Why We Should Let the Actuaries Fix Healthcare
I had an interesting conversation with a naturopathic doctor recently on her podcast. She had read that I was an actuary and fell into the tiny percentage of people who had some idea of what an actuary is and does. In her case it was thanks to an excellent article in The Epoch Times by Jeffrey Tucker (also published by the Brownstone Institute) imploring health insurers to hire more actuaries and bring rigor and reality to the industry that is sorely lacking.
The average person has no idea what an actuary does and the unique skillset that helps present a more accurate picture of what leads to good health and longevity. Here I unveil a bit of the mystery and in the process explain why a lack of basic math proficiency can be detrimental to your health.
An Actuary Explained
An actuary is a profession that involves many years of study into risk measurement using the mathematical disciplines of statistics and probability. Risk affects capital outlays so investment markets and economics are also a big part of the skillset. You’ll find a lot of actuaries in other types of insurance such as auto, home-owners or product liability. Also in life and disability insurance where the risks of death (mortality) and accident or illness (morbidity) are incorporated to ensure there is enough money to pay out benefits each year to policyholders.
Good Economics Through Incentivizing Behavior
Before getting into the realities of healthcare, it’s useful to look at another type of insurance as a comparison where actuaries are heavily involved in analyzing the risks and improving the economics: car insurance.
There are several factors that are used as a proxy for the true risk of a car accident and therefore the pricing of a car insurance policy such as:
Factor
- Where you live
- How expensive your car is
- How big the engine is
- How many miles you drive
- Any speeding tickets
Underlying Risk
- How much car crime is in the area
- Replacement cost
- How fast you can drive
- How much opportunity for an accident
- How recklessly you drive
The surest way to impact behavior is to hit you in the wallet so if you get a speeding ticket you will be required to pay a higher premium the following year. This fact turns out to be a great motivator to obey the speed limits. No one complains this setup is discriminatory – it just makes good economic sense and is actuarially sound.
A few years ago there was an outcry in the US when employers started asking smokers to contribute more towards their health insurance than non-smokers. It is now generally accepted that smoking tobacco is bad for your health and longevity. There is little backlash on life insurance companies using smoking status to price the policy. You are simply more likely to die within the policy period if you smoke than if you do not smoke.
The smoking debate took years to be accepted but is now generally recognized as a major factor in the likelihood of future illness. Employers will typically require a smoking surcharge on health plan premiums for those who continue to smoke as long as a quit-smoking program is also offered at no charge to the employee.
Smoking as a risk-factor is so well accepted these days, we have taken to aligning two other key health risks proven to cause similar amounts of deaths worldwide, to the severity of smoking:
- Inactivity such as sitting at your desk all day 1
- Eating Ultra-Processed Food (UPF) 2
Adjusting the amount an individual has to pay based on their risk profile makes economic sense in a world with finite resources. Yet neither of these risks are typically used in health insurance underwriting nor are they the first line of defense in healthcare plan design.
UPF has been proven to be every bit as addictive as cigarettes and getting started with an exercise program is painfully difficult. While employers and insurers often do provide programs to employees that address these risks, usually categorized as add-on wellness initiatives, they have not gone far enough to impose penalties and hit employees in their wallets to incentivize better health behaviors.
Another major risk factor for future ill health is the consumption of pharmaceutical products. Even when taken as prescribed, pharmaceuticals are the third leading cause of death in the US and Europe.
But don’t pharmaceuticals make us well? Actually no, in the majority of cases, they may address a symptom but the human body is infinitely more complex than the healthcare industry would have us believe. Part of the problem is how statistics are used in healthcare to drive treatment decisions.
Statistical Shortcomings
While other types of insurance calculate the likelihood of a single claim resulting from a car accident, a broken water-pipe, a failed product, or a death, healthcare is a lot more complicated as the event leading up to the claim is not always clear or finite.
You do not wake up one day with cancer or heart disease, diabetes or an autoimmune condition. These events that lead to healthcare claims have been years in the making and constitute a convergence of almost infinite risk factors. The statistical studies that make the headlines are usually testing a narrow range of factors but draw wide conclusions.
For a start, a clinical trial is usually very restrictive in the scope of people it will enroll. Individuals of certain ages or with any co-morbidities are excluded arguably to limit the effect being tested. But then the results are quoted as applying to whole populations.
Another concern with clinical trials is when the focus is solely on the product’s effect on a specific biomarker (like LDL-cholesterol) or symptom (like pain) without any measure of health or longevity being considered.
A third concern is the manipulation of the resulting statistics themselves. Take for example a clinical trial that tests the efficacy and safety of a medical intervention; let’s say an injectable shot that is meant to reduce the odds of you getting sick from a virus. Let’s say there are 20,000 people who get the shot and 20,000 people who do not. Both groups are monitored for a period of time, say three months. In the no-shot group 200 people got sick with the virus. In the shot-group only 10 people got sick. You are likely to read a headline that the shot is 95% effective. 10 compared to 200 is calculated as 10/200 = 0.05. 1-0.05 = 0.95. Or you might hear you are 20 times more likely to get sick without the shot. Either of these statements are an incorrect assessment of the true risk.
In both groups a very small percentage of the whole got sick so the comparison is actually 19,800 remained virus-free compared to 19,990 which is less than 1% difference (19800/19990), statistically insignificant.
It is also important to dig into other aspects of the clinical trial like adverse outcomes and all-cause mortality. If the ultimate aim of an intervention is to improve health and longevity then all-cause mortality is a critical statistic. In the above example (based on a real high-profile case) you were unlikely to hear of the higher number of heart attacks or deaths overall in the shot-group.
