The widespread perception among many Americans is that attention deficit hyperactivity disorder is overdiagnosed. This was fueled by a regular update to a dataset the U.S. Centers for Disease Control and Prevention (CDC) releases every few years called the National Survey of Children’s Health. The recent data showed — not surprising to anyone — that diagnoses of ADHD in children 2-17 years old increased since the last survey.

This release caused the New York Times to blare in a headline that 1 in 5 of all boys in the U.S. had ADHD. (Which turned out not to be true, but you wouldn’t know it unless you scrolled all the way to the bottom of the article and read the “correction.”)

In fact, if you looked at all the data the CDC released, you’d notice similar increases across the board of childhood diagnoses — increases in the rate of diagnosis of autism (up 37 percent from 2007), depression (up three percent from 2007), and anxiety (up 11 percent from 2007). But for some reason, the New York Times only covered the changes to ADHD diagnosis rates.

So is there an actual overdiagnosis in ADHD? Or is it more complicated than that? Let’s find out.

Let’s Ask Therapists to Analyze Case Stories

One attempt to get at the answer of whether this data represents an “over”-diagnosis or not was Katrin Bruchmüller’s study (et al., 2012) which presented four short case vignettes (short stories describing a patient’s symptoms and presentation) to 463 German child psychologists, psychiatrists and social workers. Only in one vignette was enough information to diagnose ADHD definitively; in the other three, information was missing to make a diagnosis according to the ADHD diagnostic criteria.

Despite the lack of information, therapists diagnosed between 9 and 13 of the girls in the latter three vignettes as having ADHD. It was worse for boys — between 18 and 30 percent of them were diagnosed, despite the lack of symptoms meeting the official ADHD diagnosis.

Here’s the thing, though — therapists also missed the clear ADHD diagnosis in 20 percent of boys and 23 percent of girls (even though they were instructed to make a diagnosis). In other words, the rate of diagnostic error among these same clinicians is at least 20 percent.

And that’s the second problem with this study — therapists were instructed to make a diagnosis. When given a survey and asked to make a diagnosis, what are most therapists likely to do? Follow the instructions and make a diagnosis. The survey was, in my opinion, poorly constructed with an unintended response bias — that is, it was biased toward getting therapists to make a diagnosis (even though in 50 percent of the vignettes, no diagnosis could be made).

The other clear limitation of this study is that it’s an experimental study, asking therapists what they might do in some hypothetical example. It’s not a naturalistic data analysis of what therapists actually do in their consulting office. Is a therapist really going to spend that much time thinking or rethinking their choices on a research survey, compared to what they might do if it were their own real-life patient? ((Yet another limitation of the study is that it’s German; we don’t know if we’d find the same or similar results if American therapists were surveyed, as each culture brings its own cultural baggage into the equation.))

So while this study adds another datapoint, it still fails to answer the question conclusively. Sciutto and Eisenberg (2007) concluded that there does not appear to be sufficient justification for the definite conclusion that ADHD is systematically overdiagnosed:

“No studies [exist] that compare the diagnoses being given in actual practice to the diagnoses that should have been given based on standardized comprehensive assessments.”

Bruchmuller et al. claim their study provides that data. But it doesn’t, since it doesn’t measure anything about clinicians’ actual practice.

So, sorry, but Sciutto & Eisenberg’s claim still stands — the research is decidedly mixed on whether ADHD is overdiagnosed or not.

Do Screening Measures Contribute to the Problem?

Some have suggested that overuse of screening measures — especially as a standardized practice for anyone who presents with a physical concern to their family physician — contributes to an epidemic of overdiagnosis.

But the research shows differently… Screening assessments, when used in a primary care setting, could actually help reduce the fact that most doctors miss the symptoms of depression in their patients (up to 50 percent of depressed patients aren’t recognized) (Egede, 2012; Vöhringer et al., 2013). If it’s true for depression, it wouldn’t surprise me that it might also be true for other mental disorders, such as ADHD.

