Published online June 22, 2007
PEDIATRICS
Vol. 119
No. 3
March 2007, pp.
e538-e543
(doi:10.1542/peds.2006-1686)
Inaccuracy in Pediatric Outpatient Blood Pressure Measurement
Amber Podoll, MDa,
Michelle Grenier, MDb,
Beth Croix, RNc and
Daniel I. Feig, MD, PhDd
b Cardiology
d Renal Sections
a Department of Pediatrics, Baylor College of Medicine, Houston, Texas
c Renal Section, Department of Nursing, Texas Children’s Hospital, Houston, Texas
 |
ABSTRACT
|
|---|
OBJECTIVE. Hypertension is common in the pediatric population. There is increasing evidence for early hypertensive target organ damage that may lead to substantial long-term morbidity. Because a critical aspect of any screening program for hypertension is the ability to measure blood pressure accurately, we compared typical blood pressure measurements at a vital sign station with those that were obtained following recommendations set forth in "The Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents."
METHODS. We compared the blood pressure measurements that were obtained with standard practice vital sign station screening with those that were obtained by trained personnel in accordance with Fourth Task Force recommendations. A total of 390 children were evaluated at 580 visits to the Pediatric Hypertension Clinic at Texas Children’s Hospital.
RESULTS. Seventy-four percent of the readings were higher at the vital sign station, and only 12% differed by <5 mm Hg for both systolic blood pressure and diastolic blood pressure. The mean difference between vital sign station and examination room was 13.2 ± 8.9 mm Hg for systolic blood pressure and 9.6 ± 7.6 mm Hg for diastolic blood pressure. Multiple regression analyses revealed that age, gender, race, obesity, first versus subsequent visit, essential versus secondary, or white coat hypertension and antihypertensive medications made no statistically significant difference in the lack of correlation of the readings.
CONCLUSION. These results suggest that if pediatricians use vital sign station screening for blood pressure, children with elevated initial measurements must be reevaluated in the examination room.
Key Words: blood pressure • outpatient management • clinical practice
Abbreviations: BP—blood pressure • VSS—vital sign station • EXR—examination room • ABPM—ambulatory BP monitoring • WCH—white coat hypertension • SBP—systolic BP • DBP—diastolic BP
Hypertension affects >30% of the world population and is a primary risk factor for cardiovascular disease, stroke, and premature death.1 Numerous studies demonstrate that the target organ damage can be ameliorated and the risk for future disease can be decreased dramatically by early and effective blood pressure (BP) control. Implicit in any program to identify and evaluate patients for hypertension is the ability to measure BP accurately.
Although less common than in adults, between 2% and 4% of children have hypertension.2 Overt cardiovascular disease and end-stage renal disease secondary to hypertension are very rare in children, but evidence for target organ damage is not.3 Studies of newly hypertensive children demonstrate that as many as 30% have increased left ventricular mass4 and >5% have overt left ventricular hypertrophy.5 Fifteen percent to 20% of hypertensive children have proteinuria,6 and a recent report suggested that hypertensive children have a variety of reversible somatic symptoms.7 Autopsy data suggest that children with traditional cardiovascular risk factors have accelerated atherosclerosis,8 an ominous indication for future disease.
Because of the potential for severe long-term morbidity and the rising prevalence of hypertension in the young, it is becoming increasingly important to screen and evaluate children who are at risk for hypertension. The fundamental component of any screening effort is the act of measuring BP and interpreting its significance. We compared, in a large outpatient population, the standard practices for pediatric BP screening with BP assessment in strict accordance with the best practice guidelines put forth by the Fourth Task Force on the Diagnosis and Management of High Blood Pressure in Children and Adolescents.1
|
METHODS
|
|---|
Patients
We retrospectively analyzed the charts of all children who were referred to the Pediatric Hypertension Clinic at Texas Children’s Hospital between January 2003 and December 2004. A total of 390 children were seen in a total of 580 visits.
