They Said What?

Home » Posts tagged 'Cerner'

Tag Archives: Cerner

Virgin Pulse has the best health risk assessment

Remember when the sergeant asked for a volunteer to step forward and Moe and Larry took a step back?

Image result for three stooges in the army

Looks like Virgin Pulse has won the best-HRA contest the same way. Yes, it is a complete waste of employee time and employer money. In HRAs, that’s table stakes. It earns a solid “C” not because anyone will learn anything of any value, but rather because the other two are worse. The VP HRA offers so little advice that none of it is really bad. It that sense it is better than Cerner’s (“F”) and Optum’s (F+).

Here’s what Virgin Pulse’s HRA gets points for not doing:

Tell employees to eat nonfat yogurt that is full of sugar (like Cerner):

Imply that employees on opioids should get more opioids (like Optum)

Advise employees against using olive oil (like Optum)

Further, it is actually coherent, instead of containing questions like… (Optum again)

…in which both non-asthmatics and completely uncontrolled asthmatics would give you the same answer.

Virgin Pulse’s “advice”

“Eat a healthy breakfast.” Who can argue with that? It’s like a manager advising the batter to hit a home run. Obviously the key is in the execution, and naturally Virgin Pulse provides none. In that sense, it is the best HRA because the others give such bad advice.

What constitutes a healthy breakfast? Can’t be just any old cereal. Those are full of sugar. Ah, maybe something like Kellogg’s Smart Start with Antioxidants, which is “heart-healthy.”

Image result for kellogg's smart start antioxidants cereal

Oops. That’s on Eat This Not That’s “worst healthy cereals” list.  Because it is — you guessed it — full of “heart-healthy” sugar.

Kellogg’s Special K Protein Snack Bars? 10 grams of added sugar apiece (and you’d likely eat 2–they’re small) with six different sugars dispersed throughout the ingredients label.

Eggs?  For most people, eggs are a healthier choice than most packaged cereals. Of course, there are exceptions — a bit less than 1% of the population is genetically predisposed to high cholesterol. But wellness vendors would lose money if they actually spent the time to address these nuances. Of course, in this case, the “nuance” is that familial hypercholesterolemia (FH) is a much better predictor of cardiac events than anything in any HRA. As in about 40% of everyone with untreated FH will infarct before age 60.

Oatmeal?  Probably the greatest consensus around that. (And that’s the breakfast of choice here at TheySaidWhat? World Headquarters.) Yet plenty of cereal companies, and fast food chains, have turned oatmeal into junk food. Dunkin’s “Brown Sugar Flavored Oatmeal” packs in 28 grams of sugar (none of which is actual brown sugar, as luck would have it).

Orange juice? That counts as a serving of a fruit according to the American Heart Association (AHA) and Virgin Pulse is a big one for telling you to eat up to 11 servings of fruits and vegetables a day, with “1/4 cup” of juice counting as a serving. Using the AHA math, an 8-ounce glass of juice would appear to get you 4/11ths of the way there.

And ads for “healthy” cereal sometimes show a glass of orange juice on the side:

Image result for cereal ads in magazines orange juice

On the other hand, NPR and Consumer Reports advise against considering juice as a healthy alternative. The sugar is “natural,” but there is an awful lot of it. (I myself am not a nutritionist, and don’t even play one on TV, so I don’t have an opinion on whether natural sugar is OK and added sugar isn’t. Except I would point out that grape juice contains natural sugar while grape juice concentrate is an added sugar. Both can’t be right.)

So which is it? Drink more juice, or don’t? Which constitutes “eating a healthy breakfast” ? I don’t know and I suspect neither does Virgin Pulse. The difference, of course, is that TSW isn’t charging lots of money to give employees advice.

Completing the Virgin Pulse HRA 

Obviously, employees are going to lie on HRAs, and Virgin Pulse’s is no exception. I was invited to watch a group of employees completing theirs, For each answer, they gamed the system. What is the optimal amount of stress to claim? Too much might harm their career. Too little and someone might give them more work.

How much should they claim to drink?  “Not at all” might cause their (well-known!) employer to think they’re lying, so they all decided to cut their true totals in half.

Servings of fruits and vegetables? Most of them just made up a number that would sound good to an employer.