Which is a part of the solution — and a part of the problem. Lots of people get into mental health treatment through their primary care physician, but that may not always be a good thing. Whether it’s because a doctor is lazy (or simply a lazy diagnostician) or people are lazy, treatment often ends there too — with a quick prescription and no followup care. Most people either don’t fill the prescription, or take it for a few months, see little change, and discontinue it on their own (Egede, 2012).

“When depression [for instance] is “over-diagnosed”, it is usually the result (in my experience) of hasty and inadequate assessment — not use of a “screening” instrument,” suggests Dr. Ron Pies, a professor in the psychiatry departments of SUNY Upstate Medical University and Tufts University School of Medicine.

Furthermore, as Phelps & Ghaemi (2012) note, absent a universally agreed upon set of clinical criteria and a corresponding biological validator or biomarker, how do we objectively determine what is “over” diagnosis of a disorder to begin with? More than we’d like? More than a society “should” have? The research evidence suggests that there actually is probably both some overdiagnosis, and underdiagnosis of most kinds of mental disorders.

Journalists’ Bias Doesn’t Help

Some people in the media appear to already know the answer — despite science’s mixed and inconclusive findings. That’s easy to fix when you’re a reporter, however — you simply leave out any disagreeing viewpoints and data. The reader is none the wiser, unless they go and do the research themselves.

An article entitled “A.D.H.D. Seen in 11% of U.S. Children as Diagnoses Rise” by Alan Schwarz and Sarah Cohen is one such example. Using some fresh data from the CDC, it let us know that “11 percent of school-age children overall have received a medical diagnosis of attention deficit hyperactivity disorder.”

For comparison’s sake, in 2003 7.8 percent of children had ever had an ADHD diagnosis, with the highest prevalences noted at 14.9 percent of 16-year-old teenage boys, and 6.1 percent of 11-year-old girls. Medication use for ADHD, according to the CDC, has nearly doubled in the past decade, from 4.3 percent of school-aged children in 2003 to 7.6 percent of children (2-17 years old) in 2012.

So in a decade, diagnoses have apparently gone up just over 3 percent. Not as sexy a headline — nor anywhere close an epidemic of overdiagnosis — when you put it into that context. Medication use is up a lot more, but there are also a lot more ADHD medications available than there were a decade ago (and with them, more direct-to-consumer advertising, which may spur some to ask for a medication first).

The media’s hyperbole and inaccuracies in reporting on this issue don’t help the matter any, either. Look, for instance, at the three editorial notes editors at The New York Times had to make about an article about this issue earlier this year:

Correction: April 1, 2013

An earlier version of the headline with this article referred incorrectly to the rate of A.D.H.D. diagnosis in boys in the United States. Nearly one in five high school age boys have been diagnosed, not boys of all ages.

This article has been revised to reflect the following correction:

Correction: April 2, 2013

A headline on Monday about the marked rise in diagnoses of attention deficit hyperactivity disorder, according to new data from the Centers for Disease Control and Prevention, described incorrectly the disorder that saw the increase. It is A.D.H.D. — not hyperactivity, which is present in only a portion of A.D.H.D. cases. The article also misidentified the organization that plans to change the definition of A.D.H.D. to allow more people to receive the diagnosis and treatment. It is the American Psychiatric Association, not the American Psychological Association.

This article has been revised to reflect the following correction:

Correction: April 3, 2013

An article on Monday about the marked rise in diagnoses of attention deficit hyperactivity disorder misstated the increase in the past decade of children ages 4 through 17 diagnosed with A.D.H.D. at some point in their lives. It is 41 percent, not 53 percent.

It seems to me that there was a clear effort here to exaggerate the claims regarding the data. And not just one correction needed to be made, but three — which is pretty unusual for the prestigious New York Times.

When journalists — whom we expect to be unbiased and objective reporters of the data — can’t get even the basic facts straight, it makes you wonder. Who can we turn to for objective reporting on this issue?

Part 2 of this article, where I cover the recent BMJ study and share my conclusions, is here.