Clinical Setting and Staff
Medical assistants were employees of Texas Children’s Hospital and received standard institution training and orientation in clinic care, family relations, and vital sign measurement. This included instruction in BP cuff size selection and appropriate positioning of the patients. The 4 clinic medical assistants each had >2 years of experience at the beginning of the study period, and there was no personnel turnover during the evaluation period. Being a retrospective study, it was not possible to validate the technique of individual technicians before or during the study period.
Vital Sign Station BP Measurement
On children’s arrival at the clinic, a clinic medical assistant assessed children in a vital sign station (VSS) that was adjacent to the waiting room. The clinic routine was to measure height and weight immediately followed by simultaneous measurement of BP and temperature in a chair next to the scale. BP was measured with a Dinamap (GE Healthcare, Fairfield, CT) oscillometric device with the patient seated. The clinic used 6 different Dinamap devices. In January 2003, we used 2 Dinamap 6100 and 2 Dinamap 8100 monitors. During the study period, the 2 Dinamap 6100 monitors routinely were replaced with new Dinamap Pro 400VS monitors because they had been in service for 5 years. The monitors are tested and maintained by the Biomedical Instruments Division at Texas Children’s Hospital. Routine maintenance includes testing and calibration every 3 months. Reported measurements were the mean of 2 upper arm BPs. The medical assistants were instructed in selection of cuff size at the time of their orientation, but there was no independent assessment of the accuracy of their selections. When the pressures differed by >10 mm Hg, a third BP was measured and the reported pressure was the mean of the closer 2 values. BPs of obviously fearful or crying children were not included for the analysis.
Examination Room BP Measurement
Children were seated in a quiet examination room (EXR) for at least 10 minutes before measurement of their BP. All BPs were measured by trained and certified staff and done in accordance with Fourth Task Force recommendations of technique and cuff sizes.1 BP was measured with Welch Allen (Skaneateles Falls, NY) aneroid sphygmomanometers that were calibrated with mercury devices every 2 months. BPs were measured 4 times at initial visits and twice at subsequent visits. When pressures differed by >10 mm Hg, an additional BP was measured and the reported value was the mean of the closer values. BPs of obviously fearful or crying children were not included for the analysis.
Definition of hypertension is in accordance with the "Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents."1 Evaluation and diagnosis of children with hypertension in our program has been described previously.7,9 Children with initial BP in the pre- to stage 1 hypertension range had history; physical examination; urinalysis; renal ultrasound; ambulatory BP monitoring (ABPM; with 90217 monitor [SpaceLabs, Redmond, WA]); and laboratory evaluation that included complete blood count, electrolytes, serum urea nitrogen, creatinine, glucose, uric acid, thyroid function tests, fasting lipids, aldosterone, and direct renin. Echocardiogram was performed in all children with confirmed hypertension. Children with stage 2 hypertension at initial evaluation did not have ABPM, to avoid delay of therapy, and in addition to the studies that were done for stage 1 had renal perfusion scanning for renal scars; noninvasive renal angiography (generally computed tomography); and 24-hour urine collection to screen for pheochromocytoma, Liddle syndrome, apparent mineralocorticoid excess, and glucocorticoid remediable aldosteronism. Children received a diagnosis of essential hypertension when they had confirmed hypertension by Task Force and ABPM criteria10 for stage 1 hypertension and Task Force criteria alone for stage 2 hypertension and the laboratory and imaging tests all were normal. Children received a diagnosis of secondary hypertension when their evaluation yielded a specific cause for their hypertension. Children who had apparent stage 1 hypertension in clinic and had normal BP by ABPM criteria received a diagnosis of white coat hypertension (WCH). A small number of children who were referred to the Hypertension Clinic had BPs that fell into range of prehypertension at initial evaluation. These children had ABPM and were grouped with those who had WCH, if normal BP by ABPM criteria, because all children who were referred to the Hypertension Clinic had been found previously to have >95th percentile BPs on at least 3 consecutive visits to another medical facility.
