Alfred Lewis, JD is an author of several healthcare outcomes books, operates the website, “They Said What? Because the Wellness Industry’s Pants Are On Fire,” and is founder and CEO of Quizzify of Waltham, MA.
Tell me about yourself and what you do.
I am CEO and quizmeister-in-chief of Quizzify, which is a an employee health literacy company. As we say, wiser employees make healthier decisions. However, I believe we are having this conversation because of my personal blog, which is called, “They Said What?” in which wellness vendors, diabetes vendors, and related vendors are critically analyzed to in fact show that they usually don’t achieve what they claim to achieve.
You’re offering $3 million to any company that can convince an impartial panel that their program can save employers money. Do you have concerns about having to pay up?
None whatsoever. The entry fee is $300,000, and believe me, it’s worth taking a one in a million shot with this impartial panel of five judges, of which I only get to appoint one and the burden of proof is on me. They don’t have a chance, which explains why nobody has tried to take me up on it.
Is it lack of knowledge or intentional deception that motivates wellness companies to sell services to employers without having sound science behind them?
Confucius put it very well. He said, and in those days it was all gender specific, that, “When a man makes a mistake and it’s pointed out to him and he doesn’t correct it, he is telling a lie.” So at this point, these folks know they are lying. They have made the gamble, and it’s a good gamble, that vastly more people are going to read their ads that are going to read my website. So what they do, and they’ve gotten very good at this in the last couple of years, is simply ignore my postings instead of responding to them so as not to create a news cycle and a whole discussion.
Is the available science good enough that they could do it right if they really wanted to?
I would say that for wellness generally, it is mathematically impossible to save money. There are not enough wellness-sensitive medical events. Even if you were to reduce 100% of them, you could not pay for most wellness programs. I’m not going to say it’s impossible, but it has clinically never even gotten close to that 100%. The typical reduction in risk is 0%, somewhere between minus 2% and plus 2%, while you would need a mathematically impossible 100% to 150% reduction to break even.
Most vendors are counting on the fact that most employers have absolutely no idea how many of their employees go to the hospital every year for diabetes. I could tell you if you like, unless you want to take a guess. Out of 1,000 people under the age of 65, how many go to the hospital with a primary diagnosis of diabetes in the insured population?
I’ll say two.
Actually, that’s very close. It’s more like one. Occasionally I run health and wellness trivia contests at conferences. How does the radiation in the CT scan compare to the radiation in an X-ray? But I also throw in that specific question. If you added all the diabetes events and all heart attacks together in a typical employer population, what would the rate be per thousand? In fact, it would be two, if you put both of those together. The guesses that I get are usually somewhere between 20 per thousand and 200 per thousand.
What about the perception of the incidence of chronic disease in general?
It’s not my take, it’s the world’s take. Because I do this show of hands thing, I do these trivia contests all the time. The employer benefits community thinks it is between about 20 and 200 of these events per 1,000 employees. Which of course makes no sense whatsoever. This is just what they say because they get bombarded with information talking about all the people who have diabetes and all the expensive chronic disease. Let’s take those two things one at a time.
A lot of people do have diabetes. They may not even know it. It’s not going to become an issue for them for many years after they find out. If in fact an employer intervenes, they may possibly be able to control it. But what they’re doing is saving Medicare money down the road because virtually nobody goes to the hospital with diabetes before the age of 65. Yet employers want to start paying for medication for these folks, so it’s a net increase in cost.
And then your other point of chronic disease. I’ve written extensively on this fallacy that 86% of cost is chronic disease. If you read it carefully, you’ll find that they are saying that that 50% of adults have chronic disease. Now if you’re defining chronic disease that broadly, you’re including a whole lot more things besides the things that a wellness vendor can get to. You’re including arthritis. You’re including hypertension. Who doesn’t have hypertension?
If you put all that together and say, “Let’s count every dollar that someone with hypertension spends on healthcare.” So someone with hypertension breaks their leg, you count that. You probably don’t even get to 86%, but most of that is also going to be in the over-65 population. In the under 65-population, the major drivers of costs are birth events and musculoskeletal.
The wellness vendors have done a great job of moving the goalposts. It used to be they would say, “You’re going to get a three-to-one financial return.” Then they started saying, “You’ll get a one-to-one return.” Now they’re saying, “There is really no financial return, but the employees will be healthier.”
If you actually look at the health of the employees … I’m not going to name names, except to say that there are a handful of vendors, generally the ones validated by the Validation Institute, that get more than a trivial improvement in health. There are other vendors — and I don’t mind naming names, Interactive Health and Wellsteps come to mind — where employees actually get worse as a result of these programs.
If that’s the case, won’t those companies eventually get fired for failing to deliver?
