Article, Cardiology

Left ventricular dysfunction screening in hypertensive patients with N-terminal pro-B-type natriuretic peptide and electrocardiogram

Unlabelled imageAmerican Journal of Emergency Medicine (2012) 30, 214-217

Brief Report

Left ventricular dysfunction screening in Hypertensive patients with N-terminal pro-B-type natriuretic peptide and electrocardiogram?,??

Alexander T. Limkakeng Jr. MD a,?, Weiying Drake MD a, Giselle Mani BA a,

Debbie Freeman RN a, Randall Best MD a, L. Kristin Newby MD b, Abhinav Chandra MD a

aDepartment of Surgery, Division of Emergency Medicine, Duke University Medical Center, Durham, NC 27710, USA

bDepartment of Medicine, Division of Cardiology, Duke University Medical Center, Durham, NC 27710, USA

Received 21 September 2010; revised 10 November 2010; accepted 14 November 2010


Objective: Early recognition of left ventricular hypertrophy is important because Antihypertensive treatment decreases morbidity and mortality. The ideal screening method for left ventricular hypertrophy in Hypertensive emergency department (ED) patients has not been identified. Our objective was to determine the diagnostic accuracies of electrocardiogram (ECG) and N-terminal pro-B-type natriuretic peptide (pro-BNP) for left ventricular hypertrophy individually and in combination in hypertensive ED patients.

Methods: Prospective diagnostic study in an academic urban tertiary care hospital ED with annual census of 65 000 visits. Inclusion criteria are as follows: adult ED patients with systolic blood pressure greater than or equal to 160 mm Hg or diastolic blood pressure greater than or equal to 100 mm Hg on 2 or more measurements taken 60 minutes apart. Exclusion criteria are as follows: patients with heart failure, renal insufficiency/failure, acute myocardial infarction, or without recent or scheduled echocardiograms. All patients received echocardiograms and had pro-BNP levels measured using a RAMP point-of-care device (Response Biomedical, Vancouver, BC, Canada). We calculated Diagnostic test characteristics with 95% confidence intervals (CIs).

Results: A total of 49 patients were enrolled. The average age was 57.9 years, 26.5% were male, and 63.3% were African American. Thirty-two patients (65%) had left ventricular hypertrophy by echocardiogram. Twenty-one (43%) had ECG evidence of left ventricular hypertrophy. Median pro- BNP level was 268 pg/mL. The combination of the 2 tests provided the greatest specificity (94%; 95% CI, 69%-99.7%) and positive predictive value (94%; 95% CI, (68%-99.7%).

Conclusions: The combination of ECG and pro-BNP is a promising screening algorithm for identification of hypertensive ED patients with left ventricular hypertrophy.

(C) 2012

? This paper has been presented as a poster at the annual Society of Academic Emergency Medicine conference, New Orleans, LA in May 2009. Its abstract appeared in the corresponding journal Supplement.

?? This study was made possible through support from Response Biomedical, Inc., the maker of the NT-ProBNP RAMP device used in the study. They

provided supplies and research coordinator salary support. The authors retained full control over the data and all decisions regarding the writing and publication of the manuscript. The authors do not have any other financial interest in or other relationships that would be perceived as a conflict of interest.

* Corresponding author. Tel.: +1 919 684 5537; fax: +1 919 681 8521.

E-mail address: [email protected] (A.T. Limkakeng).

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2010.11.021

BNP in Hypertension


At least 50 million Americans have hypertension warranting some form of treatment [1]. In addition, suboptimal blood pressure control is the number one attributable risk for death throughout the world [2]. Hypertension precedes the development of heart failure in approximately 90% of heart failure patients and increases risk for heart failure by 2- to 3-fold [3].

A recent reclassification of heart failure recognizes that therapeutic interventions even before the appearance of left ventricular hypertrophy can reduce the morbidity and mortality of heart failure [4]. Current left ventricular hypertrophy Screening tests, such as electrocardiogram (ECG), have been found lacking sensitivity. Therefore, a rapid, inexpensive, screening test for left ventricular hypertrophy in hypertensive emergency department (ED) patients could identify patients most in need of urgent follow-up and treatment. N-terminal pro-B-type natriuretic peptide (pro-BNP) release is one of the earliest responses to hemodynamic pressure overload and is elevated in patients with left ventricular hypertrophy [5].

In this pilot study, we seek to determine the diagnostic accuracies of pro-BNP and ECG individually and in combination for predicting left ventricular hypertrophy in hypertensive ED patients. Our primary outcome was echocardiographic evidence of left ventricular hypertrophy.


Study design

This is an observational study of diagnostic test performance. Our institutional review board approved this study, and all subjects gave informed consent.

