Evaluation of Pediatric Patients With Mild-to-Moderate Hypertension: Yield of Diagnostic Testing
OBJECTIVE. Children and adolescents with newly diagnosed hypertension undergo various tests to define the cause and target organ consequences of the elevated blood pressure. We tested the hypothesis that the diagnostic yield of individual components of the currently recommended assessment does not justify performance for all patients with mild-to-moderate hypertension.
METHODS. A retrospective chart review was conducted of patients who were referred between July 2002 and June 2007 for mild-to-moderate hypertension, defined as maximum blood pressure at ≥95% + 20/10 mmHg. The assessment included history and physical examination, nutritional assessment, urinalysis, biochemical and fasting lipid profile, renal ultrasound, echocardiogram, and 24-hour ambulatory blood pressure monitoring.
RESULTS. A total of 249 patients were identified, and charts for 220 (88%) were available for review. There were 156 boys and 64 girls aged 13.3 ± 4.4 years. BMI was 26.1 ± 6.7 kg/m2, and 143 (65%) had a BMI of ≥90%. Results of urinalysis and serum biochemical testing were clinically normal in all cases. Among those with a lipid profile, 59 (42%) had total cholesterol values of >170 mg/dL, and 26 (19%) had severe hypercholesterolemia (>200 mg/dL). Renal sonography revealed findings plausibly associated with hypertension in 14 (8%) patients; 4 (2%) had renovascular abnormalities. Yield of echocardiography was 17%. On ambulatory blood pressure monitoring, 47 (60%) children had systolic readings of >95% at least 20% of the time, and 28 (36%) had diastolic readings of >95% at least 20% of the time.
CONCLUSIONS. For children and adolescents with mild-to-moderate hypertension, on the basis of a cutoff of 5% to 20% abnormal results to define a useful test, the initial evaluation can range from a serum cholesterol level and ambulatory blood pressure monitoring to a panel that consists of a fasting lipid profile, renal ultrasound, echocardiogram, and ambulatory blood pressure monitoring. Additional assessment should be guided by specific clinical features and the nature of the patient population.
- essential hypertension
- ambulatory blood pressure monitoring
- lipid profile
- metabolic syndrome
The cause and natural history of hypertension vary in different patient populations. Whereas essential hypertension accounts for the vast majority of cases in adults, children with an elevated blood pressure (BP) are considered more likely to have an underlying disease.1–10 As such, the recommended evaluation of pediatric hypertension involves a comprehensive diagnostic panel aimed at revealing the cause and target organ consequences of the elevated BP1; however, essential hypertension is more prevalent than previous studies suggested, a trend that is attributable in part to more widespread measurement of BP in pediatric patients and in part to a rise in obesity and the metabolic syndrome in children.11–13
Different diagnostic protocols have been drafted for children with hypertension.1,2,4–10 The most recent statement from the National High Blood Pressure Education Program recommended that the following tests be performed for all children with BP at ≥95th percentile: complete blood count, serum urea nitrogen (SUN), creatinine, electrolytes, fasting lipid panel and glucose, urinalysis, urine culture, renal ultrasound, echocardiogram, and retinal examination. Twenty-four-hour ambulatory BP monitoring (ABPM) is advised under select circumstances. Additional evaluation for very young children with stage 1 hypertension and for a child or adolescent with stage 2 hypertension should include obtaining plasma renin activity, imaging of the renal vasculature, and plasma and urine steroid and catecholamine levels.1 We are unaware of any previous studies that assessed the diagnostic yield of this costly battery of tests, especially for pediatric patients with mild-to-moderate hypertension, who are generally healthy; therefore, we performed a retrospective chart review of children who were referred for evaluation of mild-to-moderate BP elevation to test the hypothesis that several components of the comprehensive test battery do not have a high enough yield to warrant performance for all children under these circumstances.
This retrospective chart review was conducted for all patients who were evaluated in the Pediatric Hypertension Center at Schneider Children's Hospital during the period from July 2002 through June 2007. A database maintained in the Division of Nephrology, which includes patients' gender, age, and initial diagnosis/complaint, was scanned to identify children and adolescents who were in the diagnostic category of hypertension and were seen during the defined time interval. Ethnicity was not recorded in the clinic chart or in the divisional database.
The Pediatric Hypertension Center is a multidisciplinary diagnostic module that convenes monthly. It is designed to expedite the workup of children and adolescents with mild-to-moderate hypertension, defined as a maximum BP at >95% level + 20/10 mmHg. Patients with more severe hypertension, acute complaints, or evidence of significant kidney disease are evaluated promptly in the nephrology clinic. The evaluation includes a complete history and physical examination, nutritional assessment, urinalysis, fasting complete metabolic and lipid profiles, renal ultrasound, echocardiogram, and 24-hour ABPM in cooperative children who are older than 10 years.
Medical charts were culled from active files or retrieved from off-site storage. The following data, when available, were recorded: date of initial visit; age; gender; family history of hypertension; symptoms; height (cm); weight (kg); BMI (kg/m2); BP in the left arm, right arm, and leg; urinalysis; serum potassium; bicarbonate; SUN; creatinine and total cholesterol concentrations; estimated glomerular filtration rate (eGFR) by using the height-serum creatinine formula appropriate for age and gender14; renal sonogram results; echocardiogram findings; systolic and diastolic means (24 hour, day, and night); percentage of systolic (SBP) and diastolic BP (DBP) readings at >95th percentile in the ABPM; and treatment prescribed at the initial visit. Because of age and technical problems, testing was not complete for all patients.
All information was entered on a preapproved form. The data collection sheet contained a number assigned to each patient for the chart review and the elements enumerated in the previous paragraph. A data link sheet containing patient name, gender, age, date of initial visit, and assigned study number was created and stored in the Division of Nephrology. The deidentified data collection and storage methods were in accordance with Health Insurance Portability and Accountability Act guidelines. The protocol was approved by the institutional review board of North Shore-Long Island Jewish Health System.
Continuous data that were normally distributed are presented as means ± SD and were analyzed by the Student's t test. The χ2 and Fisher's exact tests were used to evaluate differences in proportions between patient subgroups. Differences were considered statistically significant at P < .05.
During the study period, 249 patients were seen in the Pediatric Hypertension Center, and the medical charts were available for 220 (88%) of them. The demographics, clinical features, and laboratory results are summarized in Table 1. The cohort was composed predominantly of male patients, and the majority (77%) were adolescents between ages 12 and 21 years. The mean weight percentile was 60 ± 33, and 51% had a weight at ≥90th percentile. Similarly, 55 (26%) of the patients had a BMI of >30, and 65% had a BMI percentile at ≥90th.
Forty-one (19%) patients reported symptoms, the most common complaint being headache (n = 14), followed by chest pain and fatigue (n = 4). Other symptoms such as syncope, dizziness, migraines, shortness of breath, numbness, and abdominal pain were very infrequent, affecting 1 to 2 patients. A family history of essential hypertension was noted for 70% of the children.
The urinalysis was negative for hematuria in 201 (96%) of 210 patients. No proteinuria was detectable on the dipstick urine testing for 172 (82%) patients; follow-up quantification of proteinuria in an early morning urine specimen was normal for all patients with a positive dipstick or initially high quantitative result (Table 2).
Serum potassium (≤5.7 mmol/L), SUN, and creatinine concentrations and eGFR were normal in all cases on the basis of age-appropriate values. The serum bicarbonate level was elevated (≥30 mmol/L) in 6 cases; however, it was not associated with hypokalemia (<3.5 mmol/L) in any case and was therefore not considered clinically relevant. The mean total cholesterol concentration in the patients who had a fasting lipid profile performed was 168 ± 33 mg/dL; 59 (42%) had mild hypercholesterolemia >170 mg/dL, and 26 (19%) had severe hypercholesterolemia >200 mg/dL (Table 2). In a follow-up survey of 38 patients who were older than 10 years and were still actively followed, 17 (45%) had mild hypercholesterolemia. In addition, 7 (18%) had hypertriglyceridemia (>150 mg/dL) and 9 had low high-density lipoprotein cholesterol (<40 mg/dL in male and <50 mg/dL in female patients). These 2 tests identified only 4 patients who would not have been detected solely on the basis of measurement of the total cholesterol level.
Renal sonography revealed findings that might be associated with hypertension in 14 (8%) patients. Four (2%) children had renovascular abnormalities. Other abnormalities included solitary renal cysts (n = 3), hydronephrosis (n = 2), and altered echogenicity (n = 2) (Table 2).
Nineteen (11%) patients had left ventricular hypertrophy (LVH), and 10 (6%) had other significant abnormalities, including 4 cases of aortic coarctation, 2 with a dilated aortic root, and 1 instance each of Takaysu arteritis, dilated left ventricle, hyperdynamic left ventricle, and hypertrophic cardiomyopathy. In total, significant abnormalities, including LVH, were found in 29 (17%) patients who had an echocardiogram performed (Table 2).
The 24-hr mean BP for patients who underwent ABPM was 124 ± 11/71 ± 9 mmHg. The mean value during waking hours was 125 ± 13/72 ± 10 mmHg and 123 ± 13/69 ± 10 mmHg during night hours. Forty-seven (60%) patients had systolic ABPM measurements at ≥95th percentile (BP load) at least 20% of the time, and 28 (36%) had diastolic measurements ≥95th percentile at least 20% of time. A total of 33% of patients had both SBP and DBP measurements at ≥95th percentile at least 20% of time (Table 2).
The yield of abnormal results from the various diagnostic tests is summarized in Table 2. A total of 100 patients had at least 1 abnormal test. Of these, 73 had a single abnormal result, which suggests that routine implementation of a broad panel of tests will yield a large majority of normal findings. The results were similar in all age groups. Thus, there were no differences in the percentage of abnormal test results for children <12 versus those ≥12 years of age, except for a higher incidence of mild hypercholesterolemia and an elevated diastolic BP in ABPM in the younger group (Table 3). Moreover, when the younger age group was further divided into 2 subgroups, 0 to 5 and 6 to 11 years of age, the findings were similar (data not shown). Finally, the yield of the diagnostic test battery was not influenced by the presence of clinical symptoms (Table 4).
This retrospective chart review is one of the first studies to evaluate the yield of individual components of the recommended diagnostic battery applied to the outpatient assessment of mild-to-moderate hypertension in children and adolescents. The high percentage of children with a positive family history of hypertension (70%) and the low frequency of test results indicating a secondary cause of high BP suggest that essential hypertension is very prevalent in the pediatric population. This renders the performance of a comprehensive diagnostic battery for children who are clinically well and who have mild-to-moderate hypertension less urgent. There is precedent for assessment of the cost-effectiveness of the workup of common nephrology problems in children, namely, hematuria.15 The rising incidence of obesity and of hypertension and obesity in pediatric patients makes our study timely. Although the report is a retrospective analysis, this type of assessment is a mandatory first step in the development of sound guidelines that can be applied prospectively in clinical practice. In the following paragraphs, we address each component of the workup: serum and urine chemistry, lipid profile, renal ultrasound, echocardiogram, and ABPM.
Serum electrolyte levels were virtually normal in all patients. Measuring serum potassium level in children with hypertension is advised because it may be altered with adrenal or kidney disease1,4; however, only 3 patients had potassium levels of >5.1 mmol/L (n = 152), and none had levels of >5.7 mmol/L or hypokalemia. Furthermore, SUN, serum creatinine concentration, and eGFR were consistently within normal limits. Similarly, a urinalysis performed by a pediatric nephrologist, including determination of protein excretion in an early morning sample, is likely to be normal in all cases. The initially false-positive urine protein measurements underscore the problem incurred by excessive laboratory testing for children with mild-to-moderate hypertension. Abnormal results necessitate repeat determinations, which heighten the financial and psychological costs of the workup. There is still uncertainty about the predictive value of microalbuminuria for cardiovascular risk in adults.16 Thus, we suggest that measurement of microalbuminuria for children with new-onset hypertension be deferred until there are data to confirm its clinical utility. On the basis of our results, routine urinalysis and serum biochemical profile do not have diagnostic utility for a child with mild-to-moderate hypertension.
The only blood test that had a significant yield of abnormal results was the fasting lipid profile; 42% of the patients had a total cholesterol level of ≥170 mg/dL, and 19% had a level of ≥200 mg/dL. The correlation between high levels of serum cholesterol and an elevated BP is consistent with previous studies.13 There is a linkage among hypertension, hypercholesterolemia, and obesity as part of the metabolic syndrome in children.11–13 Our findings, in which 65% of the patients had a BMI of ≥90 and 26% had a BMI of >30 kg/m2, concur with this trend. There is no definition of the metabolic syndrome in children that is uniformly accepted in the United States.17 The International Diabetes Foundation has published criteria that include normative values for weight, BP, triglyceride, high-density lipoprotein cholesterol, and glucose.18 Our report focuses on weight and cholesterol in conjunction with hypertension. It was impractical to measure glucose or insulin levels under defined conditions in the Pediatric Hypertension Center. A follow-up survey of our sample confirmed that determination of total cholesterol alone detected 17 (81%) of 21 patients with a serum lipid abnormality; therefore, assessment of BP weight, and total cholesterol should suffice to detect most cases of metabolic syndrome in children with mild-to-moderate hypertension.
Renal sonography has been recommended for all pediatric patients with essential hypertension to diagnose parenchymal and renovascular disease. The latter category has a reported incidence in hypertensive patients of up to 10%.4 In our study, 14 (8%) patients had abnormalities on their renal ultrasound, and 4 patients (2%) had findings indicative of renovascular disease. The ultrasound results in the other 10 cases did not influence patient treatment. There were no cases of polycystic kidney disease or reflux nephropathy; therefore, it does not seem that renal ultrasound is a high-yield screening test for all children with hypertension. It could be reserved for patients with abnormal ABPM results or in situations in which there is a high degree of clinical suspicion of renovascular hypertension or renal parenchymal disease.
Echocardiograms detect aortic coarctation, an important cause of hypertension, and LVH, a marker of cardiac structural change secondary to hypertension. In our sample, 17% of the patients who underwent echocardiography had an abnormality. Echocardiography has been recommended to screen specifically for LVH, a surrogate marker of long-term cardiovascular prognosis for patients with hypertension.19 The incidence of LVH in children with hypertension has been reported to be as high as 40% to 45%.20,21 In our patients, the incidence of LVH was lower: 11% of the 178 who had the test performed. There is no proven link between LVH and clinical outcomes in children with essential hypertension. Nevertheless, we recommend use of echocardiography as a screening modality to diagnose aortic coarctation and to detect LVH. In addition, regression of LVH may be helpful as a marker of successful treatment of hypertension.
The utility of ABPM is a topic of ongoing debate. In some studies, it is considered “a research tool…in select high-risk patients10” and in others “very useful in the evaluation of hypertension in children.”1 The debate is whether ABPM improves the accuracy of the diagnosis of hypertension in children. ABPM was very useful, because 60% of children who were evaluated had systolic hypertension (>95% for age and gender) in the ABPM record ≥20% of the time and an elevated DBP ≥20% of the time in 36% of patients. It is worth noting that 33% of the patients had both abnormal SBP and DBP load by ABPM. This was very similar to the rate of hypertension, 30 (38%) of 78 patients, on the basis of mean 24-hour ABPM levels defined by Soergel et al22; therefore, ABPM should be included in the initial battery of tests in the diagnostic evaluation of pediatric hypertension to help guide decision-making in the need for additional evaluation and treatment.
Retinal examination was not included in our battery of tests. There are reports of relatively frequent abnormalities in young patients with hypertension (6–15 years)23,24; however, most abnormalities were low grade, and severe changes are more common in children with markedly elevated BP levels.25 In addition, recent reports did not list retinal examination as an integral part of the first-level evaluation of children with newly diagnosed hypertension.26 Absent new findings pertaining to this issue, we think that it is reasonable to defer a retinal examination for patients with mild-to-moderate hypertension.
Our findings should be applicable to all pediatric patients with mild-to-moderate hypertension, because the data did not show a difference in the percentage of abnormalities in any particular age group. The higher incidence of ABPM abnormalities in younger children may reflect increased anxiety provoked by this procedure compared with adolescents. We cannot explain the higher incidence of hypercholesterolemia in the children who were younger than 12 years. It does not seem that symptomatic children have a higher prevalence of diagnostic abnormalities than children with asymptomatic hypertension. Thus, our findings about the yield of diagnostic tests are germane to all healthy children who present with mild-to-moderate hypertension.
There are a number of limitations to our study. As a retrospective review rather than a prospective clinical investigation, not all of the examinations were performed in a uniform manner, and patients were not treated with an identical comprehensive workup. In addition, patients were often missing specific tests or aspects of the examination, which were either not performed or not recorded; however, this inconsistency is likely to introduce a bias against the study hypothesis (ie, in favor of higher diagnostic yields of performed tests) because tests were more likely not to be performed in cases of low-grade clinical suspicion of significant hypertension. Second, the patients were not seen in a primary care setting and instead were referred to a tertiary care center for evaluation; however, again, this would tend to introduce a bias against our hypothesis that extensive testing is not required routinely, because the cohort may have had more severe or persistent BP elevation that warranted referral to the Pediatric Hypertension Center. Third, the database did not include information about ethnicity, a relevant consideration in assessing children with hypertension; however, recent studies (eg, cardiovascular impact of white coat hypertension27) were also missing this information without compromising the validity of the findings.
Most children and adolescents who are referred for evaluation of mild-to-moderate hypertension probably have essential hypertension, as evidenced by the presence of a positive family history for hypertension, an elevated BMI, and hypercholesterolemia in a large percentage of patients. The current recommendation for the diagnostic workup can be safely streamlined to a more limited battery of tests in most cases of mild-to-moderate hypertension. In Table 5, we summarize the diagnostic battery that would be indicated on the basis of a cutoff value of 5%, 10%, and 20% abnormal results as a criterion for a useful test. This information can be applied and tailored to match the particular referral population and the specific setting where the evaluation is being performed. More extensive testing should be guided by specific clinical indications or abnormal findings in the first-line tests. This recommendation is not applicable to children with severe hypertension or those with significant symptoms, for whom a thorough workup is urgently warranted. Any specific test should be performed when the history or clinical findings point to a specific entity. This information should be useful to general pediatricians, who are responsible for the care of an increasing number of patients with hypertension and/or obesity.
We thank Bernard Gauthier, MD, for thoughtful review and comments about the manuscript.
- Accepted July 29, 2008.
- Address correspondence to Howard Trachtman, MD, Schneider Children's Hospital, Division of Nephrology, Room SCH 365, 269-01 76th Ave, New Hyde Park, NY 11040. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Mild-to-moderate hypertension is increasingly prevalent in pediatric patients. Current recommendations suggest that a broad battery of diagnostic tests be performed to evaluate these patients; however, the yield of abnormal findings for each test has not been assessed.
What This Study Adds
This retrospective chart review provides an assessment of the yield of diagnostic tests that are performed to assess children and adolescents with mild-to-moderate hypertension. The findings are useful in formulating more cost-effective strategies for evaluating these patients.
- ↵National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics.2004;114 (2 suppl 4th report):555– 576
- ↵Brewer ED. Evaluation of hypertension in childhood diseases. In: Barratt TM, Avner ED, Harmon WE, eds. Pediatric Nephrology. 4th ed. Baltimore, MD: Lippincott, Williams & Wilkins; 2004:1179–1197
- Rowan S, Androgues H, Mathur A, Kamat D. Pediatric hypertension: a review for the primary care provider. Clin Pediatr (Phila).2005;44 (4):289– 296
- ↵Goonasekera CD, Dillon MJ. The child with hypertension. In: Webb N, Postlethwaite R, eds. Clinical Paediatric Nephrology. 3rd ed. Oxford, England: Oxford University Press; 2003:151–161
- Sorof JM, Lai D, Turner J, et al. Overweight, ethnicity, and the prevalence of hypertension in school-aged children. Pediatrics.2004;113 (3 pt 1):475– 484
- ↵Feld LG, Meyers KE, Kaplan BS, Stapleton FB. Limited evaluation of microscopic hematuria in pediatrics. Pediatrics.1998;102 (4). Available at: www.pediatrics.org/cgi/content/full/102/4/e42
- ↵Jassal SK, Langenberg C, von Muhlen D, Bergstrom J, Barrett-Connor E. Usefulness of microalbuminuria versus the metabolic syndrome as a predictor of cardiovascular disease in women and men >40 years of age (from the Rancho Bernardo Study). Am J Cardiol.2008;101 (9):1275– 1280
- ↵Daniels SR, Greer FR, Committee on Nutrition. Lipid screening and cardiovascular health in childhood. Pediatrics.2008;122 (1):198– 208
- ↵Daniels SR, Loggie JM, Khoury P, Kimball TR. Left ventricular geometry and severe left ventricular hypertrophy in children and adolescents with essential hypertension. Circulation.1998;97 (19):1907– 1911
- ↵Feld LG, Corey H. Hypertension in children. Pediatr Rev.2007;28 (8):283– 298
- Copyright © 2008 by the American Academy of Pediatrics