Everyone seems to know employees lie on HRAs (including, of course, Virgin Pulse) except the employers that still use them, whose quest to create a culture of health ends up creating a culture of deceit. No need to take my word for this. Simply complete this chart to see what percentage of your employees with something to lie about are indeed lying — and what the odds are that an employee with something to hide will tell the truth about it:

Let’s try completing this chart, using the most recent winner of the C. Everett Koop Award, Wellsteps.

Result: in the so-called best wellness program, almost 2/3 of employees lie, though in all fairness to Wellsteps, some of those could be the same employee lying twice. And extrapolating from Wellsteps’ result, the odds of an employee coming clean about a bad health habit are about 1-in-4. That’s the average of smoking and drinking. Generally the more socially acceptable habits will show a higher percentage of employees telling the truth.

Keep in mind, too, that there is another way to lie on HRAs, which is understatement. You may recall from Wellsteps that the 23% who drink only imbibe 1.3 drinks a day, while 3% who do admit to smoking indulge in cigarettes only 4.27 times a week, perhaps taking a break on weekends and major holidays, such as Beethoven’s Birthday.

Image result for beethoven's birthday

Due to the lying and uselessness, should employers drop HRAs altogether?

In the immortal word of the great philosopher Curly Howard, soitenly.

Image result for curly howard



New study says employees should stop drinking. Ignore it.

Apparently the wellness industry does not have a monopoly on invalid research.

A study came out in The Lancet–the British equivalent of the New England Journal of Medicine — finding that the only safe level of alcohol consumption was: none.  As the principal investigator said: “Alcohol poses dire ramifications for future population health in the absence of policy action today.” This finding generated myriad headlines like this one at CNBC:

And how often do those two outlets agree with Fox News?

One thing you learn if you hang around wellness promoters long enough is that oftentimes a close perusal of the study in question shows the opposite of what the authors intended. Or, as we often say: “In wellness, you don’t have to challenge the data to invalidate it. You merely have to read the data. It will invalidate itself.” And the same is true here.

For example, Denmark leads the world in the number of drinkers — and has life expectancy higher than about 90% of the world’s countries. The lowest alcohol consumers? Pakistan — which ranks #130 in life expectancy. You might say: “Wait, aren’t there many other factors involved in life expectancy?” And the answer is, of course there are. None of those were controlled for in any way in this meta-analysis.  To begin with, the more people drink, the more other unhealthy habits they are likely to have.

But that’s not the crux of what is wrong with this study. Two other things should lead wellness professionals to the opposite conclusion: that light drinking is perfectly OK. The remainder of this post addresses those.

Absolute risk vs. relative risk

Absolute vs. relative risk is one of our (many) pet peeves. Here are two other examples that we have had to smack down:

  1. The American Cancer Society warns of a 22% increase in colon cancer among people under 50, but it turns out that absolute rate of colon cancer in younger people is so low that the chances of your life being saved by screening at age 45 are about the same as your chances of being struck by lightning.  The media had a field day with that one too.
  2. Before that, speaking of colons, a study came out showing that red meat increased risk of dying from colon cancer. Once again, it turned out — using the data right in the study — that more people are killed by lightning than by colon cancer due to eating more red meat than average.  Yet once again, the media had a field day.

From the media’s perspective, this makes sense. After all, who is going to click through on a headline that says: “Low quality study finds trivial relationship between variables” ?

In the case of this alcohol study, looking behind the headlines proved equally insightful. (And thank you to Aaron Carroll of The New York Times‘ Upshot for suggesting it.)

Here is the lead-in:

Alcohol is a leading risk factor for death and disease worldwide, and is associated with nearly one in 10 deaths in people aged 15-49 years old, according to a Global Burden of Disease study published in The Lancet that estimates levels of alcohol use and health effects in 195 countries between 1990 to 2016.

Based on their analysis, the authors suggest that there is no safe level of alcohol as any health benefits of alcohol are outweighed by its adverse effects on other aspects of health, particularly cancers.

Read the first paragraph again. Two observations:

  1. almost no one dies between the ages of 15 and 49, so being responsible for “nearly” 10% of those deaths means that alcohol kills about 0.001% of people in that age bracket every year.
  2. the authors have conflated two things: alcohol and excess alcohol. Virtually all of those deaths in that age bracket were due to the latter, a fact that the authors conveniently overlooked when demonizing any level. of consumption.

