Bad Data Is Inflating the Prevalence of Depression

Credit: Pexels, Gratisography

The prevalence of depression is overblown by as much 3 times the actual rate, a problem that can be traced back to bad data, according to a new analysis.

Researchers from the Lady Davis Institute of the Jewish General Hospital and McGill University in Canada examined 25 studies on the issue published last year in several top-of-class journals, finding that all but 2 of the investigations used self-report screening questionnaires as the primary means of data collection, according to the study.

“These studies misrepresent the actual rate of depression, sometimes dramatically, which makes it very difficult to direct the right resources to problems faced by patients,” said the study’s lead author, Brett Thombs, PhD, a senior investigator at Lady Davis and a professor at McGill.

Although self-report surveys are beneficial for clinicians and researchers looking to detect people with mental health issues, they are not the ideal way to gather information used to determine the scope of depression, according to the researchers. Rather, health researchers should use diagnostic interviews to confirm cases, they noted.

“We need to conduct a more thorough evaluation in order to determine an appropriate diagnosis and whether there may be other issues to address,” Thombs said.

So, what’s the force behind this disconnect? The common culprit in many stories of bad data usage: time and money. Thombs and his team noted that diagnostic interviews cost more to undertake and also eat up more time than self-report questionnaires.

The study supports similar findings related to depression and self-report data.

But this problem is not unique to depression. Last October, for instance, researchers found that oft-used sampling techniques for brain imaging studies “significantly distorted findings” regarding brain development. The result: What the scientific community thinks it knows about the brain—and, subsequently, health disorders like autism—might be wrong.

The healthcare industry is also battling bunk data, hoping to defy the adage that garbage in will yield garbage out, especially as it pertains to electronic medical records. When harnessed correctly by institutions, data can help curb inefficiencies, produce better diagnoses, and foster better outcomes. But bad data practices can negate that effort and even further set back hospitals, experts have said.

Thombs and his colleagues found 1 factor that could be contributing to bad data on depression, at least in the research world.

“Studies with dramatic results tend to be accepted by higher-impact journals and attract more attention from the public than studies with more modest findings,” Thombs said. “This may also encourage some researchers to report results from questionnaires rather than conducting appropriate diagnostic interviews.

The Canadian Medical Association Journal published the article, “Addressing overestimation of the prevalence of depression based on self-report screening questionnaires,” today. You can read it here.