Marketing surveys help companies and organizations better understand what’s important to the people who purchase (or may purchase) their product. I worked for a few years in marketing for a company, and during that time I learned a lot about how these surveys, when done right, could provide a company with some valuable insight into their product.
But sometimes an organization’s press release about their own survey data will misconstrue their own findings. Why would an organization do that? Well, two answers spring to mind — sloppy reporting by the organizations’ PR people (which seems unlikely, given that is one of the primary purposes of these organizations), or selectively reporting the results that cast the sponsoring pharmaceutical company in the best light.
Dr. Danny Carlat over at The Carlat Psychiatry Blog has an entry on this issue, Deceptive Surveys: Latest Marketing Tactic in the Antipsychotic Wars.
In his entry, Carlat notes that Mental Health America says this about their survey data in the press release they published at the end of January 2008 trumpeting their results:
When choosing from a list of side effects considered when prescribing antipsychotic medication, diabetes was most often cited by prescribers, with 94% of psychiatrists considering it “extremely” or “quite” important.
But it’s even more interesting than that…
The only place the 94% number is mentioned in the actual Executive Summary is in relation to whether psychiatrists say they screen for various health conditions (page 15). It has nothing to do with a “list of side effects” (more on those in a minute).
While it’s true they say they screen for diabetes 94% of the time, they also say they screen for Hyperlipidaemia nearly as often, 88% of the time. (Hyperlipidaemia is the presence of too much fat lipids in the blood, often a sign of potential heart issues.) Psychiatrists screen for many other conditions far more often though — side effects associated with their medication (100%), alcohol consumption (99%), seeing other medical professionals for health issues (98%), follow-up care (98%), physical health status (97%), sleep habits (97%), medicines for other illnesses (97%), etc (page 16). The list goes on.
Psychiatrists prescribe medications for the reasons most people suspect they do — for the long-term effectiveness and safety record of the medication. Effectiveness of the medication to provide short-term relief, convenience of taking the medication and patient preferences were the next set of reasons for prescribing a particular medication. Notice how side effects isn’t even in the top 5 reasons psychiatrists prescribe a particular medication or not.
The three most “extremely important” side effects of concern to psychiatrists impacting a patient’s life are: tardive dyskinesia (59%), diabetes (50%), and acute dystonia (46%).
Yet “tardive dyskinesia” (characterized by involuntary movements most often affecting
the mouth, lips, and tongue) and “akathisia” (characterized by
restless fidgeting and pacing) are not words you’ll find in the MHA press release, while “diabetes” is. Interesting.
We’ll end by noting one of MHA’s “key findings” noted in the press release:
82% of consumer respondents feel that treatment of their overall health — not just their mental illness — is important to their recovery. Yet nearly half expect their psychiatrist to focus exclusively on their mental health (48%), rather than overall and mental health.
Yes, and you know why? Because that is what psychiatrists specialize in — a person’s mental health! Most people who see a psychiatrist also have a general practitioner they see for their general health concerns. So while I think it’s important that psychiatrists be aware of and ask their patients about health concerns, psychiatrists generally don’t treat such health concerns because that is not their specialty area. So to suggest this is some sort of insightful piece of information or “key finding” is really quite ridiculous. It’s like suggesting that nearly half of all people expect dentists to focus exclusively on their dental health. Yes, I think that would be a reasonable hypothesis.
Carlat compares MHA’s study with another study done on family caregivers by another nonprofit, World Federation for Mental Health (WFMH). I kind of found the comparison more apples to oranges.
The data were released before the MHA data, in September 2006.
It should be noted that the MHA survey was done on a completely different set of people — patients and psychiatrists. “Caregivers” are family members who have to grapple with serious mental illness in a family member, and so come at these things from a slightly different perspective. Not surprisingly, this survey found that:
Nine in ten of caregivers agree that efficacy is their primary concern when weighing treatment options for their family member and that an effective medication is needed to control symptoms, before overall wellbeing and health can be properly tackled.
Well, of course. You can’t even begin to talk about a normal life in the disorders studied — schizophrenia, bipolar disorder and schizoaffective disorder — until you’ve got a handle on stable treatment.
Other important findings from this survey:
56% say it took two years or more for their relative to find a medication that worked.
85% say that their relative tried more than two different medications before finding the one that worked and 36% say their relative had to try more than five medications.
Amazingly, nearly 86% of respondents said their family member had relapsed at least once (due to a medication change or discontinuation of the medication against the doctor’s advice). What happens during a relapse?
Caregivers say that as a result of relapse their loved ones were unable to work (72%), were hospitalized (69%), tried to commit suicide (22%) and were imprisoned (20%).
So relapse is an important theme from a caregiver’s perspective, and it can takes years of trying to find the “right” medication that truly helps a person dealing with serious mental illness.
I don’t find this information particularly biased toward any specific drug, but obviously it emphasizes drug treatment approaches over other alternatives, such as the importance of psychotherapy used in conjunction with medications for these disorders.
Sadly, the survey appears to have asked no questions about anything other than medications.
So what is the take-away message from these kinds of surveys?
Survey results can be readily manipulated in a variety of ways. A press release can highlight the concerns or findings — even when they do so inaccurately, as in the MHA case — that a sponsor is most interested in. Surveys can also position specific concerns in the questions asked to ensure they receive more attention than other concerns. Good, objective survey design is often just as important as a survey’s findings, since such design can be used to bias the answers. Not surprising, neither organization released the actual survey used. And not asking certain questions or questions on certain topics ensures those topics receive no answers, again contributing to bias slanted toward a certain perspective or conclusion.
At one time, we could count on non-profits to offer fairly objective results and data that could not be impugned. I think this is becoming less and less the case.