Margaret Riley, MD; Margaret Dobson, MD; Ananda Sen, PhD; Lee Green, MD, MPH Department of Family Medicine, University of Michigan, Ann Arbor (Drs. Riley, Dobson, and Sen); Department of Family Medicine, University of Alberta, Canada (Dr. Green) [email protected]. edu The authors reported no potential conflict of interest relevant to this article. Partial data were presented at the Spring Conference of the Society of Teachers of Family Medicine, April 2012, in Seattle, Wash.

Original Research

Recognizing elevated BP in children and adolescents: How are we doing? Not well, according to this study. But clues to practice deficiencies suggest remedies.

Abstract Purpose c Hypertension is increasing in in­ cidence in children and adolescents, but may go unrecognized by health care providers. This study assessed rates of recognition of abnormal blood pressure (BP) values in pa­ tients ages 3 to 18 years by family medicine attending physicians, resident physicians, and nurse practitioners/physician assistants. The study also explored provider knowledge and comfort with diagnosing hypertension in children. Methods c We conducted a chart review of pediatric patient visits in family medicine out­ patient clinics, in addition to a survey of fam­ ily medicine providers. Results c Providers recognized only 8% of el­ evated BP values during pediatric clinic visits. They were more likely to recognize BP values in the hypertensive range than in the pre­ hypertensive range (P30% in boys and 23% to 30% in girls.2,3 Childhood hypertension can cause target organ damage, including left ventricular hypertrophy and pathologic vascular changes such as increased carotid artery thickness.4,5 One study found left ventricular hypertrophy in 41% of children and adolescents with newly diagnosed hypertension.4 Hypertension in children is strongly associated with hypertension in adulthood, and adult hypertension is the leading cause of premature death worldwide.6,7 Early recognition and treatment of hy-

The Journal of Family Prac tice | JUNE 2 0 1 3 | V o l 6 2 , No 6

pertension in children and adolescents can reduce significant long-term health risks for these individuals. The American Academy of Pediatrics and the National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents recommend checking blood pressure (BP) at every office visit for any child ≥3 years of age.8 The US Preventive Services Task Force is currently updating its recommendation on checking BP in children. Normal BP values in children depend on the child’s age, sex, and height, and are available in tables from the National Institutes of Health.9 Hypertension in children is defined as BP in the 95th percentile or higher for the child’s age, sex, and height measured on 3 separate occasions. Prehypertension is BP in the 90th to the 95th percentile, or ≥120/80 mm Hg.8 Because BP values in children and adolescents depend on their body size, providers may not easily recognize when a BP value is abnormally high. In one study, pediatricians recognized and duly documented hypertension in an electronic medical record (EMR) for only 26% of children and adolescents whose BP values met diagnostic criteria.1 We wanted to determine whether family medicine attending physicians, resident physicians, nurse practitioners (NPs), and physician assistants (PAs) practicing in an academic center were recognizing elevated BP values in children and adolescents during health maintenance exams (HMEs) and other office visits. We also wanted to evaluate providers’ baseline knowledge of, and comfort with, the diagnosis of hypertension in children and adolescents. We hypothesized that providers are likely missing elevated BP values during outpatient clinic visits, and that this may be due to a knowledge gap, in addition to clinic systems issues.

METHODS In January 2012, the University of Michigan Internal Review Board granted this study exemption status as a quality improvement project. We identified charts of children ages 3 to 18 years seen at our 5 family medicine clinic sites during 2011.

Our EMR has a tab in which the medical assistant enters vital signs. If height and BP are both entered at a clinic visit, the EMR automatically calculates BP percentile and lists it under that tab, which is separate from the tab a provider regularly uses. The EMR does not highlight or flag a BP percentile that is abnormal. We identified pediatric visits in which height and BP had both been recorded, and indicated those with a documented elevated BP value above the 90th percentile for age, sex, and height. We also identified visits at which height was not recorded but BP was ≥120/80 mm Hg. We reviewed associated notes to see if providers had recognized abnormal values. If a provider noted and repeated the BP measurement, noted and gave an explanation for the elevated BP (eg, actively crying, febrile, in pain), or noted and planned to recheck the BP at a follow-up visit, we classified the provider as having recognized the abnormal value. We further classified elevated BP values as prehypertensive, hypertensive, or elevated (BP ≥120/80 mm Hg with no associated height recorded, making the BP percentile undeterminable). We classified clinic visits as either HMEs (well-child check-ups and sports physicals) or office visits. We classified providers according to their level of training—attending physician, resident physician, or NP/PA. For purposes of statistical analysis, we grouped patient age as younger children (ages 3-11) and adolescents (ages 12-18). Using logistic regression analysis, we assessed the likelihood of a patient’s abnormal BP value being recognized as a function of level of BP elevation, provider level of training, visit type, and age classification. We surveyed family medicine providers anonymously on their basic knowledge of how prehypertension and hypertension are diagnosed in children and adolescents, and on their comfort level in making the diagnosis in this age group (TABLE 1 ). We analyzed the number of knowledge questions answered correctly relative to provider comfort and level of training, using a one-way analysis of variance. We used Fisher’s exact test to examine the association between level of training and level of comfort. We set the cutoff value for statistical significance at P=.05.

Our EMR automatically calculates BP percentile, but it doesn’t highlight or flag a percentile that is abnormal.

C o nti n u e d

jfponline.com

Vol 62, No 6 | JUNE 2013 | The Journal of Family Practice

295

RESULTS

Unlike fixed cutoff points for adults, values that indicate elevated BP in children and adolescents vary with age, sex, and height.

296

During 2011, our 5 University of Michigan family medicine clinics saw 10,769 children and adolescents, of which 8934 (83%) had their BP recorded. Of those with a BP recorded, 3807 (43%) had no height recorded; therefore, we could not determine the BP percentile. In those visits with a BP recorded, 1009 (11.3%) measurements qualified as elevated (above the 90th percentile for age, sex, and height, or ≥120/80 mm Hg with no height recorded). In only 81 (8%) of the visits with an elevated BP reading did the provider document the abnormal BP value and indicate plans to recheck it at a follow-up visit. Participating in these clinic visits were 57 physician faculty members, 40 resident physicians, and 4 NP/PAs. Providers were 6.5 times more likely to recognize hypertensive-range BP values than prehypertensive values (odds ratio [OR], 6.54; 95% confidence interval [CI], 3.53–12.11; P