Beyond these example problems with single studies, the biggest concern, and where actuaries can particularly lend their expertise, is that real life is far more complicated than a single line of enquiry can analyze. No definitive conclusions can be drawn from studies like these carefully designed to focus on a single system or biomarker.
Basic Statistics Can Only Take You So Far
Real life has an almost infinite number of permutations than even the most well-designed research can study. So you need to look at a much larger body of evidence to get closer to how the human body functions and what will improve or detract from good health.
If we look at nutrition experiments as a case in point, it is impossible to completely standardize exactly what a large group of people eat for a long period. Yet the medical and pharmaceutical industries will tell us only a double-blind placebo-controlled trial can determine if nutrition works better than one of their drugs at addressing a particular condition. Apart from the impracticality of prescribing exact food intake for months or years, the possibility for a double-blind control group falls away when you realize how difficult it would be to disguise the food being studied so you didn’t know if you were eating it or not.
Nutrition studies therefore typically rely on food-recall surveys instead, most often at the beginning and end of the study or even just once in the middle. If you were asked to list everything you ate last week, let alone in the last month or two, how would you do? What about if you were just asked to estimate how many times you ate x?
Even if the answer to the question was fairly accurate, what about the questions that are not asked? What brand of x? (Quality heavily impacts micro and macro-nutrient levels and additional substances such as pesticides, fertilizers, added sugar and preservatives). What else did you eat at the same sitting? (which impacts absorption of nutrients).
The problem with a non-actuarial interpretation of a research study is that it is taken at face value and in isolation. In fact the infinite complexity of the human body means any single study could easily lead to a wrong conclusion. Much more context needs to be brought into the analysis.
Two of the most damaging examples are:
What's the Difference Between an Actuary and a Data Scientist?
(Sounds like the start of a joke!) A helpful article in The Actuary magazine provides some insight but basically it comes down to pure statistics versus applied to and interpreted in the real world (the actuary focusing on the latter).
The now debunked assertion that saturated fat and elevated LDL-cholesterol lead to heart disease.
The body of evidence is clear that lower LDL-cholesterol in the body (mainly through statin drug use) likely leads to dementia and other neurological problems whereas heart disease is much more aligned with sugar intake and other risks leading to excess blood clotting.
The debunked claim that depression is caused by a serotonin imbalance in the brain and that SSRI drugs are the answer.
In fact SSRI’s have been shown to block the neurotransmitter acetylcholine which can lead to dementia, they are connected with bone density loss, an increased risk for suicide, and a 620 percent increase in breast cancer. Depression stems from a variety of different root causes.
Indeed, it is the narrow analysis of pharmaceutical studies, each in isolation, which has missed the glaring reality that in general pharmaceuticals lead to worse outcomes in the long term.
Actuarially Sound Healthcare
So if pharmaceuticals are part of the problem, it questions the whole paradigm of healthcare provision. By taking a broad, multi-disciplinary approach based on comprehensive data and independent analysis, actuaries would conclude that an insurance policy covering only the conventional “standard of care” doesn’t make economic sense.
Root cause medicine, known best as functional medicine, is rarely covered by insurance today but makes much more sense actuarially. Here’s a simple comparison:
Conventional Standard of Care
Diagnosed with an autoimmune condition
Put on a high-cost drug: $30,000 a year
Side-effects lead to more drugs prescribed: $$$
Dis-regulated immune system leads to more ailments and possibly cancer a few years later
Functional Medicine Approach
Diagnosed with an autoimmune condition
Doctor takes a detailed history identifying the root cause(s)
Nutritional coach: $100 per month
High quality supplements: $100 per month
Exercise regimen
Health turned around
Autoimmune condition in remission
Health improves year over year
While the summary may seem overly simplified, it is actually the reality in practice when taking a holistic approach to health and embracing the innate healing mechanisms of the human body. The fact that this approach makes so much more economic sense seems lost on the health insurance industry.
Healthcare Must Face Reality
Any macro-economic assessment of the industry will conclude we don’t have enough healthcare services to reach every person that needs them, so we have to find a way to allocate the scarce resource. Aside from rising the price beyond many people’s means, the most obvious way is to adjust demand through incentivizing pro-health behaviors. This is just plain reality and common sense.
The “pro-health behaviors” need to apply to both the consumer and the provider. Doctors who spend 10 minutes with a patient listening to symptoms in order to prescribe a pharmaceutical is inefficient care at best. (I might call it something else entirely akin to medical malpractice in some cases). If the realities of health and human longevity were acknowledged, then the first line of defense for chronic conditions is always lifestyle management and functional medicine.
Many doctors will need to get re-trained as the efficacy of nutrition and natural therapies is not taught in medical school. Fortunately the Institute for Functional Medicine (IFM) provides excellent education courses with an introduction and sample topics available for free.
Actuaries may need to take IFM training too to understand the intricacies of the human body and interactions of its systems. Combining physiological knowledge with the actuary’s already broad skillset should provide sufficient context and help reach more accurate conclusions on what will lead to better health, longevity, and lower healthcare claims.
Reducing demand for healthcare sends shivers through the pharmaceutical and hospital markets with their power over the media contributing to the dilemma. But employers paying for health plans, and insurance companies funding policies, each need to take a stand and make sure reality and fiscal responsibility win out. Hire more actuaries and push for independent analysis in line with the rigorous approach applied to other industries.
1 Effect of physical inactivity on major non-communicable diseases worldwide (Lancet article) and Physical Inactivity May Cause as Many Deaths as Smoking (Medical News Today)
2 Association between consumption of ultra-processed foods and all cause mortality (Study) and Ultra-processed food consumption, cancer risk and cancer mortality (Study) and Scientist argues for tobacco-style warnings on ultra-processed foods (article)