Some number of them are getting shown the door, but new employers are coming in. The problem is that the vendors have figured out how to measure outcomes fallaciously in such a way that most employers and most consultants aren’t going to catch them. They compare participants to non-participants, for example. It’s been proven up, down, sideways, backwards, forwards, and eight ways to Sunday that every iota, every dollar of savings in a participant versus a non-participant comparison is due to the mindset of the participants versus the non-participants and not to the program.
How do I know that? There are several data points. Studies have benchmarked those things and found exactly that. But the most dramatic one is a company called HealthFitness Corporation that did a wellness program for a company called Eastman Chemical. They separated the groups into participants and non-participants in Year Zero. But due to a whole bunch of incompetence and delays, they didn’t get the program started until Year Two. By the time they started the programs, the participants had already dramatically outperformed non-participants.
The funny part about that is that my nemesis, the Snidely Whiplash to my Dudley Do-Right or the Lex Luthor to my Superman, was stuck with this, so he moved the goalposts. He said, “Oh, we overlooked that. That was our bad. We weren’t competent enough to realize that the program had actually started in Year Zero, not in Year Two. Therefore, you don’t know whether it’s due to the participants or non-participants.”
That turned out to be a big enough lie. And I don’t mind saying, oh, I’ll say on the record, Ron Goetzel is a liar. He can go ahead and sue me. The difference between him and me is that if he calls me a liar, I’ll have him in court the next day.
They put out a graph that shows suddenly that the program started in Year Zero, not Year Two. The people who actually did the program got upset enough with that. If you go back and look at the website now, they have in fact replaced the lie with the truth, which is that the program started in Year Two after dramatic savings had already been found.
You’ve made the case that the simplest way to measure a workplace wellness program’s success is to ask the people who signed up if they participate regularly and see benefit from it. Do most programs fail even that basic test?
There is a tool put out by the Validation Institute that is the most elegant tool for measuring the cost-effectiveness of programs that I’ve ever seen. We are big supporters of it. You ask employees two questions. How much did you use something? You may not even have to ask them that because you already know. Then, did you find it useful? Then you multiply the number of times somebody used something times the usefulness they found. That gives you an engagement score as your Y axis. On the X axis is the cost of the program. You plot the engagement score against the cost of the program and you can tell in a single graph how cost-effective your programs are as viewed by employee use, employee engagement.
You’ve come down hard on Livongo. What concerns would you have as an employer who is considering buying their their program?
I would have two ethical concerns. One is that what they called a study that they point to is essentially a paid ad. The study was done by their employees and their suppliers’ employees. They don’t say anywhere, “We paid thousands of dollars to have this study published.” If they had disclosed that, that would be acceptable. Marginally acceptable. But to essentially take out an ad in this schlock journal disguised as a study, I have an ethical problem with that.
The other thing I have an ethical problem with is that that journal did do a modicum of peer review. Not remotely as much as I’ve done, but a modicum. And they said, “There is no causality here. It is only correlation. There is a correlation between having a Livongo program and having a reduction in costs.” Livongo put out a press release that said, “This study delivered a reduction in cost,” which is a lot different from a correlation. You cannot ethically take the word “correlation” and turn it into the word “delivered.” Those are my two ethical problems.
I have some arithmetic problems as well. The two things that you should measure if you’re trying to figure out if in fact you have reduced the severity and the incidents of diabetes are, number one, what happened to insulin use? Insulin use has actually been declining because the price has gone up so much, so it shouldn’t be a heavy lift to show a reduction in insulin use. Meaning you’re getting some diabetics off of insulin, which is a cost savings, and it also shows that the type 2 diabetics are improving.
Number two, you say, how many fewer diabetics went to the hospital for diabetes than they did previously? That’s a very standard plausibility test that the Validation Institute uses, that Health Affairs has used. It’s in my book, “Why Nobody Believes The Numbers,” which was a trade bestseller when it came out if anybody wants to look at it. It has never been challenged.
Either Livongo did not know enough to measure the two primary outcomes of a diabetes study — which are, did you reduce the use of insulin and did you reduce the hospitalization rate for diabetes – or they measured them and they did not disclose them. Neither of those gives confidence in Livongo.
The third thing is that their first study said they got a 59% reduction in inpatient, which essentially means that they wiped out every single inpatient admission that did not involve birth events, trauma, cancer, or mental health. Every single one. Their second study made absolutely no reference to inpatient, but said that physician visits and physician expense went down by 26%. So essentially they had two studies, and when they put out the second one, they conveniently forgot about the first, which essentially said the opposite. That’s a red flag.
