Collins, Thomas R.
ARTICLE IN BRIEF
In a review of literature available on population-based dementia screening, investigators found scant evidence to support its use, leading them to conclude that more inquiry needs to be done to assess the benefits of the screening — as well as the potential harms that are inherent to this kind of screening.
Even as screening of the population for dementia becomes a more widely accepted step toward launching early treatment and, hopefully, helping curb disease progression, a systematic review out of the United Kingdom shows that the practice is not based on solid evidence.
A team of researchers from the University of Cambridge conducted a system search of available medical literature on population-based dementia screening — combing more than 14,000 studies through May of 2012 — and found scant evidence from 74 of the most relevant papers on the topic. They concluded that more research needs to be done to assess the benefits of this practice — as well as the potential harms that are inherent to this kind of screening.
The investigators looked for literature — either as a standalone or component of a dementia intervention compared with a routine pattern of care in general population or non-selected general practice or community care attendees. They were unable to find studies looking at potential harms caused by the screening results (such as risk of depression, anxiety, stigma or independence).
Among six of 26 papers that looked at economic outcomes, three primary studies reported on the direct cost per patient diagnosed; two related modeling studies compared a population screening scenario with other models of dementia care, and one looked at screening in older drivers.
None of those studies offered information on cost-effectiveness of screening or generalizable information on cost. The available data suggest “that substantial resources are required to screen for dementia, which are determined by the age of the screened population, the properties of the screening instrument, and the extent to which general practitioners are involved in follow-up assessments,” the investigators noted in their abstract.
“The evidence base on which the rhetoric is built is very weak,” the lead researcher Carol Brayne, MD, professor of public health medicine at the University of Cambridge, told Neurology Today. “And therefore what we've identified is a very urgent gap in need of filling because we don't really know what people think about screening, about what the consequences of screening are, and there are no trials which show that there is a benefit to screening nor what the cost and impact of that is.”
She said that policies promoting population-based dementia screening seem to be based simply on the sense that there must an inherent benefit.
“A lot of the policy, shall we say, is built more on the concerns that people have and the belief that these kinds of interventions will be self-evidently of benefit,” she said.
In particular, she said, “there's virtually no attention to harm,” even though it's an extremely important issue. She said “there often appears to be an implicit assumption that a diagnosis is always perfect.”
“All screening programs do harm because you'd never have a test that is fully accurate, 100 percent accurate,” Dr. Brayne said. “So inevitably there is misdiagnosis. Some people will be diagnosed and will not have the disorder but will experience potential harms including anxiety and overtreatment, and others will not be diagnosed who do. These things are never perfect.”
Another analysis from the same group had found a dearth of evidence to show that population screenings for cognitive impairment have a significant benefit on clinical or psychosocial outcomes. [See the Neurology Today article, “Does the Evidence Support the Benefits of Population Screenings for Dementia? Not So Much, New Analysis Reveals:” http://bit.ly/18Yu6Ia.]
Malaz Boustani, MD, MPH, who is the director of the Healthy Aging Brain Center in Indianapolis and has studied this topic for about 10 years, said he once thought that dementia screening had to be intrinsically beneficial, but his work has found that the question remains unanswered. He has also been a co-author with Dr. Brayne on several other studies, but was not involved in the current review of population-based screening.
“I was one of the people early on in my career who thought it's obvious that early detection of dementia is a good idea — you don't need science, you don't need evidence to back that up,” he said. “So one of my early mentors told me, ‘OK, just try to study and see what happens if you actually do screening.’ And what I have discovered in the real world, all that perception that I had is not backed up by evidence.”
One problem is that while most patients — some 90 percent, he's found — are agreeable to initial screening, many balk when it's found that they show signs of possible dementia, and refuse to undergo diagnostic testing.
“When we tried to find why, we noticed a lot of them were concerned about the stigma of dementia, so when it was theoretical...they had no hesitation of undergoing the screening,” Dr. Boustani said. “When the screening becomes positive and becomes real, then a lot of them, 50 percent of them, refused (the diagnostic evaluation).”
The larger the number screened, and the wider the net cast, the more potential harm that can be done, he said. “One percent is low if you're dealing just with 100 people, but 1 percent if you're dealing with a million is a lot of people,” he said.
He is now recruiting patients for a randomized controlled trial on the benefits, harms and costs of screening. “We will find out after one year if the benefits of screening outweigh the harms in comparison to the group who did not undergo dementia screening,” he said.
The costs are a crucial consideration in whether to proceed with population-based screening, he said, because the care for currently diagnosed dementia patients has been found to be poor in many instances.
Rachelle Doody, MD, PhD, director of the Alzheimer's Disease and Memory Disorders Center at the Baylor College of Medicine, cautioned that since details on the data on the systematic review had not yet been published, it was difficult to draw helpful conclusions from it.
That concern notwithstanding, she contends that screening should be reconceived. “There should be multiple levels of screening that occur within medical settings and are individualized to a person's risk. That's what I think screening should be, and I think it should occur.”
“As for some kind of screening that takes place in a random geographical area without regard to a person's medical status,” Dr. Doody added, “I don't see how that could be very valuable.”
More work needs to be done on the parameters that are the most valuable for dementia screening, she said.
“When you're talking about individualized personal screening or screening within a medical practice, I think it's very clear that we need it,” she said. “The question is, who should be screened, how much should you screen? Those details we don't have a good evidence base for.”
She said the potential harms should not be a major concern if screening is done properly. “With a rational screening approach, I don't think that people have a tremendous anxiety,” Dr. Doody said.
As far as cost, she said, “If it's worked into a medical practice, resources are not large.”