Published July 12, 2012

How Health Care Can Cost So Much, Despite Technology


By Alex Daley and Doug Hornig

Last week, amongst the hundreds of emails that fly around our offices each day, came a small and seemingly inconsequential story about an aspiring politician named Chris Collins, who is gunning for a seat in Congress representing his Buffalo suburb of Batavia. According to the clip, Collins basically told a local news outlet that cancer has been cured.

The press and bloggers latched on to a single quote from Collins that was featured on a tiny local news site, The Batavian. He said: "People now don't die from prostate cancer, breast cancer and some of the other things." The predictable jokes and rants ensued.

In all fairness to Collins, that quote was taken well out of context. He went on to make a point about the rapid increase in quality of care that we've seen in the past few decades, thanks to technology, and wondered whether some of the cost increase in health care might be attributable to that success. He continued on to say, "The fact of the matter is, our healthcare today is so much better, we're living so much longer, because of innovations in drug development, surgical procedures, stents, implantable cardiac defibrillators, neural stimulators – they didn't exist 10 years ago. The increase in cost is not because doctors are making a lot more money. It's what you can get for healthcare, extending your life and curing diseases."

Whatever you might think of his political leanings – he believes Obama isn't spending enough on Medicare and wants to tariff China – he does have a bit of a point on the advances in medicine. There are many things we can do today, thanks to technology, that weren't even possible just a few years ago. But that got me wondering: just how much of a point does he have? Is the runup in healthcare costs really a result of how many new treatment options we have? Or have costs increased for other reasons?

There is absolutely no doubt that medical care has improved in the last few years. That change may not universally reach everyone who might benefit from it nor be evenly distributed across the conditions that ail us – certainly many suffer today just as much as they always have. Still, on what we scientists might refer to as the "upper bound," what's possible to do with medicine today is multiple orders of magnitude greater than it was in, say 1972... just 40 years ago.

That date is about when the computer revolution started to take hold and turn upward. Sure, they'd been invented decades earlier, but it wasn't until the '70s that computers started to become common in commerce and industry. And the story since then has followed one clear path forward: increasing capabilities at an ever-plummeting cost. The result is that I can now do more on my smartphone – which cost a mere $200 – then I ever could have thought of doing on a $2,000 computer just a decade ago. So why hasn't the same happened in health care?

In many ways, it has. However, much of the change in spending is a result of the outcome and not of the inputs. Consider, for a second, just how long we now live…

Are We Really Living That Much Longer?

It is easy to dismiss the days of people's lives spanning a mere three decades as prehistoric. But it wasn't really that long ago. Consider that according to data compiled by the World Health Organization, the average global lifespan as recently as the year 1900 was just that, 30 years. In the richest few countries on Earth at the time, the number rarely crossed 50.

However, it was just about that time that public health came into its own, with major efforts from both the private and public sectors. In 1913, the Rockefeller Foundation was "looking for diseases that might be controlled or perhaps even eradicated in the space of a few years or a couple of decades … Technical approaches also tended to yield immediately quantifiable results that justified equivalent expenditure of funds."

The result of this concerted public-health push included nearly eradicating smallpox, leprosy, and other debilitating or deadly diseases. It also included vaccines against the six killer diseases of childhood: tetanus, polio, measles, diphtheria, tuberculosis, and whooping cough.

A simple graph illustrates the dramatic change.

(Click on image to enlarge)

In the US, the average lifespan is now 78.2 years, according to the World Bank. In many countries in the world it is well over 80. But the story isn't so simple. Like all averages, it's affected mainly by the extremes. And in the early part of the 1900s, the data point that weighed most heavily on average lifespans was child mortality. Not the least of the reasons why families were much larger then is that parents routinely expected some of their children to die.

(Click on image to enlarge)

But the flip side, as can be seen in the graph, is that for anyone lucky enough to survive childhood at the turn of the last century, life expectancy was not that much lower than it is today. For all of our advances in medicine, we only live about 20 to 30% longer. Not only is the increase quite small – relative, say, to the explosion in computing power over the same period of time – the amount of money we spend adding another year or two to the average lifespan is on the rise. In other words, each additional day we live costs more than the one before it. And, as with compound interest, the effect is amplified over time.

In essence, we have picked much of the low-hanging fruit from the tree of life.

So if, absent high child mortality, we are not living all that much longer today than we once were, where does all the money we spend watering that tree actually go?

We can get a glimpse of the answer in the following graph. Intuitively, one would think that there should be a relationship between the economic wellbeing of a country and the life expectancy of its citizens. And, as you would imagine, there is a strong correlation between wealth and health.

(Click on image to enlarge)

The important takeaway from this graph is the flattening of the curve along the top. What it means is that many countries with lower GDPs (and, consequently, less to spend on health care) can attain life expectancies in the 65-75 range. Pushing them higher than that, as the richer countries do, clearly requires much greater resources. By implication, that means spending more to do it.

The cost of battling the diseases of adulthood rises dramatically with age.

How much so? Well, per capita lifetime healthcare expenditure in the US is $316,600, a third higher for females ($361,200) than males ($268,700). But two-fifths of this difference owes entirely to women's longer life expectancy. For everyone, nearly one-third of lifetime expenditures is incurred during middle age, and nearly half during the senior years. And for survivors to age 85, more than one-third of their lifetime expenditures will accrue in just the years they have left.

Technology, contrary to most popular opinion, often helps to bring these costs down. To take a simple example, an MRI may seem expensive on the surface, but it accomplishes things that previously required a much heavier investment in time and diverse professional expertise. Or consider how a company like NxStage Medical has revolutionized the delivery of renal care. The company's home-based units save a ton of money compared with the traditional, thrice-weekly visits to a special dialysis clinic.

But the primary problem is that, overall, medical costs continue to rise faster than improved technology can serve as a countervailing force. There are three easily identifiable reasons for this:

  • Diminishing marginal returns
  • Rising costs of non-technology inputs
  • Increased quality of life

The Law of Decelerating Returns

Technology in most arenas is a field of rapidly increasing marginal return on investment, i.e., accelerating change. Things don't just get continually better or cheaper; they tend to get better or cheaper at a faster rate over time.

There is a simple concept in finance, hard sciences, and any sufficiently quantitative field – anywhere that numbers dictate behavior – called "economies of scale," and it mainly refers to the idea of changing returns over time. We refer to these as "marginal returns."

Imagine for a moment that you are a manufacturer. Once you've paid off the cost of your factory or equipment – your "fixed" costs – you maybe make a widget to sell for $10, with a cost per unit of $5. If you make and sell a thousand units, you make $5,000 profit. That is the marginal return. But now imagine that if you make 100,000 units, your costs per unit drop to $4 – you have more negotiating power over your raw materials suppliers, you can run your staff with less slack, etc. And if you can make 200,000 you save/make another dollar. That means you have increasing or accelerating marginal returns. We're used to that in technology.

Every year Intel is able to lower the costs of the processing power it sells on the market. Making the chips becomes easier with time and scale, although there is fierce competition from companies like ARM Holdings and Taiwan Semiconductor to take some share. Fortunately for Intel, a chip is a commodity product, it's the same for nearly all consumers, and the market is global. (Not to mention the small need for highly trained service practitioners, the lack of spoilage, and other nice benefits of dealing in circuit board technology.)

But medicine is not quite the same – at least not in the most important instances.

Sure, an aspirin pill is relatively cheap (inflation adjusted, the price of about $7 per 100 pills has held steady for decades) because it reaches a massive audience, it can be shipped globally, and no one pill is different than another. However, there is no universal medical cure. While a computer can solve virtually any math problem, begetting everything from spreadsheets to video games, Lipitor is applicable only to a single condition and only optimal for a subset of the human race.

Nor can one doctor treat a wide range of diseases. They're forced to specialize. Moreover, they usually only see patients from within a certain physical radius; they must undergo never-ending education and certification; they practice in expensive buildings; and they require complex equipment used only by a handful of fellow specialists. There are few places to find so-called economies of scale.

Treatment Difficulties

The simple fact is that, in our self-centered zeal to live to the age of 20, we have made a trade-off. We've left behind the diseases of youth – diseases that mostly strike once, resulting either in death or fading chances of a long life – but they've been replaced by a host of new chronic diseases. Diseases of age. Diseases of environment. And diseases of design.

These conditions – such as diabetes, ischemic heart disease, and yes Mr. Collins, cancer – are all much more complicated than their predecessors. First, none is caused by a single, easily identifiable agent. There's no virus to isolate and eradicate.  There's no pathogen sample to convert to a vaccine. These are diseases born of the complexity of our bodies, and the built-in difficulty of keeping the most complicated and mobile of all computers running for many decades on tuna sandwiches and OJ.

 While deaths from most infectious diseases – other than AIDS and a few other key conditions – are expected to drop rapidly, deaths from strokes and other ailments of age and environment, heart disease, and cancer continue to rise (although some great strides are being made in certain types of cancer, such as basal cell carcinoma and lymphoma).

(Click on image to enlarge)

These conditions cost considerably more to treat than the traditional infectious disease does. More labor is involved. More time. And available drug treatments rarely cure in a few doses, if ever.

Chronic conditions breed chronic costs. Of course they do, that's their nature. Keeping someone with lung cancer alive for twice as long as would have been the case 30 years ago is a great feat, but it comes at considerable additional cost in terms of the time devoted by the many healthcare professionals involved. (And that means troubling questions like this must be asked: If every patient can live twice as long, but it takes twice as many people-hours net to care for them, has there been a net gain for society?)

The Driving Force

Our medical progress has been won through a major increase in net costs per person. In 1987, US per capita spending on health care was $2,051. That's $3,873 in 2009 dollars. But in 2009, actual spending amounted to $7,960 per capita. Why?

Some of that is attributable, pure and simple, to rising costs that have outpaced inflation. In 1986, the average pharmacist made $31,600, or $66,260 in 2012 dollars. Today, their real average salary is $115,181, nearly double. On the other hand, it's not universal. Radiologists, for example, have seen their salaries drop from an inflation-adjusted $425,000+ to $386,000 in the same period.

Also, costs for surgeries and diagnostics are not a clear-cut contributor. Data are hard to compile as costs vary greatly, even in one region (California recently saw charges for appendectomies in the range of $1,500 to $180,000; in Dallas, getting an MRI at one center can be more than 50% more expensive than another across town). Most indications seem to point to lower, not higher, real costs over time for most common conditions. Average hospital stays post appendectomy have fallen from 4.8 to just 2.3 days in the past 25 years, for instance, thanks largely to insurance requirements, as well as better sutures, pain medicines, and surgical equipment.

As hard as procedural costs are to compare, the outcomes are much more clear cut. In cancer, which Mr. Collins refers to specifically, the improvement has been significant in some cases and less dramatic in others.

For those diagnosed with cancer in 1975-'77, the five-year survival rate was 49.1% (and only 41.9% for males). For those diagnosed between 2001 and 2007, five-year survival increased to 67.4% for both sexes and jumped to 68.1% for men. Even if you're diagnosed when over age 65, you have a 58.4% probability of living another five years.

Prognoses, however, vary widely with disease specifics. If you contract pancreatic cancer, for instance, your prospects are the grimmest. It's likely to be terminal very quickly. Among the most recently diagnosed cohort, a meager 5.6% survived for five years. That's more than double the rate from 30 years ago, but small comfort. Liver cancer sufferers' five-year survival rate has more than quadrupled, but only from 3.4 to 15%. Lung cancer is also still a near-certain killer. In the 2001-'07 group, a meager 16.3% survived for five years, only a slight tick up from the 12.3% rate of 30 years ago. Brain cancer is quite lethal, as well, with only 34.8% surviving for five years today, more than 50% better than the 22.4% rate of 30 years ago, but not great.

On the other side of the ledger, breast cancer victims are doing very well. 90% survive for at least five years if diagnosed after 2001, vs. 75% in 1975-'77. And prostate-cancer treatments have been the most spectacularly successful. Five-year survival is fully 99.9% of those diagnosed in the past ten years, vs. only 68.3% in 1975-'77.

Longer survival rates are, of course, impossible to document in recently diagnosed patients, since we're not there yet. But to give you some idea, here are the 20-year survival rates for the above cancers, taken from the NCI's 1973-'98 database: pancreas, 2.7%; liver, 7.6%; lung, 6.5%; brain, 26.1%; breast, 65%; and prostate, 81.1.%.

These are big steps forward, no question. They enhance not only the length but the quality of life, as well.

However, with each rising year of average age, we increase our medical expenses rapidly. When we eradicated the big childhood killers, we solved most of the easy problems. As a result we all live longer. And we all live to face the much more complicated and much more expensive to treat diseases of age. At that point, it isn't lifestyle changes that are keeping us alive – It's machines and doctors and medicines doing a lot of the heavy lifting in order to grant us those precious extra days. It's the Hippocratic Oath writ large: physician, thou shalt do whatever it takes to prolong life, no matter the price.

All of that costs money, and lots of it.

Mr. Collins is right, in a manner of speaking. We are not dying of the most dreaded ailments as quickly as we once were. But that's not due to much in the way of real advances in curing the major chronic illnesses of our time – heart disease, diabetes, cancer, and AIDS. The truth is that we've primarily extended the amount of time we can live with them.

Mexico's health minister, Dr. Julio Frenk, noted the irony here when he said, "In health, we are always victims of our own successes." We are living longer... and we're costing a lot more in the process.


Bits & Bytes

Novel Nanotherapeutic Busts Heart-Attack- and Stroke-Causing Clots (Fierce Biotech Research)

The obstruction of critical blood vessels due to thrombosis or embolism (basically blood clots and other blockage) is one of the leading causes of death worldwide. But it may not have to be. Scientists from Harvard's Wyss Institute have discovered how to use the body's natural clot-producing mechanism to deliver lifesaving nanodrugs directly to obstructed blood vessels, which dissolve blood clots before they can cause serious damage. So far, the technique has only been used in mice, and is still years away from being tested in humans. But I think we can all agree that it's a step in the right direction.

Terahertz Laser Scanners – Something New to Spy on You (ExtremeTech)

Apparently, Big Brother will soon have a new high-tech toy to spy on you with. Reports say that in the next year or two the Department of Homeland Security will be able to scan you from 50 meters away with a new laser scanner that is able to, according to its inventors, "penetrate clothing and many other organic materials and offers spectroscopic information, especially for materials that impact safety such as explosives and pharmacological substances." The tech is not new, it's just faster and more convenient than before. The machine fires a laser to generate molecular-level data, and the attached computer provides a readout of the information in real time – from trace amounts of drugs or gunpowder on your person or clothing to what you had for breakfast. The plan is to deploy the scanners at airports and border crossings across the US. But if the device is successful and as convenient as its inventors say, who would want to bet against it being abused and misused?

Sign Language-Translating Gloves (Gizmag)

More than 40 million people suffer from hearing and speech impairments. To communicate, they must use sign language. But most of the rest of us don't know how to sign. A team at the Microsoft Imagine Cup created the Enable Talk project to solve the language barrier between sign-language users and the rest of the world. A smartphone and a pair of the special sensory gloves are all that is needed. The gloves capture the hand movements and transmit the movement pattern – the sign itself – to the mobile device. Then the application matches the incoming pattern with stored signs and plays the sound for that sign.

An Algorithm to Pick Startup Winners (Technology Review)

Correlation Ventures is taking automation to the extreme. Rather than relying on the traditional due diligence of human analysts, this "new breed venture capital firm" relies on predictive analytics software to drive investment decisions.