Isaac Asimov once remarked that a sufficiently advanced technology was indistinguishable from magic.
Were he alive today, Mr. Asimov might also remark that both advanced technology and magic got nothing on Pharos Innovations, whose website reports a world-record 79% reduction in admissions for congestive heart failure (CHF) patient monitoring.
Pharos achieved this Nobel Prize-worthy result in CHF monitoring without actually using CHF monitoring devices, but rather just the telephone and that favorite tool of the frail elderly, the Internet. Most magical was the time this admission reduction took: 31 days.
On the graph below, you can see that the baseline ended December 31, 2007, while the full impact started February 1, 2008.
That means Pharos was magically able to find all these members’ contact information, write to them to announce the program, schedule the phone calls to the members to convince them to join the program, collect their information, conduct those phone calls, explain the system to the members, get them set up on the system, collect the information, get members to visit their doctors, and adjust lifestyles and medications…all during January.
Thanks to that lightning speed, there was literally a 90% decline between the December admissions rate and the February admissions rate, as this chart demonstrates. Overall, this chart is a dramatic rebuttal to the conventional wisdom, which would state that:
it takes a long time to make even the most minor improvements in a population through telephonic and Internet disease management, if indeed improvements are possible at all; and
a trendline that is “unchanged” does not decline 25% like Pharos “unchanged” matched cohort trendline above.
In college Al was assigned a roommate who was like the bad seed from the Richie Rich comics, a kid who, among other things, would have a snifter of cognac before bed. Once Al told this guy he was decadent. “Decadent, Al?” he countered. “Let me tell you about decadent. I spent last summer at a summer camp –everyone was there, Caroline Kennedy, everyone – where we played tennis on the Riviera and then went skiing in the Alps.”
Al agreed he had a point. “Wow, Lance, you’re right. That was decadent.”
“Al,” he replied, “I haven’t even gotten to the decadent part yet.”
Likewise, we haven’t even gotten to the real magic yet. Watch costs, validators, sponsors and articles disappear.
This Is Where the Magic Happens I: The Disappearing Costs
If admissions represent half of total costs, and other costs remain unchanged (even though reducing admissions by half would ordinarily require a significant investment in physician, pharmaceutical, therapy and other preventive expenses), a 79% reduction in admissions should yield nearly a 40% reduction in total costs.
Not so with Pharos. Pharos achieved an 85% reduction in total costs. Thus, as impressive as that 79% admissions reduction was, other costs must have declined by 91% to achieve overall 85% savings. A 91% reduction in other costs means patients could discontinue most of their meds and insulin, cancel their doctor visits, and stop getting most of their therapy…as long as they used the Pharos interactive voice response to monitor their CHF – for a minimum, according to Pharos, of 15 days.
So, in the Pharos CHF scheme, 91% of other costs just went away—no indication of why or how.
This Is Where the Magic Happens II: The Disappearing Validators and Plan Sponsors
One might ask, “Did anybody validate this?” There are three answers to that question. First, we always thought that being validated was like being pregnant – your results are validated or they aren’t. Not so with Pharos. According to a series of press releases, Pharos’s results are not just validated but strongly validated. Second, there is no mention of who the “strong validator” is. One would think that the strongest validation in population health history might be accompanied by the name of the validator, but whatever name that was, has simply disappeared.
Speaking of disappearing, there is no longer even any mention of who the “regional health plan” is for which Pharos achieved these remarkable, strongly validated results. The case study used to mention “Wellpoint,” but Wellpoint unsurprisingly made Pharos take their name off once they figured out what was going on. (Wellpoint, the country’s second largest health insurer, may not appreciate being categorized by Pharos as a “regional health plan,” like the time the Wall Street Journal was famously rejected from the Pulitzer competition because it was a “trade publication.”)
This Is Where the Magic Happens III: The Disappearing Peer Review
Let’s take stock of where we are now. Like magic, the strong validator has disappeared, the name of the “regional health plan” disappeared, and costs have disappeared. And you know what else has disappeared? The New England Journal of Medicine. NEJM published a study showing that the Pharos program not only missed the 79% admissions/91% cost reduction targets, but in fact “did not improve outcomes” at all. Not surprisingly, as they say in Argentina, that article has also been disappeared from any of their materials.
The Magic Continues
Magicians are not supposed to do this, but Pharos did recently share their secrets with someone: they just announced a contract with Humana, and it’s unlikely that a health plan of Humana’s caliber would make a deal just based on magic.
So we close by asking Pharos to share their secret sauce with the rest of us as well, in the comment field below. We are especially interested in how they and they alone can almost wipe out admissions and other costs and get more “strongly validated” than any other vendor…and how the New England Journal of Medicine could have completely missed the boat with their decidedly non-magical conclusion.