Study setting and population

Patients were enrolled in an academic urban tertiary care hospital ED with annual census of 65 000 visits from October 2007 to August 2008. Patients older than 18 years with systolic blood pressure greater than 160 mm Hg or a diastolic blood pressure greater than 100 mm Hg on 2 or more measurements at least 60 minutes apart in the ED were eligible. We excluded patients with a history of renal insufficiency or failure (as defined by a glomerular filtration rate b60 mL/min per 1.73 m2 calculated by the Modification of Diet in Renal Disease formula [6]), who were determined to have acute coronary syndrome (as defined by abnormal troponin T value on their first cardiac marker or ST-segment elevation on ECG), with a known history of heart failure, or who were unable to provide informed consent. We also excluded patients who did not have echocardiogram data in


the previous 3 months or who were not scheduled to have an echocardiogram as part of their clinical care.

Study protocol

Trained research coordinators prospectively screened electronic ED tracking boards and medical records for eligible patients. Each patient was approached for consent. After consent, research coordinators collected a 3-mL blood specimen and determined pro-BNP levels using a point-of- care platform (RAMP platform; Response Biomedical, Vancouver, BC, Canada). All coordinators received formal in-service training on the point-of-care device and had extensive formal training in venipuncture. One author, a board-certified emergency physician, blinded to pro-BNP levels, scored all ECGs dichotomously for left ventricular hypertrophy using the Selvester score [7]. In addition, a research coordinator prospectively recorded demographic, laboratory, radiographic, medication, ECG, and echocardio- gram data from patients, providers, and electronic medical records on standardized case report forms. Response Biomedical provided supplies for measuring pro-BNP, staff training, and research coordinator salary support. The authors retained full control over the data and all decisions regarding the writing and publication of the manuscript.

Key outcome measures

An echocardiogram reporting left ventricular hypertrophy was the primary outcome. We chose this because it is a marker of hypertensive damage to the heart and a precursor to heart failure. It is also routinely measured in all echocardiograms. We used it as a dichotomous outcome to facilitate diagnostic characteristic analysis. Echocardiograms were performed and read as per usual care by board-certified cardiologists blinded to study hypothesis, ECG, and pro- BNP values.

Data analysis

For pro-BNP, a receiver operator curve was calculated for predicting left ventricular hypertrophy on echocardiogram, and the area under curve was calculated. Based on this analysis, the ideal cut point was identified, and diagnostic test characteristics were calculated with 95% confidence intervals (CIs) for a dichotomized pro-BNP. Diagnostic test characteristics for ECG were calculated using similar dichotomized variable analysis with 95% CIs. Likewise, we calculated the diagnostic test characteristics of a combination of the dichotomized pro-BNP and ECG results, considering the combination positive if either individual test or both were positive. We performed a logistic regression analysis with left ventricular hypertrophy as outcome and using the continuous pro-BNP and dichotomous ECG as independent main effect variables plus an interaction

216 A.T. Limkakeng Jr. et al.

variable. To calculate odds ratios with 95% CIs, univariate logistic regression analyses were calculated for each dichotomized variable (ECG, N-terminal pro-BNP, and combination of both.) All statistics were calculated using SAS 9.2 Enterprise Guide 4.2 (Cary, NC).


Table 1 Diagnostic test characteristics of N-terminal pro-BNP (NT-proBNP) and ECG alone and in combination for left ventricular hypertrophy (LVH)

LVH (+) (-)

NT-proBNP (+) 21 5

alone (-) 11 12

LVH (+) (-)

ECG alone (+) 18 3

(-) 14 14

LVH (+) (-)

ECG + NT- (+) 15 1

proBNP (-) 17 16

Sensitivity 65% (47%-81 %)

Specificity 71% (44%-89 %)

PPV 81% (60%-93 %)

NPV 52% (31%-73 %)

LR+ 4.2 (1.9-9.4)

LR- 0.9 (0.6-1.5)

Sensitivity 56% (50%-78 %)

Specificity 82% (56%-95 %)

PPV 86% (63%-96 %)

NPV 50% (31%-69 %)

LR+ 6 (2.1-17.4)

LR- 1 (0.6-1.6)

Sensitivity 47% (30%-65 %)

Specificity 94% (69%-99.7 %)

PPV 94% (68%-99.7 %)

NPV 48% (31%-66 %)

LR+ 15 (2.2-100.5)

LR- 1.1 (0.7-1.6)

Diagnostic test characteristics listed with 95% CIs. NT-proBNP was considered positive if level was greater than 258 pg/mL. electrocar- diogram was considered positive if Sylvester score was positive. PPV indicates positive predictive value; NPV, negative predictive value; LR+, likelihood ratio positive; LR-, likelihood ratio negative.

A total of 56 patients were enrolled, of whom 7 did not have complete data and were excluded. The average age was

57.9 years. Patients were predominantly female (73.5%) and African American (63.3%). The most common symptoms reported were chest pain (50.7% of subjects), headache or dizziness (18.8%), or shortness of breath (11.6%). The range of initial blood pressure readings extended from 162 to 254 mm Hg for systolic pressures (mean +- SD, 200.1 +- 23.4) and 75 to 135 mm Hg (mean +-SD, 100.7 +- 14.9) for diastolic pressures, with a normal distribution for both. The majority (77.5%) reported taking some form of Antihypertensive medication as outpatients.

