By Vik Khanna

Suppose one day you sit in front of your work computer, click on a link supplied by your employer, and set about the task of answering a hundred or more highly intrusive health questions.  Setting aside the issue of financial penalties or rewards for doing the survey, you would trust that the instrument itself, called a health risk appraisal (HRA), would actually have a sound scientific basis, especially since its ultimate goal is to give you purportedly accurate health guidance.

Unfortunately, your trust in the validity of the tool would be quite misplaced.

HRAs are an essential screening tool in workplace wellness programs despite the fact that no body of evidence clearly demonstrates either their fiscal or clinical value and that no health services research has determined which HRA is the optimal tool.  Indeed, a recent review of HRAs concluded that they increase spending, not reduce it, and that no one has any idea whatsoever whether taking an HRA has anything to do with the delivery of health value.

By masking essential methodological truths about HRAs, wellness vendors have essentially hustled their employer clients into believing that HRAs, which frequently ask clinical questions best left to primary care clinicians or restate platitudes (gosh, I didn’t know it’s safer to drive while not under the influence), are both probative and predictive of a person’s health future.  This is just simply wrong.

Start with the problem that there are multiple instruments in the wellness marketplace.  Wellness vendors all claim to have the “best” HRA as a way of differentiating themselves from competitors, even though there is no comparative analysis of HRAs to underlie the claim.  There aren’t even good anecdotes about the value of one HRA versus another.  Because of this lack of rigor, employers have no way of knowing which HRA questions their employees will find most understandable or answer truthfully, or even which question form is most likely to elicit an accurate answer.  Survey design is a complex process, which is likely why it’s a foreign language to wellness vendors.  Consider, for example, that other surveys routinely used to gather information about lifestyle habits, such as the government’s health eating index and Harvard’s alternative healthy eating index, have been studied and validated in multiple settings, thus giving confidence that they actually gather useful data.

The closest that the wellness industry gets to a patina of credibility on HRAs comes from the National Committee for Quality Assurance (NCQA), which certifies them if they meet certain standards.  It’s a nice idea, except that the standards are process-related, and ignore the issues of reliability, validity, and evidence.  Notice that in neither of two HRA-related documents (see here and here) on the NCQA website do the HRA standards require that the instrument have been tested and validated; NCQA President Margaret O’Kane does note, however, in her letter to the Centers for Disease Control and Prevention that administering an HRA to employees without an accompanying cash incentive is likely a fool’s errand because almost no one will complete it.  In other words: people have no interest in answering the questions on HRAs, unless you bribe them to do so.  Hence, both the wellness industry and its only conceivable legitimacy cover, the NCQA, treat HRAs with studied nonchalance.  This would be cool if we were talking about high school, but is unforgiveable in an industry that is plowing ahead with the taxpayer-subsidized bamboozling of employers with something that is simultaneously easy to explain or sell and possibly worthless.

Given that recent reports raises doubts about how frequently patients lie to their doctors, with whom they are supposed to have a nearly reverential relationship, it is inconceivable that employees are uniformly truthful on anonymous surveys, the mere completion of which, as President O’Kane notes, is a winning lottery ticket.  Consider, for example, that in the fall of 2012 a Fortune 500 company told its employees that simply completing an HRA would earn them a premium discount in 2013 and that the truthfulness of their answers mattered not.  Who wouldn’t take the survey and the money?   And yet 20%-50% still don’t bother, which is why incentives have soared to $521 in recent years.

