How An Age-Old Chart Is Redefining Health Care

It’s “just” an electronic growth chart. But lurking inside is a disruptive new approach to medicine.


In a world where organ transplants and MRIs are commonplace, pediatric growth charts don’t sound very exciting. But after spending 30 illuminating minutes on the phone with Harvard researcher Dr. Isaac Kohane, a passionate intellectual who speaks with an infectious urgency, I’m a convert.


“For everything from a rare brain tumor to common garden variety obesity, the growth chart is probably the single-most useful tool to monitor the growth of kids,” Kohane tells me. And then he tells me something else. That with the advent of electronic medical records, the growth chart’s timeless utility has given way to the clumsy hand of clueless programmers, rather than doctors backed by smart designers.

So Dr. Kohane brought together the SMART Platforms Initiative (an organization backed by doctors at both Harvard Medical School and Boston Children’s Hospital to rethink electronic medical records) and service design firm Fjord to redesign the paper growth chart as an app. And then they made it open source to shake things up.

Why Growth Charts Matter

The growth chart is essentially a few line graphs on the same piece of paper. A doctor enters a child’s height and weight (and sometimes head circumference) at various ages. This child’s development can then be compared alongside the standard growth curves, at a glance.

“As we adopted electronic health records, many didn’t have growth charts at all, or they’d have stats like z-scores that say how far you’ve deviated from the mean, which isn’t nearly as sensitive as the human eye is at spotting deviations from curves,” Dr. Kohane explains.


Even slight deviations from the norms are critical, as after the age of three, children follow very precise developmental trajectories. And the way these trajectories interrelate is of the utmost importance. Normal weight mixed with delayed height can signal very specific problems, like thyroid disease or growth hormone deficiency, while normal height mixed with delayed weight might point to obesity. But electronic medical records hid these important relationships.

“The first electronic attempts made the incorrect assumption that it was not necessary to view all three measurements at once, and were created with the premise that by viewing one at a time they could show a larger view,” Brian McLaughlin, visual design lead at Fjord explains. “Doctors and other medical staff were not consulted, and the result was that they did not recognize the importance of understanding the correlation between all three measurements.”

Why They Weren’t Fixed Already

So electronic growth charts stink for both doctors and their patients. Fair enough. Why not just wait for the big medical software firms to fix them? Well that’s exactly what the industry has been doing, to no great progress. Here’s where Dr. Kohane gets really riled up at the whole medical system.

“How is it that we’re comfortable shoving deactivated viruses into people’s blood, or moving embryonic stem cells to build new organs, or using biomaterials to replace joints, and that’s routine!” Kohane rants. “That’s routine innovation in health care, but health care has taken health IT and said, ‘That’s not our problem. We’ll leave it to some high priesthood and they’ll figure it out.'”


Kohane blames the intersection of “monolithic” IT firms, running medical records on “1980s client server systems” with a culture of doctors who, for some reason, don’t demand the same innovation in professional software as they would consumer software. (Though to the industry’s credit, this is all part of a larger trend of enterprise-level software offering substandard interface for no good reason beyond that it’s corporate.)

So the boon in data visualization–something the original growth chart was a bit ahead of the curve on–has been inaccessible to our doctors, who are incapable of running technologies like modern HTML5. These closed systems are also immune to the disruption of the countless developers driving the app market and, to some extent, all of the Apples, Microsofts, and Samsungs driving personal technology forward.

“How come when we work in our professional lives, it’s back to the ’80s,” Kohane asks, “and we go back to our kids at home, we see their phones and the tablets they’re using, and all of the sudden it’s fast forward to 2013?”

A New Approach To An Old Idea

Kohane has given up on changing the industry from within. Instead, SMART teamed up with Fjord to rebuild the growth chart as what it should be today: The best parts of the old analog tool with all the benefits of electronic infrastructure.


Fjord/SMART started by paring down the charts to minimize visual clutter and improve the legibility of curves–garnering feedback from doctors as they went. It was only with this updated visual language in place that Fjord reconsidered interface and interactions (which I’m told is a total opposite from the way the company would normally approach a project but necessary because the established visual tropes of medicine require relatively strict bounds).

With the graphs in place, they enabled some pretty powerful options, like overlapping several different growth chart standards at once (developed by the CDC, WHO, Fenton, and even custom sets). They also use streamlined data inputs to minimize the mathematical errors that paper charts were prone to contain, and backend calculations double check for strange outliers that signal user error. And finally, the app includes a means to translate all this esoteric growth chart data into a clear graphic that families can understand. In other words, a single app is able to visualize information for the doctor to both better understand and better explain the data.

Kohane calls the app–which won’t make its way to the App Store but is free and open for companies to build upon and integrate with–a “statement piece.” It’s meant to disrupt the industry by showing purely what’s possible today.

“The prevailing meme to health-care standards is we have to pay hundreds of millions to large health-care IT vendors, and they’re going to offer safe ways to explore the data,” Kohane says. “Our expertise as clinicians is secondary because this [computer stuff] is a completely new medium! But the fact is, with the right tech, we can make things look like we want them to look. Adoption would be nice, but it’s secondary.”

To learn more about the SMART program, go here.

About the author

Mark Wilson is a senior writer at Fast Company who has written about design, technology, and culture for almost 15 years. His work has appeared at Gizmodo, Kotaku, PopMech, PopSci, Esquire, American Photo and Lucky Peach