Q&A With Andy Kessler on the Disruption of Doctors

Andy Kessler is a former tech guy, former Wall Street analyst, former hedge fund manager. He has written three books on the investment world, chock full of hilarious and revealing stories, insightful analysis and wisecracks.

In pursuit of the next big investment score, he wrote The End of Medicine: How Silicon Valley (and Naked Mice) Will Reboot Your Doctor, published this week. It  has Kessler's trademark bawdy stories and acid observations, but it is an outsider's examination and diagnosis of an industry that looks poised for change.  We traded emails with Andy to learn a bit more.

 

Q: Andy, what got you interested in looking at the medical business and what got you to thinking that it is ripe for disruption?

I've spent over 20 years tracking the technology business for Wall Street, as an analyst and then as an investor.

The only thing I really learned was to find the silicon. Once you find a market that is attacked by silicon, you only have to wait for big markets to come into being from nothing. Silicon gets cheaper every year by 30%, it halves in price every two years. If you find something that works today, but is too expensive, then wait a bit and the fireworks start.

I was bored looking at traditional markets, computing, telecom, wireless, even music and video. I drifted around for a while, sniffing at other interesting things when I learned that a friend was diagnosed with cancer, by accident, as he had banged his head skiing and an X-Ray and CT scan showed a tumor on top of his neck. A brother-in-law had a heart attack. I wondered if silicon could be found in medicine to be able to detect disease much earlier. I was astounded by what I found.

Ripe for disruption is an understatement. $2 trillion a year is spent on healthcare in the U.S. and it is quickly on its way to $3 trillion. Doctors drive medicine and are a large component of costs which rise by double digits each year. If silicon can make its way into diagnosis, perhaps costs could stop rising and at the same time provide better care.

Q: Are doctors to become the bank tellers of their industry, to be replaced by better, smarter, automated processes or online procedures that customers would prefer to deal with?

Think about what happened to telephone operators. Switches came along and performed their function better and cheaper. Eventually, switching happened so quickly, telephone calls could become cheap and ubiquitous. A machine embedded the intelligence of the operator.

Same for bank tellers. A slightly more complicated task, ATMs embed the intelligence of the teller into silicon and software and lower the cost of transactions.

Same for auto mechanics. They fix the problem, but silicon diagnoses what is wrong with autos today, embedding the intelligence of the mechanic into the system. At $60-75 per hour charged by service stations for labor, the payback for an automated diagnosis system is quick.

The same thing happened to stock traders…So why not doctors. I went for a physical and my doctor took my blood pressure, looked into my ears and the took out a rubber hammer and

banged my knee. $440 for nothing. Was I going to have a heart-attack? He couldn't tell me. Could I have a tumor the size of a golf ball on top of my neck? No way to tell looking in my ears. He was flying blind.

But new technology exists to do powerful diagnosis, and embed the intelligence of front line doctors into silicon and software. 256 slice CT scans that turn into high resolution 3D images that can be flown through looking for atherosclerosis to avoid heart attacks. Biomarkers and molecular imaging, all silicon based, will be able to detect unique proteins from cancer cells five years early. This will change what is spent on chronic care by huge amounts.

The trick is to get down the learning curve and make these tests cheap and routine.

Q: Ok let's cut out doctors. I am all for that. But they do a lot of thinking, don't they? Just like equity analysts. Can we really automate doctors? While we are at it, can we automate stock picking and eliminate analysts? Apart from in commodities, analysts don't scale well either.

First, not all doctors. Just the front-line doctors.

My friend had a tumor the size of a cocktail olive found on the top of his neck, by accident. If he hadn't banged his head skiing, he would probably be dead right now.

The idea is to look inside each of us. Front line doctors, internists, family doctors, can't do that. They can only guess.

Nature is the screen for heart disease and cancer. When you have a heart-attack, the system declares that you have arteriosclerosis. Couldn't you have told me ahead of time and avoid the midnight ambulance ride?

Cancer is worse. You piss red or your shoulder hurts or you’re throwing up dinner and after exhaustive tests, they find a tumor and declare you have six months to live and even chemo and wonder drugs are a long shot.

Using technology, we can scan for all sorts of proteins in our blood and do a much better job of finding cancer early. Five years early. Doctors can't do that. There are no symptoms. They would never know. But cheap technology can know. Smaller, cheaper, faster, better. In this case it is cheaper and better than doctors.

The security analyst comparison may not work, but the trader comparison does. You can trade in milliseconds huge blocks of stock anonymously, while using a trader requires phone calls and paper trails and confirms and moving the market, etc.

Specialists on the NYSE and doctors may be in the same boat: Progress bypasses them.

Q: Maybe this question is too simple, but from a customer's point of view, what is the broken part of the medical business and what might be the easiest way to fix it? Ie Convert the macro observations into a business idea.

Doctors are human and don't scale. Get them out of the equation and you have a business that can get smaller, cheaper, faster, just like everything else silicon valley touches.

Consumers can do their own tests, and then only see specialists when they need them. (No pun on the NYSE, but same concept!)

Q: Cancer is a big killer today, yet doctors and patients will tell you that despite all the advances in chemotherapy, it is still a crapshoot. Did you come across any technologies that might change that?