Statistical Analysis
Primary end points were the difference between VSS systolic BP (SBP) and EXR SBP and the difference between VSS diastolic BP (DBP) and EXR DBP. Secondary end points were the BP differences in the subgroup populations separated by gender, race, age (in 6-year groups), BMI (normal weight, overweight, and obese), and first or subsequent visit. Statistical significance testing of the continuous variables was by Student’s t test. Analysis of variance was used to assess the differences between subgroups and the whole population. Correlation between VSS and EXR BP measurements was assessed by Pearson calculations. Statistical analyses were performed using Statistica 7.0 software (StatSoft Inc, Tulsa, OK).
|
RESULTS
|
|---|
Population Characteristics
The characteristics of the study population are shown in Table 1. The population was a pediatric cohort with an age range from 1 month to 18 years (mean: 12.9 ± 4.2 years). It was similar ethnically to the Houston metropolitan area with the exception of fewer children of Asian descent (35% black, 40% white, 23% Hispanic). There were more boys than girls (228 and 162, respectively), and the diagnoses of essential hypertension, secondary hypertension, and the combination of prehypertension and WCH were represented evenly. Slightly more than half of the population, 53.9%, was overweight (BMI >85th percentile for age), and 25.6% were obese (BMI >95th percentile).
BP Measurements
The BP data for the population is shown in Tables 2 and 3. In Table 2 the SBPs and DBPs shown are the means for all patient BPs at each site, VSS or EXR, whereas the 
SBP and DBP are calculated by subtracting the EXR BP from the VSS BP for each visit and determining the mean of the differences. Because some EXR BPs were greater than VSS BPs and there was wide variability in the difference of the age and the size of the patients in the population, the SBP does not equal the difference of the mean VSS SBP and the mean EXR BP. At the time of the initial visit, 23 (5.9%) of 390 children had VSS SBPs and DBP <95th percentile, 6 (26% of those normotensive at VSS) of whom had SBP or DBP >95th percentile in the EXR. Sixty-six children (17% of the new patient visits, including 17 children who had normal VSS BPs) had EXR SBPs and DBPs <95th percentile. A total of 55 (14.1%) of 390 of the children were misclassified as hypertensive or normotensive and another 81 were misdiagnosed as to stage of hypertension by the VSS techniques.
The mean differences for BPs at VSS and EXR were 13.2 ± 8.9 and 9.6 ± 7.2 mm Hg for SBPs and DBPs, respectively (P > .0001 for both). Subgroups of the population separated by gender, race, age (by 6-year intervals), type of hypertension, body habitus, or first versus subsequent visits revealed SBP ranging from 11.9 to 14.2 mm Hg and DBP ranging from 8.8 to 11.8 mm Hg, none of which was statistically significantly different for the SBP and DBP for the whole population.
Because of the observation that when serial BPs are measured in children the first BP often is somewhat higher,11 we repeated the analysis, excluding the first VSS measurement. Table 3 compares the last VSS measurement with the mean EXR data. The mean difference in SBP and DBP are 12.9 ± 10.9 and 9.5 ± 7.1 mm Hg, respectively, differences that are not statistically different from those using mean VSS measurements. When the subgroups gender, race, age, diagnosis, BMI range, and visit number were compared, the only statistically significant differences between the values in Tables 3 and 2 are that the SBP for white patients and the DBP for black patients were smaller when only the last VSS measurement was used. Both values still were statistically significantly different from EXR measurements (P < .001 for both). The similarity in VSS values likely is because when the 2 VSS measurements were >10 mm Hg different, a third measurement was performed and the closer 2 were averaged (see "Methods"). This practice eliminated many of the high outlier values that often are seen as initial BP measurements.
Analysis of variance for variables that affected SBP revealed statistically significant effects (P < .05) only for VSS SBP and VSS DBP and no significant effect of age, gender, race, diagnosis, laboratory parameters (uric acid, hemoglobin, mean cell volume, creatinine, and potassium), body habitus (height, weight, body surface area, BMI, and BMI percentile), glomerular filtration rate, birth weight, or prescription of antihypertensive medications. Analysis of variance for variables that affected DBP revealed statistically significant effects (P < .05) only for VSS SBP, VSS DBP, and serum potassium (P = .043). The magnitude SBP and DBP was not statistically significantly different for essential or secondary hypertension or WCH.