Thirty-two patients (65%) had evidence of left ventricular hypertrophy on echocardiogram. All of these echocardio- grams were performed after patient enrollment as part of usual clinical care. Among these patients, the most commonly reported symptom was chest pain (31.9%; compared with 18.8% of those without left ventricular hypertrophy). Twenty-one (43%) had ECG evidence of left ventricular hypertrophy. Median pro-BNP level was 268 pg/ mL, with a positively skewed distribution of values.

A receiver operator curve for pro-BNP as a continuous variable was calculated with an area under the curve of

0.70. Based on these calculations, an ideal cutoff point of 258 pg/mL was identified. The diagnostic test character- istics N-terminal pro-BNP and ECG alone and in combination are summarized in Table 1. The combination of the 2 tests provided the greatest specificity and positive predictive value.

In logistic regression analysis, the overall model was significant. Electrocardiogram was an independent predictor of outcome, but pro-BNP was not. There was no significant interaction between ECG and pro-BNP. The odds ratios calculated from univariate logistic regression with dichoto- mized variables for ECG, pro-BNP, and the combination of both were 6.0 (95% CI, 1.44-25.0), 1.0 (95% CI, 1.00-1.01),

and 5.5 (95% CI, 1.53-19.7), respectively.


In this proof-of-concept study, we have demonstrated the feasibility of left ventricular hypertrophy screening in hypertensive ED patients using pro-BNP and ECG. Al- though individually, these tests demonstrated only modest

sensitivity and specificity for identifying left ventricular hypertrophy, in combination, they demonstrated high specificity and positive predictive value.

The role of the ED in screening for hypertensive end- organ damage has been widely debated. It is known that many undertreated hypertensive patients are seen in the ED [8], and the long-term consequences of undertreated hypertension are well established. However, hypertension screening in the ED may increase cost and impede patient flow. Therefore, a rapid, inexpensive ED screening paradigm is desirable.

Researchers have long observed that BNP levels are higher in patients with left ventricular hypertrophy [5]. Although its use for screening hypertensive patients in outpatient settings has been described [9], the use of BNP for hypertensive screening in the ED has not been extensively studied.

This pilot study suggests that due to its high specificity and positive predictive value, a combination of pro-BNP and ECG could be used to prioritize patients for more intensive antihypertensive therapies. These 2 tests are very rapid, inexpensive, and noninvasive. They would thus make an ideal tool for the busy emergency physician who wishes to identify high-risk hypertensive patients without slowing down departmental flow. Although its modest sensitivity would indicate that many patients with left ventricular hypertrophy would be missed by this screening method, currently, screening and referral from the ED occur so infrequently that even identifying only the highest-risk

BNP in Hypertension

patients would be an improvement. This may particularly be the case when outpatient referral resources are scarce or



overburdened. This finding may have significant public health implications.


It is known that plasma BNP levels are affected by factors other than ventricular wall stretch [10]. Although we were able to control for some confounders of BNP level such as renal dysfunction, we were unable to control the timing of medications taken because of the ED setting. Our study is limited by convenience sampling and small sample size. We are thus unable to control for age or weight. In addition, because of small sample size, we are unable to assess the role for BNP screening in subpopulations of left ventricular hypertrophy such as concentric remodeling or eccentric hypertrophy. It should also be noted that almost 74% of our patients were female, possible because of this small sample size.


A combination of ECG and pro-BNP demonstrates high specificity and positive predictive value for left ventricular hypertrophy on echocardiogram. Future study with more rigorous control and increased sample size is warranted.

  1. Hajjar I, Kotchen TA. Trends in prevalence, awareness, treatment, and control of hypertension in the United States, 1988-2000. JAMA 2003;290(2):199-206.
  2. Guilbert JJ. The world health report 2002–reducing risks, promoting healthy life. Educ Health (Abingdon) 2003;16(2):230.
  3. Senni M, Redfield MM. Heart failure with preserved systolic function. A different natural history? J Am Coll Cardiol 2001;38(5):1277-82.
  4. Hunt SA. ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure). J Am Coll Cardiol 2005;46(6):e1-e82.
  5. Kohno M, Horio T, Yokokawa K, et al. Brain natriuretic peptide as a cardiac hormone in Essential hypertension. Am J Med 1992;92(1): 29-34.
  6. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130(6):461-70.
  7. Selvester RH, Wagner GS, Hindman NB. The Selvester QRS scoring system for estimating myocardial infarct size. The development and application of the system. Arch Intern Med 1985;145(10):1877-81.
  8. Umscheid CA, Maguire MG, Pines JM, et al. Untreated hypertension and the emergency department: a chance to intervene? Acad Emerg Med 2008;15(6):529-36.
  9. Seki S, Tsurusaki T, Kasai T, et al. Clinical significance of B-type natriuretic peptide in the assessment of untreated hypertension. Circ J 2008;72(5):770-7.
  10. Cowie MR, Jourdain P, Maisel A, et al. clinical applications of B-type Natriuretic Peptide testing. Eur Heart J 2003;24(19):1710-8.

Leave a Reply

Your email address will not be published. Required fields are marked *