Attention, They Said What Nation:
We are available for forensic consulting (taking appellations and kicking posteriors), recoupment of fees, and expert witness work, hourly or on contingency (except for expert witness, where it is not allowed). So if you think a vendor has snookered you, who you gonna call?
We. Never. Lose.
And here is an example of why vendors fold…
Accolade recently announced that they had paid Aon a zillion dollars for their actuaries to decide how much savings they could claim without leaving too many clues that the claimed savings figure is fabricated.
They drew that trendline at 8.3%, and decided the savings was also 8.3% in 2019. Could this be an example of “trend inflation,” where, as Optum’s Seth Serxner says: “The choice of trend has a large impact on financial savings”?
Nah, an actuary would have to be both stupid and dishonest to inflate a trend willy-nilly. it would be too obvious a ruse. You’d show savings in:
- every single disease category
- every single comorbidity/risk category
- every single resource use category
To begin with, here is Aon’s trend vs. the sample Accolade clients:
We’re not calling Aon liars, or saying that Accolade’s payments to Aon create a conflict of interest for Aon’s consultants, or saying that Accolade chose Aon instead of the much less expensive and more credible Validation Institute, which backs its validations with a unique Credibility Guarantee, because they wanted Aon’s actuaries to fabricate savings. (The Validation Institute would have to pay out on a ton of Credibility Guarantees if they fabricated savings.)
Quite the contrary, we’re sure that Accolade wanted Aon to publish results that were completely on the level. That’s why they paid them so much money, to check and double-check their work.
Every single disease category
Even so, as we pore through Aon’s report, we can’t help but see some questions that you might want to ask, just out of idle curiosity. By way of background, Accolade is a care coordination/navigation company. If you have an issue, like whether something is covered (benefits questions being the majority of the incoming phone calls), you could call them. They might be helpful. Come to think of it, they are so helpful that — get ready — they save money in every single disease category:
Accolade seems to save substantial sums of money in mental categories — “mental anxiety” and “mental mood” (as opposed to physical anxiety and physical mood, I suppose) — because massive numbers of mentally anxious and mentally moody employees are apparently calling them.
Just suppose for a minute that you were mentally anxious or mentally moody? Would your first impulse would be to call a random number on the back of your insurance card and take a stranger’s advice to save your employer money? That course of action wouldn’t jump to my mind in that situation. But maybe that’s just me. So your question might be:
What do you tell these mentally anxious and mentally moody callers that saves all this money?
Maybe they advise these anxious and moody callers to stop spending money on therapists or meds, and instead just tell them to, respectively, calm down or cheer up. A few other categories might raise questions too, like:
How do you save money in “Upper GI/Esophagus”? Do employees call you and say: “My esophagus is acting up again”?
And that brings us to cancer, where the mind-blowing 18% savings explodes still farther, in another chart showing 2 years of data, to 26%. How do you save 26% in cancer? On average you would be telling every employee — including the 38% to 50% who (according to Accolade) never call even with minor administrative claims issues — not to bother with that fourth round of chemo. And I’m sure they’ll trust Accolade’s judgment on that.
Every single comorbidity/risk category
Aon’s actuaries also sliced the data by number of comorbidities. Initially, they did not show savings in every single comorbidity category. For some reason they struck out with employees who have 3 comorbidities.
We’ll blow up the right-hand “Cost Ratio” column — meaning the % savings — for you:
You could ask:
Why do your phone-answerers do such a great job on people with two or four comorbidities, but strike out on people with three?
And, my personal favorites are about people with, respectively, 0 or 1 chronic conditions:
How do you save 8% on people with (1) zero chronic conditions, (2) no reason to call Accolade, and (3) who already spend less than $2000/year on healthcare to begin with?
How do you save 9% on people with only 1 chronic condition, like maybe mental anxiety or mental moodiness, or perhaps a cranky esophagus?
Accolade/Aon’s answer to these questions on page 15 of their report is:
Only the results for members with 0 or 1 chronic conditions represented statistically credible reductions vs. market controls.
Wait…are they claiming savings only in the cohorts where there was basically no savings to be claimed? Surely not even Aon’s actuaries could be that stupid.
And yet, when you combine the 2,3 and 4 condition categories into one bar, and do the same with the 0 and 1 condition categories, it turns out that they are saying there is no difference at all between the cohorts who could have benefited from a call to Accolade, vs. the ones with no reason to call. (OK, there is a difference – 0.1%.)
Before you assume this is a textbook example of trend inflation, you might ask:
How is it that you save roughly the same amount on the healthiest employees as on the sickest employees?
Surely the data is accurate — if you pay actuaries that much money it should be — but a cynic might expect that care coordination/navigation would be more effective where there is actually the need/opportunity to coordinate/navigate care. Just sayin’…
Every Single resource use category (almost)
Finally, Aon cut the data by resource use.
The good news is that Accolade is keeping patients from using healthcare services, thus saving money in every category of utilization. There is 11% less inpatient use. Outpatient and physician expense declined as well. So another question might be:
Where are you sending employees to get their care?
By contrast, check page 147 of Why Nobody Believes the Numbers. There is an example from Quantum Health where inpatient use declined, but lower-cost resources like doctor visits increased. That increase is a “plausibility check” on Quantum’s shockingly valid claim to have reduced overall spending. Their claim would not have been plausible if every category of resource use declined. As the book says: “if you insulate your house, you’ll save money, but not on insulation.”
The best news? At the very end, Accolade broke their streak of 16 diseases, 4 comorbidity categories, and 5 resources showing savings –by getting a whopping $13 more generic drug use per employee. So, to paraphrase the immortal words of the great philosopher BIll Murray, they’ve got that going for them.
What does Aon have to say about all this?
Initially, they proudly announced the initial “savings” on the Healthcare Hackers group. (Ping me at email@example.com if you want to join that group.) When I and others humbly pointed out just a few of these questions about their analysis, Aon’s Jim Winkler replied that they “stand by” their results.
I learned long ago with the wellness industry not to bother to try to argue with these people because you can’t prove something to someone who just got paid a zillion dollars to “validate” the opposite. So I simply offered to bet a million dollars their results wouldn’t stand up to scrutiny from a panel of reputable health economists.
And that’s when they folded. Not a peep out of them, though other people on the Hackers also asked. And then I noticed that Aon’s chief actuary looked at my linkedin profile.
Never heard from him either, to collect his $1 million. I guess he decided that instead of “standing by,” he should swallow hard and tacitly admit he goofed. (We will assume for now these were all honest mistakes. The way you’ll know is that, having now seen this posting, they will correct their errors.)
Speaking of swallowing hard, if you’re finding Aon’s analysis tough to swallow, you may be right. And you should also call Accolade to fix your esophagus.