclosure: From 1985 – 2020 I was part of a team that built a moderately successful health benefits business, based in part on our fierce opposition to HMOs.

Two Questions: (1) If HMOs were such a great idea, where did they go wrong? (2) If HMOs were such a disaster, why would anyone want to bring them back?

Truth to tell, HMOs went off the rails early on…the moment they stopped being about facilitating quality healthcare and began focusing on reducing the cost of health benefits. They forgot, if they ever really knew, healthcare’s Prime Directive:

“The cheapest healthcare is not always the best but the best healthcare is always the cheapest™.”

The best healthcare is always the cheapest? Really? Is the best automobile the cheapest? The best wine? The best house? Sadly, no! So what makes healthcare different?

Most goods and services generate their own demand. If we make better wine, more people will drink it and the price will go up. Same with houses and cars. But not healthcare. Quality healthcare, properly delivered, reduces the need for future healthcare…and therefore reduces cost long term. How so?

Streamlined access to quality care means that progressive abnormalities can be diagnosed earlier…and interrupted. The ‘cheapest healthcare’ is the care we don’t have to deliver down the road because we diagnosed a potential problem (e.g. diabetes, cancer) early and acted quickly and competently to eradicate the underlying cause.

Conversely, the most expensive healthcare is the care we didn’t deliver, either because we failed to detect the symptoms, or misdiagnosed them, or mistreated them, thereby allowing the patient’s condition to deteriorate (arithmetically) while the cost of treating the worsened condition rose (exponentially).

Ah, the magic of compound interest: miraculous in your 401(k), diabolical in your health benefit plan!

***

I came of age during the great credit card explosion of the 1970s. If you were 18 years old and not in jail, banks you never heard of sent you plastic with ‘astronomical’ credit limits and ‘easy’ repayment terms. No one paid attention to the interest (18% plus) charged on unpaid balances. 

Every time I got a new card, or a credit limit increase on an existing card, I would immediately take my closest friends to dinner at the most expensive restaurant that would let our disheveled selves pass the maître d’. Little did I know that I was managing my personal finances in the same way my country was financing healthcare.

Like the superannuated hippies that we are, we are doing what feels good in the short term (saving money, containing costs), giving barely a thought to the long term debt we’re piling up in the process.

***

So why resurrect the discredited HMO model now? Because those pioneers got some things very right: 24/7 walk-in access, low copays, zero paperwork, a single health record, and a cloud of specialists delivering coordinated patient care through a front line cohort of primary care docs. Of course, we already know that this approach won’t work…unless it is accompanied by a deep commitment to healthcare’s Prime Directive (above) and to these Seven Pillars of Healthcare Wisdom:

1. Maximize the use of AI Bots in the patient in-take process. Create a single, thorough, baseline medical history, organized in a relational format and shared by all physicians (and Bots) in the new mega-practice. 

2. Based on that history, empower AI to review the patient’s regimen of Rx drugs, non-prescription supplements, and diet to identify possible adverse interactions and to recommend safer and potentially more effective combinations.

3. Conduct tests as indicated and offer timely diagnoses developed and reviewed by a team of specialists that includes an AI Bot.

4. Present the patient with a robust set of treatment options, including the null option, along with the potential benefits and known risks of each. AI will play a critical role here by ensuring that no viable intervention is ever overlooked.

5. Undertake patient-selected interventions as quickly as possible, using AI to deliver elements of care whenever medically advisable. 

6. Initiate a Bot led follow-up process to track a patient’s progress and to assess compliance with the prescribed after-care regimen.

7. Recommend AI Identified lifestyle modifications, including dietary supplements and complementary (alternative) healthcare, that may reduce the patient’s need for much costlier acute healthcare services down the road. 

In short, we are proposing a healthcare model based on the traditional HMO concept, now enhanced by a new Patient First ethic and by a commitment to the maximum responsible integration of AI at every step in the patient care process. 

This is not AI-only healthcare, God forbid, but it is also not just AI-assisted. Perhaps call it AI First! Membership in such a system should be viewed as a perq, akin to enrollment in a concierge practice, not as a penalty necessitated by budget cuts. 

‘Good health and long life’ as an employment related benefit? Better than foosball and an espresso machine! “Who’d a thunk it?” (Hairspray) Someday soon, access to this model of care will become a factor in an applicant’s employment decision… and once that happens… 

***

What makes me so sure that AI First healthcare will deliver major improvements in quality of care and similar reductions in cost? Well, to practice medicine in the United States, physicians need to pass the US Medical Licensing Examination (USMLE), a rigorous and standardized assessment of clinical knowledge and decision making capability. 

Passing the USMLE requires the applicant to score 70% or better on each of the test’s three parts. AI Bots invariably achieve nearly perfect scores. It’s your loved one! Would you rather she be treated by an MD who crammed 10  years ago just to get a 70%…or by an AI Bot that scored over 90% and could do so again at any hour of any day?

Still not convinced? Ok, what if you could have the benefit of Dr. Tom’s famous medical intuition and exemplary bedside manner and AI’s near perfect factual retention and medical reasoning. And please, don’t make me be tedious: Don’t make me point out that AI is always 100% current on medical advances, doesn’t drink or take drugs, and never gets in a fight with a significant other (except in 2001 - A Space Odyssey).

“Yes, but how does adding the cost of AI to Dr. Tom’s already substantial salary save money?” Let me count the ways:

1. AI integration will cost a bit upfront but the cost per use after that is nominal. But AI will allow Dr. Tom to consult on twice as many cases, doubling his revenue stream; and he’ll be more than happy with a 20% pay bump.

2. Up to 30% of all money spent on healthcare in the US is spent delivering care that is not appropriate for the patient’s actual condition. AI can reduce this % to single digits overnight.

So what? Well, when a set of symptoms is misdiagnosed, three things happen: (a) money is wasted delivering services that are of no value to the patient, (b) inappropriate treatments may trigger their own adverse side effects, and (c) the patient’s actual condition, now untreated, will progress, exponentially increasing the cost of future care. 

Jeez Louise! You make one mistake! And pay for it three times? Lightbulb: Why not invest in technology that will let you get it right the first time? Hmm.

Need hard evidence? Each week, the New England Journal of Medicine (NEJM) – one of the world’s leading medical journals – publishes a record of certain complex cases treated  at Massachusetts General Hospital. Recently, Microsoft created a series of interactive diagnostic challenges based on 304 recent NEJM case studies.

AI correctly diagnosed 85% of these complex cases. A team of 21 board certified specialists, each practicing for at least 5 years, correctly diagnosed just 20%. Over this admittedly rarefied subset of cases, AI enhanced quality of care by 400% and therefore reduced cost by 75%.

In another study (2/10/2026), 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.) ‘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. 

“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.”

***

Coach, I’m not asking you to throw away your playbook. Just put me in the game. Let me show you what I can do. Personally, I’m sick of always losing. How about you?

***

 Van Gogh’s The Courtyard of the Hospital at Arles offers a rare, empathetic glimpse into the environment of 19th‑century healthcare, portraying the hospital not as a cold institution but as a place where nature and recovery coexist. By filling the courtyard with blooming flowers, bright pathways, and warm color, he transforms a medical setting into a space that feels restorative, suggesting the importance of surroundings in emotional and physical healing. The painting becomes a quiet argument for humane healthcare, showing how even in moments of vulnerability, thoughtful environments can offer comfort and dignity.