The next problem that arises is no wellness vendor can show that using an HRA in an employee group over time (say three to five years), causes any actual improvement in health or health-related quality of life (HRQOL).  HRA proponents are fond of pointing to studies showing that employees answering the surveys make beneficial lifestyle changes.   Studies such as this one, however, only report a “desired” result of either static or improved health status.  Self-reported worsening of health status is left out of the equation, because vendors typically exclude increases in risk among people who were previously low risk.  Vendors also usually count only people who completed the HRA twice, thus ignoring the ones so discouraged that they dropped out of the wellness program.  Likewise, no one maps self-reported changes for HRA takers against actual medical claims for wellness-sensitive events for the entire employee group.  The main reason that the mapping is not done is that it would surely sink the HRA ship.  Most people have relatively little interaction with the medical care system.  In particular, there are very few costly wellness-sensitive events (such as coronary artery bypass surgery or angioplasty) in most employee groups; a large proportion of medical care problems are benign and self-limited; people are more inclined now than ever before to self-treat; and many, many people are actually still quite healthy (which does not mean that they should treat their good fortune with indifference, at least not if they want it to last).  For those with outright wellness-sensitive disease or elevated risk factors, the answers to an HRA are far less revealing than claims analysis and far less useful than conservative, evidence-based medical management.

So, how, then should employers gather the data necessary to build a sustainable, cost-efficient, participatory wellness program?  First, employers should distance themselves from HRAs and biometrics (which Al Lewis and I will write about elsewhere), and, in fact, separate themselves completely from strategies such as these that undergird health contingent wellness programs.  Asking employees to expend time, energy, and money taking annual (or, monthly or quarterly, as some wellness vendors propose, with even less evidence to support the approach) surveys that lack validity is purposeless.

Second, employers must understand their group demographics.  Age, income, and education are strongly linked to both health status and behaviors.  Knowing whether your group is older or younger, less educated or more educated, and better paid or lower paid can go a long way towards the reform of the workplace culture and development of a robust participatory wellness program that is less costly, less intrusive, and more likely to raise employee morale, which is essential to program participation as well as recruitment, retention, and productivity.

Third, employers should have their health plan, or in the case of a self-insured company the third party administrator, provide them with a thorough claims analysis.  At a minimum, this analysis should show the inpatient, outpatient, and prescription drug spending stratified by the total number of events, total costs, per-event costs, age bands, and gender distribution, for the top 10 primary diagnosis codes and then for a separate list of codes that are specifically related to wellness.  Vendors never track wellness-sensitive medical events even though they run wellness programs, and benefits consultants are fine with this omission. This data view will, relatively quickly, inform the employer about whether it needs a wellness program or a better managed care plan.

A far more useful tool for understanding HRQOL than HRAs is the SF12, which is actually a valid survey and the accepted gold standard for understanding HRQOL.  There is even a companion SF10 that parents can complete for their children, which should matter to employers given that kids are often covered lives.  These two tools will help employers understand whether their workers suffer from poor HRQOL.  If they don’t, and the claims analysis view shows little that wellness could impact, then the employer can justifiably make a shift in planning from a health contingent program to a participatory one.

Employers could even start this process by randomly sampling employees with the SF12 and/or SF10.  Sampling is a far more credible tool than bribing people with “incentives”.  When the federal government wants to know about the population health status of Americans, they don’t give everyone an HRA and take biometrics.  They sample the population using, for example, NHANES and the BRFSS.  If we build health policy for the entire country with these (and other data sets) that sample, how is it possible that an employer needs (possibly deceptive and certainly coerced) answers from every single person in the workforce in order to design a wellness program?

Last year, in a meeting of health industry titans, I asked how many of the dozen or so attendees had taken the HRA that their employees took.  Every hand went up.  I then asked who in the group had ever had an illness that an HRA would have helped them avoid.  Every hand went down.  Finally, I asked how many had healthy product selection and pricing policies for the food served to employees.  All hands remained down.  The goal of reducing the demand curve for medical care through actual improved health and the compression of morbidity is unreachable as long as the focus is on the false economies of invalid surveys rather than the imperative of health-driven cultures.

If you’re reading this, make sure to also read it’s sister piece by Al Lewis- Caution: Wellness Programs May Be Hazardous to Your Health.

Vik Khanna is a St. Louis-based independent health consultant with extensive experience in managed care and wellness.  An iconoclast to the core, he is the author of the Khanna On Health Blog.