Reading a bit farther in…

They estimate that, for one year, in people aged 15-95 years, drinking one alcoholic drink a day [1] increases the risk of developing one of the 23 alcohol-related health problems [2] by 0.5%, compared with not drinking at all (from 914 people in 100,000 for one year for non-drinkers aged 15-95 years, to 918 in 100,000 people a year for 15-95 year olds who consume one alcoholic drink a day)

Hello? A 0.5% increase in relative risk? And the increase in absolute risk (not calculated) is 4 per 100,000 people a year — or 0.004% a year.  Even two drinks a day increases absolute risk only by 0.06% a year. (Once you get beyond two drinks a day, the chance of harm accelerates exponentially…but that’s not news.)

What the he** are employees going to consume instead?

Our biggest beef with this study is the same as with just about every wellness program: everything is off-limits. Even foods that are OK in moderation for most people — like full-fat dairy, salt, oils, cholesterol/eggs and red meat — are singled out for criticism by health risk assessments. And now alcohol.

Unfortunately, the more foods you demonize, the less likely it is that any employee will pay any attention to any of your dietary pronouncements.  And to the extent they do. well what are they going to eat instead? Here is Cerner telling people that non-fat yogurt is a “healthier choice.”  Trivia question: What added ingredient makes nonfat yogurt taste good?

Here is Optum railing against oils:

And Cerner, once again, this time incriminating dietary cholesterol, which of course has no impact on blood cholesterol for most people:

Finally, here is Interactive Health hyperventilating about something-or-other in their HRA feedback to an employee.  We don’t know what it is other than, given the provenance, it’s wrong. Fortunately, no employee is going to plow through this anyway.



Treat this alcohol finding the same way you would treat advice from most health risk assessments: ignore it.


Cerner’s HRA scores an “F+”

My recent request to review health risk assessments (HRAs) brought a number of responses. I’m grading the HRAs I was able to complete, on both advice and readability.

Cerner’s, the first to be reviewed, receives “F” in advice and, as will be shown below, D in readability, for an average of F+. Advice is scored as follows:

  • A: Virtually all the advice is up to date and correct
  • B: The advice is generally correct
  • C: There is good and bad advice in roughly equal amounts
  • D: Bad advice outweighs good
  • F: Employers using this HRA should caution employees not to take the advice

For each HRA being graded, unless otherwise indicated, I am completing them exactly the same way. I am somewhat understating the quality of my diet and the amount of exercise I do, in order to get “mainstream” advice.

High-risk for alcohol and drug use is incorrect. 10 mg of Ambien, taken as directed by a physician, does not create high risk in the absence of other drug and alcohol risk factors.

Ironically, there is a risk in telling people they are high-risk when they aren’t, which is that they will simply lie, in order to move their score left, out of that red segment, and out of the embarrassing range. As a result, they would not exposed to advice they should be getting if indeed they are at risk for drugs and alcohol.

And then they realize it’s OK to lie in general on HRAs. These lies take place on a massive scale, invalidating the entire HRA instrument.  In an attempt to create a culture of health, companies are creating a culture of deceit.

This blood pressure advice is incorrect. Someone with a “pulse pressure” of 25 (110 – 85) is not at “moderate risk” of anything. This is an emergency situation, likely indicating heart failure fluid overload. (This is not my blood pressure, by the way, or I’d be practically dead.)  

This advice is wrong on five dimensions.

  1. People should not be taking blood pressure advice off a computer-generated algorithm, especially one that hasn’t been updated in years. That’s what doctors are for.
  2. It’s not at all clear that lowering sodium reduces blood pressure in the large majority of adults. This should not be offered as a fact. In some people it makes a difference. Advising people on their own situation is what doctors are for. 
  3. “If you limit your sodium to 1500 mg a day, you can lower your blood pressure even more.” Actually the Framingham Offspring Study says the reverse. It turns out that people who say they consume 4000 mg/day of salt have the lowest blood pressure. (The link is to a lay article that itself links to scholarly material.) Not clear which is right, but advice this controversial should not be passed off as a fact when the evidence conflicts.
  4. It is possible that this Cerner-recommended low-sodium diet increases cardiovascular risk. Other studies, largely older ones, say this diet reduces risk. It’s safe to say there is no consensus. But HRAs, which employee are being financially coerced into doing, should not be offering controversial advice as fact.
  5. Telling people to avoid salty snacks may encourage people to eat sugary snacks. The HRA is quite deficient in warnings about sugar.

