Blood Pressure Screening in Children: Do We Have This Right?
Article Outline
The manuscript by Stewart et al,1 published in this issue of The Journal, reports that although blood pressure measurements in children at the time of emergency department visits for a nonurgent condition were often elevated (26%), follow-up did not uncover a single case of “true” hypertension. The authors conclude, “The yield of measuring blood pressure at triage to the emergency department in children with nonurgent problems appears to be extremely low.” Hypotension, certainly a concern in the emergency department, is predicted by other findings, and true hypertension was not diagnosed in the population studied. In fact, a large number of false-positive results resulted from the triage measurements performed.
See related article, p 478
The report actually may draw attention to an even more important question, which goes to the heart of screening for disease. Is screening for hypertension in childhood, the way it is recommended today, a useful individual health (or even public health) strategy?
Provocative? Perhaps, but let's explore why the question may be worthy of debate. Pediatricians, as much as any physician group, are familiar with screening. Almost every newborn in the United States undergoes screening for a variety of conditions with a drop of blood obtained soon after birth. The decision to include screening for a condition is based on a number of well described criteria.2, 3 The goal of screening should be to (1) identify a previously unknown or undiagnosed condition in an apparently healthy or symptom-free patient; (2) offer presymptomatic intervention because such intervention will reduce the likelihood of disease development, progression, morbidity, or death; (3) offer counseling for risk; or (4) perform research. Perhaps with the exception of research, the reason to screen also involves the principle that with presymptomatic diagnosis there is indeed an effective treatment or option available to the patient.
The consideration of whether a screening test should be used is based on a number of principles. The test should be able to distinguish those with the condition from those who do not have it. The sensitivity should be high, thereby detecting those who truly have the condition (high positive predictive value). The specificity should be high, thereby excluding those who do not have it (high negative predictive value). Furthermore, a good screening test is accurate, reproducible, easy to perform in the field, and preferably inexpensive. Not every screening test meets all these criteria, but the risk benefit ratio of performing a screening test should undergo this type of evaluation.
The 2004 National Blood Pressure Education Program (NHBPEP) Working Group 4th Report recommends that blood pressure measurements be performed in all children over 3 years of age when these children are seen in a medical setting.4 The fundamental tenet of such a recommendation is that emergency departments or outpatient clinics represent the chance to measure blood pressure and thus find children who are presymptomatic patients with a treatable condition. Furthermore, there is clear evidence that the risk to health of hypertension increases with time leading to known sequelae of hypertension such as cardiac, renal, and central nervous system injury. Does the NHBPEP Working Group 4th Report recommendation meet the criteria for a good screening test?
Using the criteria above, measuring blood pressure is the way we use to determine whether a patient has hypertension. It must be recognized that, unlike may other conditions, it may be necessary to keep measuring because the early changes in blood pressure may occur at almost any age. This is distinctly different from many conditions where the screening test, when it is performed, can immediately be followed by the definitive test. If a test result is negative, further testing later in life is not necessary. In this regard, hypertension screening is more like colon cancer or breast cancer screening and not like cystic fibrosis or metabolic disease screening. Also, there are interventions that can be used that are effective. However, the prevalence of hypertension in the pediatric population is not high when screening the whole population is the recommendation. This makes a high positive predictive value difficult to achieve. In the study by Stewart et al, the positive predictive value was 0! What about the test method itself? Are blood pressure measurements as we do them routinely accurate? Increasingly, automated blood pressure machines are employed in all settings. Although difficult to assess, reports have demonstrated that automated blood pressure measurements overestimate both systolic and diastolic blood pressures.5, 6 This will further reduce the positive predictive value, but it is assumed to assure that all patients with true hypertension will have a positive screening result. While, as Stewart et al point out, this “assumption is implicit in the NHBPEP guidelines,” this remains unstudied. Is the method of screening reproducible? This too is unstudied. Will the blood pressure measurement performed at an emergency department or an outpatient department be similar if performed at a different site on the same patient? We do not know. We do know that blood pressures in the same patient do vary depending on time of day and a variety of influences on the patient. Concepts such as white coat hypertension or masked hypertension (findings consistent with hypertension when using ambulatory blood pressure measurements which are not measured by a single random measurement) further put the screening approach in doubt by potentially further reducing the positive predictive value (white coat hypertension) or by reducing the negative predictive value (masked hypertension).7, 8
Well, at least the cost of the screening test is not high, correct? A single blood pressure measurement with an automated machine especially in a patient being seen for other reasons probably does not add much cost. The exact cost is not well studied and the cost of further screening and evaluation should be factored into the cost of screening. This too remains unstudied. Indeed, the increasing use of ambulatory blood pressure measurement because of accuracy may become an important factor when considering screening, diagnosis and follow up for hypertension.
Do we have a good strategy for screening children for hypertension? Maybe because blood pressure measurements are common and “easy” to perform, it seemed a good strategy to screen in the outpatient or emergency department. Furthermore, the recommendations were predicated on the notion that the approach was the best chance at being able to screen children. But is this good policy? Not necessarily. The NHBPEP guidelines should undergo more thorough testing. Other approaches should be analyzed such as screening populations as risk (mother and father with hypertension, for example). The Stewart et al study is “a single center study, conducted over a brief interval. The results should be generalized with caution.” But recommending screening of an entire population is large undertaking and deserves careful analysis including a careful determination of the risk/benefit ratio. There are costs to undiagnosed hypertension but we should be very certain of the costs of screening for hypertension as well. Such an analysis might help us do what is best for children. It is time to reconsider this screening approach.
References
- . The value of routine blood pressure measurement in children presenting to the emergency department with non-urgent problems. J Pediatr. 2008;153:478–483
- . Principles of health screening. In: Wallace HM, Gould EM, Oglesby AC editor. Maternal and child health practices: problems, resources and methods of delivery. 2nd ed.. New York: John Wiley and Sons; 1975;p. 569–576
- Presidents Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research, 1983, screening and counseling for genetic conditions. Washington DC. Government Printing Office.
- . The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents pediatrics. Pediatrics. 2004;114(Suppl):555–576
- . Comparison of auscultatory and oscillometric blood pressures. Arch Pediatr Adolesc Med. 2001;155:50–53
- . Inaccuracy in pediatric outpatient blood pressures measurements. Pediatrics. 2007;119:E538–E543
- . How common is white coat hypertension?. JAMA. 1988;259:225–228
- . Masked hypertension. Hypertension. 2002;40:795–796
PII: S0022-3476(08)00401-0
doi:10.1016/j.jpeds.2008.05.014
© 2008 Mosby, Inc. All rights reserved.
Refers to article:
- The Value of Routine Blood Pressure Measurement in Children Presenting to the Emergency Department with Nonurgent Problems , 27 May 2008