The other red flag is that every single other wellness vendor in the universe looks at physician visits and physician expense as a good thing. You’re getting people to go to the doctor more. It’s questionable whether that is a good thing, but that’s what everybody looks at. You’re getting people to go to the doctor more, so they’re doing more prevention, et cetera. The idea that you could be titrating all these diabetics’ meds, managing all these diabetics, and somehow have vastly lower physician expense is something they would have to do a great job of explaining to me.
That brings us to the final item, which is that some of what they do appears to be in conflict with other guidelines. This is also in my company Quizzify’s diabetes Q&A, which is reviewed by doctors at Harvard Medical School and carries the Harvard Medical School shield on it as a result. That is, that type 2 diabetics should not obsess with checking their blood sugar. That’s more of a type 1 thing, to check your blood sugar every few hours or every day or something.
It’s quite clear that there is no difference in outcomes between type 2 diabetics who do that and type 2 diabetics who check it vastly less regularly and just have a healthier lifestyle. They don’t have any kind of sentinel events, like a change in their meds or a big change in their weight or some kind of medical event of some type. You just don’t have to check it that often.
But Livongo brags about how many times they get their type 2 diabetics to check their blood sugar. Maybe it’s a coincidence or not, but they are allied with companies that provide medication and other supplies to diabetics. So I would have them explain why they are doing something different from what the literature says.
The manufacturer of Oxycontin pitched their product in referencing a friendly, somewhat obscure research letter that wasn’t peer reviewed. That’s what drug companies do – cite the positive papers in their advertising even if they are scientifically shaky. Is this a healthcare problem beyond just wellness programs, where we aren’t critical enough consumers of literature?
The Oxycontin thing was kind of funny. The doctor was not getting paid to say it and he was actually specifically referring to patients who are already in the ICU. They found something that happened to say what they wanted to say, and like you said, they ran with it.
This one is a little different, because they basically paid a bunch of their employees and they got their suppliers to write this article. Then they paid a journal to publish it. The payments to the journal have never been disclosed to investors. It does say who wrote the article. It does say that the employees and the suppliers wrote the article.
But here’s the thing. Most people, when they see the term “peer reviewed,” that checks the box for them. That says, “Oh, this is legitimate.” But I could give you 15 or 20 peer-reviewed articles in the wellness and the diabetes literature that are essentially incorrect on their face.
Anybody can challenge data. The issue is invalidating data. Can you look at data, and on its face, prove that it’s incorrect? With most wellness data, you can. In fact, I often say in wellness that you don’t have to challenge the data to invalidate it. You merely have to read the data and it will invalidate itself.
Many of those studies are peer reviewed, and many of never should have passed peer review. Oftentimes there are entire journals out there like the Wellness trade journal that have never asked me to peer review anything because they know true peer review would just shoot down everything that they put in it.
Employers talked a lot about coalitions and group purchasing to reduce their healthcare costs, but they haven’t accomplished much. Are wellness programs a half-hearted attempt to rein healthcare costs without addressing provider charges?
Let me take that question and put it into two parts. One is that wellness was very easy to put in place. You could say to your CFO, “Oh, look, we’re doing wellness. This will solve our problems.” Because for wellness, you didn’t have to negotiate with your suppliers or anything like that. You just layered in a new cost item and claimed that it would save money.
A guy by the name of Dave Contorno in our industry, a very capable guy, says the way to save money is to spend less of it, not to add on programs. Like Yogi Berra once said, “We don’t know where we’re going, but we’re making good time.” It was a panacea. There was even a guy — I don’t mind telling you his name, because he said it publicly — by the name of Bruce Sherman, who claimed in a conference that wellness could reduce industrial waste. When you get to that level, you’re just in fantasyland.
The second point that you made is, what should employers do? I would direct you to a book by a guy by the name of Dave Chase. It’s called “CEO’s Guide to Restoring the American Dream: How to deliver world class healthcare to your employees at half the cost.” He points out that, in fact, you can reduce costs by 20 to 40%. It’s been done. It’s not a question of finding solutions — the solutions have been put into place. It’s just a question of putting these proven solutions into place. Things like reference-based pricing and employee education, which is of course what we do. There are new levels of new types of pharmacy benefit managers that don’t have these massively complex contracts with all sorts of rebates that the employers never see, but rather just take wholesale prices and mark them up. All sorts of things have been done. All you have to do is do them and you will see.
When I work with David Contorno or Dave Chase, I use a little formula with them. Which is, X plus Y equals 20%. X is the reduction in cost and Y is the increase in employee satisfaction with the healthcare program measure, however they want to measure it. Those two figures will add up to a 20% improvement. So if you really want to ratchet your costs, you can do that with no improvement in employee satisfaction. Or at the other extreme, if you feel that a really good program is great for attracting employees, you can keep the cost the same but then basically create low co-pays and low monthly contributions and get your employees much more satisfied with the program.