Mostly, I spent time looking at early detection, the ability to find cancer early enough so that treatment is simple and effective, directed radiation, surgical removal, heating, freezing, melting via directed ultrasound, etc. Detected early, cancer tumors are microscopic, a million cells vs. later stage tumors the size of a cocktail olive or golf ball with a billion cells. These microscopic tumors can be treated without affecting much surrounding cells, almost impossible with later stage larger tumors, hence chemo. So the earlier, the better.

Having said that, I did see imaging techniques involving probes tagged with radiation, that find these early microscopic tumors and  then light them up for PET scans. There is research going on in increasing the radiation dose tagged to these probes, meaning that  they not only find cancer tumors, but directly apply radiation to  them. Interesting, but very early.

Q: Ok let's invest in some medical companies or startups. What are your picks for either companies or technologies to consider?

There really isn't a business model for early detection, not yet.  There is more of a negative for expensive pharmaceuticals and  treatment centers, the day that heart disease, stroke and cancer begin to subside in numbers.

However, sometime in the next several years, as the silicon and software and algorithms developed for early detection begin to get cheap enough, these business models will  emerge, and you will see venture capitalists step up to fund interesting companies and IPOs of the hot new early detection players. 3-D imaging, computer aided detection, biomarkers, molecular imaging probes, nanotech scan devices for antibody chips, these are all areas that we will see companies emerge.

**Other Information**

The Disruption Group lists the results from disruption, including new revenue streams and sustainable, high return on equity.  The website also has a CEO Guide to the Benefits of Disruption (pdf)

Comments

[John Haughom, MD]

I just finished Mr. Kessler's book and found it fascinating. I have now recommended it to dozens of friends in healthcare. While his criticisms of our industry and profession are uncomfortable, they are also largely accurate.

A comment about electronic medical records. Based on over ten years of experience implementing them, I can tell you Mr. Kessler is right. They are a piece of the puzzle but EMRs will not by themselves revolutionize healthcare. However, the digital data in them will be a rich source of knowledge to further the exact type of research and care outlined in the book. I have felt for a number of years that the solution to healthcare's problems lie in earlier diagnosis and treatment of major diseases. The premise of this book is right on the mark.

[Bern Grush]

I'd want to think more about Kessler, he's right -- but there are a million risks. Even while reading this simple Q+A, I started seeing the designs for online dignostic systems (which are no new idea). I, like CSS, would also want to read his book first. David Lilienfeld's comments are good, and I agree. But they are orthogonal to what I think Kessler is saying. To Lilienfield's point the loss of context and history due to our abysmal record keeping is legion. I have have had the same tests on my fluids within a week of each other for no discernible reason except that the costs of obtaining the results from last week was more trouble (or less lucrative?) to the practicioner of this week. I retain the same family doctor (although I'd liek to move) simply because he has massive records that no other physician would ever dig through, if I moved them. I actually rely on a combination of his memory and mine.

My before-reading-the-book opinion: neither of these things will happen soon. There is too much cultural habit and established value tied up in doing both elements the current way.

[Andy Kessler]

Lot's of folks think fixing medical records will solve all of
healthcare's problems. I'm a bit skeptical. I think it can help, but not be game changing. Google didn't fix the Dewey Decimal system and save libraries. They changed the game. Similarly, in healthcare, the trick is to keep people from getting sick in the first place. It may be genomics, which would involve medical records, but it is more likely environment, which requires early detection.

[David Lilienfeld]

I offer my comments from the perspective of a physician-pharmacoepidemiologist who has practiced. The savings to be incurred by record linkage and practice evaluation is considerable. Unfortunately, HIPAA restricts the ability to do so absent the specific consent of those whose records will be linked. Since an analysis of 100K+ records is needed in most analyses, HIPAA creates quite an impediment. Prior to HIPAA, such analyses disclosed the effects of NSAIDS in causing upper GI bleeding, and by reducing the use of those medications, reduced the demand for health care considerably. There are lots of similar examples, and many of these directly relate to the development of practice guidelines to reduce wastage in the health care delivery system. That wastage reduction would provide significant cost savings--savings needed if our society is going to address the twin challenges of Alzheimer's disease and other disorders of aging, and the developing epidemic of Type II diabetes among teens and young adults.

[css]

HIPAA is an excuse - by that logic GLB would have shut down the banking industry a long time ago. Didn't happen there, and shouldn't here.

What health care has lacked is a market - there's been no need (at least until now) for the system to be either market- or cost-competitive - and no incentive for investors to push these ideas forward.

If the payoff is there for investors, HIPAA will be nothing but a speed bump. But I believe we need a (free) market to make that a reality. Nationalized health care could very easily push things in the OPPOSITE direction. Since I haven't read the book yet can't speak to whether Andy addresses that point, but if nothing else these ideas should contribute to the debate.

[David Lilienfeld]

Your interview with Andy Kessler is certainly interesting, but you both miss the point: the issue in the U.S. health care system hasn't been about the technology to move information for some time. Maybe 15-20 years ago, but not today. Today, the problem is one of access to information. One of the challenges presented by HIPAA is the inability to link records and thereby derive needed information from the raw data of health records. Kessler ignored this key issue. One can probably justify the investment in new IT in health care based on cost savings per se, but the amounts involved pale in comparison with the likely increases in overall health expenditures we will see as a society. The only way to address those increases is changing how we use the data we have available to work smarter. HIPAA (and the general concerns in society about privacy) suggest that such changes are not in the offing--at least not near term.

[Andy Kessler]

I did Q&A with Michael Urlocker on his website, On Disruption, about The End of Medicine...