Despite blood pressure screenings, hypertension in children is often missed, while other children get evaluated and sometimes treated for high blood pressure readings that turn out to have been transient (often induced by kids’ fear of doctors). That has cardiologists like Justin Zachariah, MD, MPH, concerned.
“We’re both overdiagnosing and underdiagnosing hypertension,” says Zachariah, of the Boston Children’s Hospital Preventive Cardiology Clinic. “There must be a problem in the way we’re measuring it.”
Hypertension, or high blood pressure, is being seen more and more often in kids. Its prevalence 15 years ago was about 1 percent; now it’s nearly 5 percent, according to 2011 data from the American Heart Association, likely due to unhealthy diets and lack of exercise.
That may still not sound like much, but these children are on a dangerous trajectory: an increasing number are hospitalized, and as many as one in five people will have hypertension by the time they hit their 20s and 30s. That sets them up for an increased risk for cardiovascular disease at younger ages—such as a disturbing 30 to 40 percent increase in stroke among 15- to 34-year-olds.
The diagnostic problem has two parts. One part—now being addressed through training programs at Boston Children’s—is how blood pressure (BP) is measured. Ambulatory BP monitoring would eliminate the problem of high readings in the doctor’s office, but insurance companies are reluctant to pay for it. Also, clinicians measure children’s BP in different ways, each having an impact on the reading. Is the child sitting, standing, or lying down when that cuff is inflated? Are his legs dangling? To make matters worse, many busy clinics resort to automated measurements, 60 to 70 percent of which are wrong.
A second problem is recognizing when a BP value is truly out of the normal range, and whether it’s part of a persistent pattern. That’s much harder than it sounds.
“BP cutoff values vary widely in kids, by age, sex and height,” says Zachariah. “As a result, doctors have to consult a very busy chart to determine their percentile.”
To illustrate his point, here’s just a piece of that four-page chart, which is just for the subgroup of 1- to 5-year-old boys:
Zachariah envisioned a software application that could automatically integrate contextual information—age, sex, height—from the patient’s chart, to help clinicians interpret the readings. But since pediatrics is just a small fraction of health care activity and spending, commercial electronic medical records vendors were hesitant to take it on. So Boston Children’s innovated its own solution.
With the blessing of the hospital’s CHAMPS medical records program, Zachariah turned to SMART, a government-funded project that aims to create modular, iPhone-like apps that can be readily shared and used by different health care IT systems. First conceived by Zak Kohane, MD, PhD, and Ken Mandl, MD, MPH, of the Children’s Hospital Informatics Program (CHIP), SMART seeks to shift healthcare IT from (in its words) monolithic, slow-to-evolve systems toward more nimble, shareable technologies.
The SMART team, headed by Josh Mandel, MD of CHIP and fellow software engineer Nikolai Schwertner at Harvard Medical School, got input from a broad range of Boston Children’s clinicians, representing Preventive Cardiology, the Renal Program, Adolescent Medicine, the Optimal Weight for Life (OWL) program, the Emergency Department and the Primary Care Center.
The result was Blood Pressure Centiles. The app not only displays the child’s correct BP percentile, factoring in age, height and sex, but graphs it over time. Color coding reveals at a glance whether individual readings are normal (green), prehypertensive (yellow), hypertensive (red) or hypotensive (blue).
As explained in this tutorial, users can zoom in on a group of readings for more information and apply various filters to help them interpret the measurements—such as looking only at BPs measured in the legs, sitting down or by machine.
A hospital-wide rollout has begun, accompanied by a quality improvement study to see if the app improves recognition of hypertension. Eventually, SMART will make it available to electronic medical record system users in other institutions.
“The solution that has been developed is portable to multiple other platforms, not just our own,” notes Zachariah. “It’s a child-specific solution that many pediatric hospitals have been clamoring for, but that official vendors neglected, presumably because it is not a priority on the adult medicine radar.”