Al Lewis is the author of Why Nobody Believes the Numbers, co-author of Cracking Health Costs: How to Cut Your Company’s Health Costs and Provide Employees Better Care, and president of the Disease Management Purchasing Consortium
Vik Khanna is a St. Louis-based independent health consultant with extensive experience in managed care and wellness. He is Wellness Editor-At-Large for THCB and author of the forthcoming Your Personal Affordable Care Act: Making Yourself Scarce in the Dysfunctional US Healthcare System. Vik and Al are co-authors of Surviving Workplace Wellness, THCB’s first e-book.
Yes, Pharos did some stuff with ICCC where they did do that matching and pre-post and showed very high savings. However, I had access to all the Iowa data for all of Medicaid. It appeared that — measured validly — ICCC did achieve noticeable event reduction. However, it was nowhere near what Pharos was claiming.
The problem is that even if your intervention achieves something, if you measure it invalidly the news story ends up being how invalidly you measured it, rather than what the pony-in-the-pile actually was.
Yes, thank you! So this is/like the Pharos and Iowa Chronic Care Consortium study, because sounds like identical pre/post design design, matching, etc. https://thehealthcareblog.com/blog/2013/07/30/stop-the-presses-a-disease-management-program-worked/?
Those two are excellent questions that Randy Williams should answer, but he is not going to because it would mean acknowledging that this posting exists.
My best guesses: they applied some kind of algorithm, like having two or more CHF admissions. That gave them their study group baseline. Then they applied the same algorithm to the control group one year later. The obvious fallacy (obvious to everyone but them) is that the first group would have come way down in costs and admissions anyway in the next year, due to regression to the mean. So their matching is meaningless.
And just to make sure that there is no possibility that this study could be confused with valid, in the study period, they only counted people who self-selected into the program (though the matched control was passive so it counted everyone).
So the study was doubly meaningless. Welcome to the world of population health, Melissa…
Randy, did I miss anything here?
Hi Dr. Lewis, Some questions for you:
I have no training in evaluation, but based on what I learned in my health care outcomes class in my nursing graduate program, but here goes:
1. I agree with Ms. Riddell most clinicians will tell you that HF admissions almost always go up in December with holidays and eating high sodium foods, etc. So the drop is big, but it happens.
2. Can uou share how the groups were matched? Do you have admissions data for both groups in 2007? Otherwise this would be a case of regression to the mean.
3. Since self-selection could be a big error, did you control by randomly assigning patients to the intervention and control groups?
Yet another flaw: comparing December’s admissions to February’s. Perhaps CHF admissions go down every February and the effect was amplified by selecting lower-risk participants than the matched cohort group.
Thanks, folks (and Vik!) Thing is, these vendors just make this too easy. Nonetheless, I have to admire their self-control in not attempting to defend themselves but rather just hoping –usually correctly — none of their customers or prospects will notice, and that attempting a defense will simply prolong the news cycle.
However, the intelligence required to arrive at that conclusion also implies that they are smart enough not to have arrived at their outcomes results by mistake.
Great article Al – love the sarcastically incredulous tone on the journey to the punch line(s)!
Thanks Al & Vic for making me laugh out loud. Even without your exquisite expose, I must say I never really understood Pharos’ value prop. Ask yourselves, would this intervention change your behavior???
BTW, when you are done skewering the DM, sorry, the population health field, I hope you will turn your sharp eyes and wit to what is going on in the world of health apps – most are badly in need of some “strong validation.”
Are you sure Pharos actually exists? Maybe somebody made up the story and put in the impossible numbers to let everybody know it was a big joke. Did their press releases come out on April 1?
Arthur C. Clarke said it — not Asimov.
Yikes! I can’t tell you how many times I’ve misattributed this to Asimov. I guess it’s like DM nd wellness outcomes, in that you can make stuff up and no one calls you out on it.
If this blog is healthcare’s 60 Minutes, these guys are Andy Rooney. Thanks for making us smile
I haven’t attended your presentation – the quote is very useful in many different situations.
Another one of my favorite Wizard of Oz quotes (actually a paraphrase):
“I can’t give you a brain, but I can give you a diploma”
Sounds like Mr. Flyer has attended one of my presentations, where I propose those exact words…along with Groucho Marx’s: “Who are you going to believe, me or your own eyes?”
Wizard of Oz: “Pay no attention to that man behind the curtain.”
The Humana wrinkle is interesting. They are an excellent health plan, and I just today reviewed one of their vendors and was very impressed both with the vendor and with Humana’s metrics for evaluating their outcomes. So we must be missing something in the Pharos “secret sauce” and I look forward to hearing what it is.