Correlation of BP Measurements
Figure 1 shows the correlation between VSS and EXR BPs. Figure 1A shows the linear relationship between VSS and EXR SBP with a calculated correlation coefficient of r = 0.7199. There seems to be somewhat less disparity at the highest BPs, but this was not statistically significant for more than or <160 mm Hg. Figure 1B shows the relationship between VSS and EXR DBPs. The correlation coefficient is only r = 0.5947, reflecting a wider disparity between measurements than seen in SBP.
View larger version (32K):
[in this window]
[in a new window]
|
FIGURE 1 SBP (A) and DBP (B) for each patient are compared by scatter plot for each of the 2 measurement settings/techniques. VSS BPs are on the x-axis, and EXR BPs are on the y-axis. The calculated correlation coefficients between the VSS and EXR measurements are shown above each graph.
|
|
|
DISCUSSION
|
|---|
The goal of the study was to compare BPs that were obtained in a typical pediatric practice setting with the more rigorous methods that are used for the establishment of pediatric normative BP data and for pediatric hypertension trials. Because hypertension in children is defined by comparison with existing normative tables, dramatic differences in technique may confuse the diagnosis. The analysis that we performed does not compare 1 device with another or 1 operator with another but rather the BP measurements that are obtained by standard practice of nonspecifically trained medical assistants using an oscillometric device at a central VSS with those that are obtained by rigorous adherence to Task Force recommendations. The latter practice is comparable to the measurements that are used to generate the pediatric normative data that are used for the diagnosis of hypertension in children.
The magnitude of the differences in VSS and EXR measurements is large. With a mean difference >13 mm Hg for SBP and 9 mm Hg for DBP, the technique differences can lead to substantial misdiagnosis: normotensive children being identified as hypertensive or children with stage 1 hypertensive being identified as having stage 2. In most children, the VSS practices led to an overestimation; however, application of a correction factor, such as subtraction of 13 from SBP, would be unwise because 24% of VSS BPs actually were lower than EXR values. Furthermore, the variance of the differences between measurement techniques was large, rendering correction factors useless.
The degree of difference between VSS and EXR measurements was preserved through all analyzed subgroups. Gender, race, age, and degree of obesity had no significant effect on the differences in SBP or DBP. The lack of impact of age or BMI suggests that incorrect cuff size use is unlikely to explain sufficiently systematic overestimation of BP. We also did not see evidence for an exaggerated WCH effect at the VSS because children with WCH had similar differences in SBP and DBP to all other groups.
This study has several important limitations. First, it was performed at a single center. That it was performed in a hypertension clinic, where there is an emphasis on the measurement of BPs, might be expected to increase the consistency of VSS measurements compared with clinics in which BP is measured less regularly. Second, we did not control for cuff size selection. It theoretically is possible that systematic use of smaller cuffs at the VSS contributes to the differences seen; however, this requires that the 4 medical assistants were consistent in such errors during the 2-year study period. If this were to be the case, then it further emphasizes the importance of training and technique in the ascertainment of pediatric BP data. The third limitation is that the VSS BPs were always done first, rather than having randomized order to the BP techniques. A decline in BP values with repeated measurements was reported previously in children,11 and this may account for some of the differences that were seen between VSS and EXR measurements. This possibility emphasizes the importance of rechecking BPs with rigorous technique rather than relying on the more efficient VSS-type methods.
Because hypertension in the pediatric population is associated with significant target organ damage and morbidity, accurate BP screening and appropriate diagnostic evaluation is critical. The current recommendations are based on the careful use of aneroid devices, with appropriate cuff size on children who are relaxed and in a seated position for at least several minutes. Currently, if BP is measured at all, then the use of automated oscillometric devices at a VSS is a common practice that is efficient but inaccurate. Ideally, all pediatric groups would conform to task force recommendations, but at the very least, children with elevated VSS BPs should be reevaluated in the EXR by trained personnel. These results should not be interpreted as a reassurance that elevated BP that is detected in pediatric screening visits should be assumed to be normal. It is true that common screening practices will overestimate the BP; however, epidemiologic studies indicate that 2% to 4% of the pediatric population2 and between 15% and 30% of obese children have hypertension.12–15 Children need to be screened and assessed accurately for treatment to prevent significant long-term morbidity. An awareness that differences exist in routine oscillometric BP measurements that are taken on entrance to a VSS and aneroid BP values that are obtained in a quiet setting should decrease error in pediatric BP assessment.
|
ACKNOWLEDGMENTS
|
|---|
This work was supported by National Institutes of Health grants DK-064587 and DK-071223 (to Dr Feig).
|
FOOTNOTES
|
|---|
Accepted Oct 10, 2006.
