ARTICLE IN BRIEF
An assessment of online tests for dementia found that most — 12 of 16 — had poor scores for scientific validity and reliability.
Virtually every neurologist in clinical practice today — and most general practitioners as well — has probably had at least one patient, or family member of a patient, come in to the office concerned about the results of some kind of “online test for Alzheimer's.”
Now, a new study released at the Alzheimer's Association International Conference in mid-July confirms that these “tests” fail important examinations themselves: those for scientific validity, reliability, and ethical standards.
Scientists at the University of British Columbia (UBC) assessed 16 widely available online tests for Alzheimer's disease, using an expert panel of geriatricians, neuropsychologists, neuroethicists, and human-computer interaction specialists to score them on a scale of 1 (very poor) to 10 (excellent) on key factors. All 16 of the tests flunked ethical standards, scoring “poor” or “very poor” because of problems such as confidentiality and privacy policies that were either difficult to understand or entirely absent, undisclosed commercial conflicts of interest, and inappropriate wording of test outcomes.
And most of them (12 of 16) had equally poor scores for actual scientific validity and reliability. In other words, they were pretty useless for the actual purpose of diagnosing Alzheimer's disease (which should come as no surprise to neurologists).
“There were no real outliers, no subpopulation of tests that were really good,” said lead investigator Julie Robillard, PhD, a postdoctoral fellow at the National Core for Neuroethics at UBC. “We're not just dealing with bad information, but bad information overall; the likelihood that your patient will click on something that will help them is very low. It's not about using the 'right one,' it's about not using them at all.”
Some tests went beyond simply being “not useful” into the territory of downright harmful. “Some were predatory: they were designed so that it would be difficult to do well on the test, and then the person would be marketed products for the ‘prevention’ of Alzheimer's disease,” said Dr. Robillard. “We're dealing with a very vulnerable population.”
She pointed to research released in December 2012 by scientists at the University of California, Los Angeles, indicating that even healthy, cognitively normal older adults have greater difficulty than younger individuals in detecting untrustworthy information. (The research appeared in the Dec. 3 edition of Proceedings of the National Academy of Sciences.) “They identified a region of the brain [the anterior insula] that was less active in older adults,” Dr. Robillard said.
COGNITIVE ISSUES ON THE INTERNET
“It's not surprising that all of these websites offering so-called Alzheimer's tests have cropped up,” said Clinton Wright, MD, scientific director of the Evelyn F. McKnight Brain Institute at the University of Miami Miller School of Medicine, who was not involved with the study.
“When it comes to dementia, it's relatively easy to put something online that allows someone to assess their abilities, whether that assessment is valuable or reliable or not. You can't put a test on the Internet that can claim to find out if you have cancer or heart disease, beyond ‘risk factors.’ Dementia is a little bit special in that regard: as a cognitive issue, it lends itself to the Internet.”
Dr. Robillard advises neurologists to discuss the diagnostic process explicitly with their patients. “They need to understand that there is no one single test that can give you an answer about Alzheimer's disease or dementia,” she says. “A memory test is only one part of a multifactorial evaluation,” Dr. Robillard said.
Patients and families will still want to seek out information online, and it's important to validate that desire to self-educate and promote the most reliable sources of information. “There are some potential benefits,” said Dr. Robillard. “Evidence has suggested that people feel more empowered when they think they are getting health information that is tailored to them. They also feel support from the aspect of community that is found in many online health forums. Clinicians should direct their patients to trusted online sources of peer-reviewed information, and reputable support sites, and away from these often-predatory ‘test’ sites.”
Even if your patients aren't asking you about something that they found online, it's important for clinicians to be aware that they're probably doing a fair amount of surfing related to their condition. “There are some people who understand and are very realistic about the limitations of what they find on the Internet, and others who believe everything they read,” said Dr. Wright. “Our patients and their caregivers are generally not trained to judge the quality of the evidence, and they don't even know that they need to be trained. I think this study is very important, and more work needs to be done in this area.”