It’s not news that Americans spend a lot on healthcare. Nothing has seemed to work at changing this. Preventive care and technology, such has AI, has not put a dent in this spending. Same for the push for healthier lifestyles. There is much blame to go around–from drug companies, to insurance companies, or people being overbilled. Pessimistically, however, I propose that high healthcare spending is an inevitable consequence of people acting rationally, which makes the problem much more intractable and less amenable to technology or policy-based approaches.
2 years ago, a Reddit user describes going to the ER due to severe stomach pains. She felt dumb because nothing was found, wasting her time and the doctor’s time. A doctor who replied reassured her that she was justified in going to the ER despite being a false alarm and not appendicitis as originally feared:
This exchange explains why healthcare spending cannot be reformed with policy or technology. The patient spent no money, yet the doctor had to be compensated for his time to try to diagnose her. The hospital had to also accommodate her. If emergency room treatment is assumed to be a public good, then raising prices in an attempt to dissuade non-emergency visits won’t work. So the usual economic models of inelasticity vs elasticity or supply vs. demand go out the window.
Hospitals try to recoup some of these costs, but unlike consumer goods or elective procedures (e.g. cosmetic surgery), there is a great disconnect between the delivery of said service and having to pay for it. With healthcare, it tends to be the opposite of how consumers typically pay for things: the service is provided before payment, and the price is negotiated afterwards. This creates an incentive from the perspective of the consumer to demand more of the service.
The loss of potentially decades of life from undiagnosed cancer or other illness makes the ER visit worthwhile despite the wait. The assumption is long wait times will dissuade non-emergency visits, as those patients self-select to leave before seeing a doctor, but we’re talking lives at stake, not a Disneyland ride.
Moreover, diagnosing illnesses is rather impervious to non-medical technology and is labor intensive. How can patients or even doctors know if there is really an emergency (except for obvious things like gunshot wounds) or not? They cannot. Despite decades of training, the only way a serious illness can be ruled out in someone presenting with symptoms is to visually inspect the body, such as scans or possibly observation. This means tens of billions of dollars spent annually on running tests on people who have symptoms but are healthy, or tests or scans on asymptomatic people as part of routine checkups or screening.
As those who have consulted Dr. Google can attest, benign conditions can have the same symptoms as even terminal conditions. It does not help that abdominal pain and nausea are both symptoms of something as harmless as indigestion, or as deadly as pancreatic cancer. A headache with sensitivity to light are commonly cited symptoms of brain cancer, but also can occur together for no reason. No amount of AI-based diagnosis or flowcharts will rule out cancer, which can only be done by visualizing the body and confirming it is not there.
Doctors are acting rationally in not turning away these patients, because there is always the possibility something is wrong that can only be picked up with tests. It may actually be cheaper in terms of less wasted manhours to run tests at the first visit than have the patient return repeatedly if symptoms get worse.
The potential downside is huge at the individual level for both parties. For the doctor: possibly being sued or loss of employment. From the perspective of the patient: being turned away despite being severely ill. Running tests however has a much smaller downside felt on everyone, patient and staff alike. But it adds up to the tune of billions of dollars annually in extra spending.