Address correspondence to Daniel I. Feig, MD, PhD, Department of Pediatrics, Renal Section, Baylor College of Medicine, 1102 Bates Ave, MC3-2482, Houston, TX 77030. E-mail: dfeig{at}bcm.tmc.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES
|
|---|
- The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents.
Pediatrics. 2004;114(suppl)
:555
–576[Free Full Text]
- Sorof JM, Turner J, Martin DS, et al. Cardiovascular risk factors and sequelae in hypertensive children identified by referral versus school-based screening.
Hypertension. 2004;43
:214
–218[Abstract/Free Full Text]
- Cuspidi C, Macca G, Salerno M, et al. Evaluation of target organ damage in arterial hypertension: which role for qualitative funduscopic examination?
Ital Heart J. 2001;2
:702
–706[Medline]
- Pontremoli R, Nicolella C, Viazzi F, et al. Microalbuminuria is an early marker of target organ damage in essential hypertension.
Am J Hypertens. 1998;11(pt 1)
:430
–438
- Daniels SR. Hypertension-induced cardiac damage in children and adolescents.
Blood Press Monit. 1999;4
:165
–170[Medline]
- Daniels SR, Meyer RA, Strife CF, Lipman M, Loggie JM. Distribution of target-organ abnormalities by race and sex in children with essential hypertension.
J Hum Hypertens. 1990;4
:103
–104[ISI][Medline]
- Croix B, Feig DI. Childhood hypertension is not a silent disease.
Pediatr Nephrol. 2006;21
:527
–532[CrossRef][ISI][Medline]
- Berenson GS, Srinivasan SR, Hunter SM, et al. Risk factors in early life as predictors of adult heart disease: the Bogalusa Heart Study.
Am J Med Sci. 1989;298
:141
–151[ISI][Medline]
- Feig DI, Johnson RJ. Hyperuricemia in childhood primary hypertension.
Hypertension. 2003;42
:247
–252[Abstract/Free Full Text]
- Soergel M, Kirschstein M, Busch C, et al. Oscillometric twenty-four-hour ambulatory blood pressure values in healthy children and adolescents: a multicenter trial including 1141 subjects.
J Pediatr. 1997;130
:178
–184[CrossRef][ISI][Medline]
- Sorof JM, Poffenbarger T, Franco K, Portman R. Evaluation of white coat hypertension in children: importance of the definitions of normal ambulatory blood pressure and the severity of casual hypertension.
Am J Hypertens. 2001;14(pt 1)
:855
–860
- Boyd G, Koenigsberg J, Falkner B, Gidding SS, Hassink S. Effect of obesity and high blood pressure on plasma lipid levels in children and adolescents.
Pediatrics. 2005;116
:442
–446[Abstract/Free Full Text]
- Manzoli L, Ripari P, Rotolo S, et al. Prevalence of obesity, overweight and hypertension in children and adolescents from Abruzzo, Italy.
Ann Ig. 2005;17
:419
–431[Medline]
- Falkner B, Gidding SS, Ramirez-Garnica G, Wiltrout S, West D, Rappaport E. The relationship of body mass index and blood pressure in primary care pediatric patients.
J Pediatr. 2006;148
:195
–200[CrossRef][ISI][Medline]
- Angelopoulos P, Milionis H, Moschonis G, Manios Y. Relations between obesity and hypertension: preliminary data from a cross-sectional study in primary schoolchildren—The Children Study.
Eur J Clin Nutr. 2006;60
:1226
–1234[CrossRef][ISI][Medline]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics
TOP
ABSTRACT
METHODS
