The less good (bad)
There are a lot more older people needing new medical treatments, resulting in even longer lives and inevitably lots of other costly health problems.
In my own family, my grandfather was diagnosed with prostate cancer and told to return to the farm and prepare to die. My father succumbed to prostate cancer with a modest amount of palliative care. Should I face the same diagnosis, I will be opting for expensive Da Vinci robotic surgery.
Then we have doctors drowning under ballooning patient volumes. A 2023 study calculated that a GP with 2500 patients would need to work 27 hours a day to deliver all the acute, chronic, and preventive care recommended by “The Guidelines”. For those with 3000 patients (common in New Zealand), it’s 32 hours a day.
Meanwhile, many guidelines are about as current as fax machines and ironically health remains a big user of fax.
Agreeing guidelines can take years, while medical knowledge was reportedly doubling every 73 days in 2020, up from seven years in the 1980s and a leisurely 50 years in the 1950s.
This tsunami of new knowledge means healthcare professionals just can’t keep up.
Doctors are also just exhausted.
Forty-five per cent of all US doctors are burned out, 25% reporting they are depressed. In the UK, nearly half of all doctors report struggling to provide sufficient patient care on at least one occasion each week. In New Zealand over two-thirds of healthcare workers report emotional exhaustion.
And the problem is likely to only get worse.
The World Health Organisation predicts the global health workforce will reach 53.9 million by 2030. But this falls far short of the projected 80 million health workers needed, leaving an enormous gap of 26 million.
Simply put, healthcare is collapsing under the weight of its own success – here and everywhere else in the developed world.
The ugly
Then we come to the ugly.
Medical errors kill millions around the globe. Each year more than a quarter of a million die in America alone.
Every year diagnostic errors affect at least 1 in 20 patients in the US, causing permanent disability or death. In the UK, diagnostic error is estimated to occur in around 5% of primary care consultations.
In the European Union, 22 of 30 million patients with rare diseases don’t yet have a diagnosis, and eight million wait an average of 10 years to receive it.
So very, very ugly!
The medicine (cure)
If the healthcare sector were a patient, it would be in ICU in New Zealand and around the world. Healthcare costs have ballooned from an OECD average of ~5–6% of GDP in the 1970s to ~10.2% by 2030, with the US expected to be 20.3% by 2033 and continue to climb.
Many pundits suggest Band-Aid solutions, like more GPs, more IT, another hospital or two, etc – all of which are like putting fingers in a dam. Essentially fiddling with the deck chairs while the ship slowly sinks. Simply put, we need current health systems to better treat a lot more patients while making far fewer mistakes and tragic errors and health professionals need help to do so.
Something like a Star Trek Tricorder that turned Bones McCoy from a bit of a dope into an amazing doctor.
Fortunately, help is at hand. AI is proving equivalent or better at doing most steps of a patient’s journey. For example, interpreting symptoms, presenting treatment options, interpreting decades of history, reading X-rays, generating documentation, orders, and referrals, and much more.
MedQA tests demonstrated this as human doctors probably average around 76%. That’s at the bottom of the class when compared with the major AI LLMs and these AI scores continue to get better.
Future healthcare looks like patients logging symptoms, nurses adding vitals and observations, and then AI working its magic to deliver a definitive diagnosis and treatment(s) for a GP to approve.
So five minutes of GP time instead of 15, thereby trebling doctor productivity with all scripts, orders and documentation then being generated automatically.
However, what’s really fantastic is that an empowered nurse plus AI will get it right far more often than a doctor alone. They will often be able to treat many people without the patient ever needing to see a doctor.
More accurate diagnosis and treatments will eliminate further doctor and ED visits and hopefully drastically reduce misdiagnosis deaths. And every patient will receive the best treatment(s), as recommended by the latest research.
We’re seeing isolated cases of AI in action in healthcare. But today’s healthcare AI is fragmented. There are lots of small, smart tools solving little problems.
But the next generation will weave them all together across the entire patient journey, integrating data from labs, genomics, wearables, environment, and family histories to provide insights no human could muster alone.
The question is: will the medical profession embrace this AI utopia?
Maybe not?! The medical community has historically scoffed at, rejected, or ignored major innovations such as stethoscopes, surgical anaesthesia, handwashing, blood transfusion and many others.
How ministers of health and other leaders of healthcare economies will overcome this resistance could be challenging and is beyond my pay grade.
But if we want to get healthcare off the front pages and out of the political football arena, embracing AI is our best and only realistic option. It is the only path out of the healthcare morass all countries find themselves in.
Catch up on the debates that dominated the week by signing up to our Opinion newsletter – a weekly round-up of our best commentary.