The New York Times got a lot of attention last week with a dramatic headline: “Severe Mental Illness Found to Drop in Young, Defying Perceptions.” The article was a report on a study, published in The New England Journal of Medicine, about trends in kids (6 to 17) getting treated for mental health issues from roughly 1996 to 2012.

Over that period of time the percentage of kids getting some kind of outpatient mental health care rose, from 9.2% to 13.3%. So who, wondered researchers, was that care going to? Was it, as some would argue, going to kids who don’t need it?

To get at an answer, they divided the kids into two groups—more severely impaired and less severely impaired—based on parents filling out scales on 13 kinds of impairment. And they found, as the Times headline points out, that there was actually a decline over that period in reports of kids in the severe category, from 12.8% to 10.7%.

What does that mean about who’s getting the increased care? Here’s where it went:

  • Services to kids with more severe impairment rose from 26.2% to 43.9%.
  • Services to kids with less severe impairment rose from 6.7% to 9.6%.

So the percentage of kids getting services who have severe impairment is more than four times that of less impaired kids, and the percent increase over the period is greater, 67.5% compared to 43%.

And here’s the thing: The increase in care is impressive, but still, less than half of kids with severe impairment and less than 10% of kids with less severe impairment are getting treatment.

The Times also notes the fact that the number of kids with severe impairment in this study are lower than other studies which have put the prevalence of severe impairment in kids at around 22%. Here’s where it gets confusing, because the scale these researchers used for impairment, called the Columbia Impairment Scale (CIS) is different from the criteria used by other researchers.

Many other prevalence studies are based on diagnostic criteria—either via parent reports of a diagnosis by a health care provider or direct diagnostic interviews by a professional of a sample population of kids.

As Kathleen Merikangas, a National Institute of Mental Health researcher who has conducted some of the most significant prevalence studies of children and adolescents, tells the Times, “Right now we have all these different agencies doing surveys, each using a different method. It’s a nightmare. We need to do better.”

And what of the decline in severe impairment? The study’s lead author, Dr. Mark Olfson of Columbia University, offers several possible explanations for this trend: prevention (maybe parents are doing something that’s protecting kids) and increased treatment.

But here’s Dr. Olfson’s final conclusion in the study results: “Despite the increase in treatment, many young people with severe mental health impairment received no care. In schools and primary care settings, improvements in the identification and referral of young people in the greatest need of treatment could provide community benefit.”

Speaking of perceptions of mental health, as the Times is, it’s important not to ignore the obstacles—including, frequently, stigma—that prevent kids from getting the care they need. Recognizing this would have made for a different and more challenging headline.