This advice to switch to lowfat dairy is more likely to cause harm than to create benefit. Full-fat dairy is preferable to fat-free for most people. As this summary, with links to the studies, shows, full-fat dairy probably offers protection against diabetes.

And don’t overlook their complete rookie mistake: telling employees to eat “low-fat or nonfat yogurt,” which of course is full of sugar.

The advice regarding saturated fat in meats is controversial but is presented as fact. There is a whole body of research saying the opposite of what Cerner says. Once again, we aren’t taking sides except to note that coercing employees to complete HRAs implies that the HRAs should be accurate.


Further, Cerner’s HRA, as Yogi Berra might say, contains a lot of omissions. The following pieces of advice should be included, but aren’t:

  • Advice not to text while driving (in lieu of the seat belts question*)
  • Health literacy information on finding hidden sugars
  • Health literacy information on finding hidden salt (who knows how much “1500 milligrams” is, assuming that advice were good?)
  • Eggs are one of nature’s healthiest foods for most people
  • Shingles vaccine addresses the most easily avoidable risk for people over 50
  • Questions on opioid awareness (not opioid use, which people will just lie about) and information on how addictive they can be

*Not a useful question and speaks to a lack of updating. While decades ago, a few employees might have benefited from the advice to buckle a seat belt, no employee in recent history has ever not buckled their seat belt, taken an HRA on which they admitted not buckling their seat belt, read the printout that said they should buckle their seat belt, realized for the first time that not buckling a seat belt was a bad idea…and then started buckling their seat belt.


The scale for readability is:

  • A: Intuitive — didn’t even think about any readability issues
  • B: Readable enough
  • C: Readable with exceptions
  • D: Not readable in many instances
  • F: Exercise in frustration

Cerner earns a “D” only because, as we will see in a future posting, it is better than Optum’s (F) and Interactive Health’s (F). Absent those two HRAs, Cerner would get an “F”.

Cerner says I am “at HIGH risk for chronic health issues.” But on the right it says I am “moderate risk.” That is confusing and decidedly unhelpful in and of itself. But wait. There’s more…now how much risk do I have?

Having sweated bullets while learning I was high risk and moderate risk, I rest easily to learn that I am at “LOW risk for chronic health issues.” 

Eventually I guessed what meaning they were trying to convey, but the typical employee (who, according to Cerner, has to be told to buckle a seat belt) is not going to figure this out.

Why couldn’t they just ask if you are feeling down or depressed? (Not that any employee is going to answer that honestly, in any event.) 

“Has little interest or little pleasure in doing things bothered you?”  Does this mean that if having “little interest or little pleasure in doing things” doesn’t bother you, you write “no”?  I had to read this several times to figure out what they were asking.

It doesn’t help that Cerner’s writing style uses a lettering system for no reason. What do those “a’s” and “b’s” add?  They do that everywhere, to add to the general confusion:

Lots of extra letters, for no reason at all. And “stroke” and “heart attack” are not “conditions.” They are “events.” Heart failure (a serious disease with an average life expectancy of 5 years) should not be lumped in with angina, which can be controlled for many years with diet, exercise and medication.


There is far too much advice in here – salt, dairy, meat, drugs, blood pressure just for starters – for anyone to internalize and act upon. It is a tenet of human behavior that the more advice is given, the less is taken. Plus, more importantly, a large chunk of the advice is at least arguably wrong, if not definitely wrong.

Further, the two things that are most important for employees to learn about — sugar and opioids — are pretty much ignored. A good HRA will laser-focus on sugar harms. Likewise, opioids, the leading killer of people under 50. While an HRA can’t cure an addict, questions like: “How long does it take for the first signs of opioid addiction to start, when used as directed?” could measure people’s opioid awareness and then provide easily digestible information on what the actual answer is (3 days). Quizzify has placed its pain management-and-opioids quiz in the public domain, which could fill that gap.

The graphics with all the HIGH and LOW risk scores are very confusing.

Virtually everyone with a drinking problem will lie about it. (Just check the national statistics against the results of this HRA.) This will encourage dishonesty in other answers, and in general create the culture that it is OK to lie to one’s employer.

This HRA, and HRAs in general, make the mistake of giving the same advice to everyone who answers a certain question(s) the same way.

This health risk assessment will not provide a net benefit to a population. It is recommended not to use it, or at a very minimum insist that whoever uses it double-check any advice offered with Quizzify or another source that is up to date and accurate.

%d bloggers like this: