Heart Fail Rev (2014) 19:439–451 DOI 10.1007/s10741-014-9445-8

Performance of BNP and NT-proBNP for diagnosis of heart failure in primary care patients: a systematic review Ronald A. Booth • Stephen A. Hill • Andrew Don-Wauchope • P. Lina Santaguida Mark Oremus • Robert McKelvie • Cynthia Balion • Judy A. Brown • Usman Ali • Amy Bustamam • Nazmul Sohel • Parminder Raina

•

Published online: 27 June 2014 Ó Springer Science+Business Media New York 2014

Abstract National and international guidelines have been published recommending the use of natriuretic peptides as an aid to the diagnosis of heart failure (HF) in acute settings; however, few specific recommendations exist for governing the use of these peptides in primary care populations. To summarize the available data relevant to the diagnosis of HF in primary care patient population, we systematically reviewed the literature to identify original articles that investigated the diagnostic accuracy of B-type natriuretic peptide (BNP) and N-terminal proBNP (NTproBNP) in primary care settings. The search yielded 25,864 articles in total: 12 investigating BNP and 20 investigating NT-proBNP were relevant to our objective and included in the review. QUADAS-2 and GRADE were used to assess the quality of the included articles. Diagnostic data were pooled based on three cutpoints: lowest and optimal, as chosen by study authors, and manufacturers’ suggested. The effect of various determinants (e.g., age, gender, BMI, and renal function) on diagnostic R. A. Booth Department of Pathology and Laboratory Medicine, University of Ottawa, Ottawa, ON, Canada S. A. Hill
A. Don-Wauchope
C. Balion Department of Pathology and Molecular Medicine, McMaster University, Hamilton, ON, Canada P. L. Santaguida
M. Oremus
J. A. Brown
U. Ali
A. Bustamam
N. Sohel
P. Raina (&) Department of Clinical Epidemiology and Biostatistics, McMaster University, 1280 Main Street West, MIP Suite 309A, Hamilton, ON L8S 4K1, Canada e-mail: [email protected] R. McKelvie Department of Medicine, McMaster University, Hamilton, ON, Canada

performance was also investigated. Pooled sensitivity and specificity of BNP and NT-proBNP using the lowest [0.85 (sensitivity) and 0.54 (specificity)], optimal (0.80 and 0.61), and manufacturers’ (0.74 and 0.67) cutpoints showed good performance for diagnosing HF. Similar performance was seen for NT-proBNP: lowest (0.90 and 0.50), optimal (0.86 and 0.58), and manufacturers’ (0.82 and 0.58) cutpoints. Overall, we rated the strength of evidence as high because further studies will be unlikely to change the estimates diagnostic performance. Keywords Diagnosis
Primary care
B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP)
Systematic review

Introduction The diagnosis of heart failure (HF) remains a difficult clinical challenge in all settings. Unlike patients presenting to emergency rooms with symptoms of acute HF, patients presenting to primary care settings often have mild or no obvious symptoms, or they present with only risk factors for the condition. Early detection of left ventricular systolic dysfunction is important because early treatment has been shown to delay the progression to overt HF [1]. In primary care, diagnosis is based predominantly on clinical symptoms, as more specific diagnostics such as ECG or echocardiography are often not readily available. Therefore, identification of patients with HF in primary care using readily available diagnostic tests is key to early detection and treatment. B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) have emerged as promising markers for HF diagnosis, prognosis, and treatment. These peptides are

123

440

Heart Fail Rev (2014) 19:439–451

secreted into the bloodstream by cardiac myocytes in response to increased ventricular wall stress, hypertrophy, and volume overload. Increased levels of these peptides are seen in patients with HF, both symptomatic and asymptomatic, and it is well established that a low result can exclude HF [2]. While many national and international guidelines recommend the use of natriuretic peptides as an aid to diagnosis of HF in acute settings [3, 4], few specific recommendations exist for using these peptides in primary care populations [5, 6]. To summarize the available data relevant to diagnosing HF in primary care patient populations, we systematically reviewed the literature to identify original articles that investigated the diagnostic accuracy of BNP and NT-proBNP in primary care settings. We also included an investigation of the effect of various determinants (e.g., age, gender, renal function) on diagnostic accuracy. Our specific research questions included the following: In patients presenting to a primary care physician with risk factors, signs, or symptoms suggestive of HF: (a) (b) (c)

What is the test performance of BNP and NTproBNP for diagnosing HF? What are the optimal decision cutpoints for BNP and NT-proBNP to diagnose and exclude HF? What determinants (e.g., age, gender, comorbidity) affect the test performance of BNP and NT-proBNP for diagnosing HF?

Methods A broad literature search strategy was implemented to reflect the wide scope of this review. Full details of the methods can be found in the accompanying overview and methods paper. Inclusion and exclusion criteria We included studies that recruited patients presenting to primary care physicians with signs or symptoms of HF, or patients who were at risk of HF at the time of presentation. We used the American Academy of Family Physicians’ definition of primary care, i.e., care provided by physicians as a point of first contact and continuing care for persons with any undiagnosed sign, symptom, or health concern not limited by problem origin, organ system, or diagnosis [7]. We excluded studies where all subjects are B18 years of age, studies where subjects had known acute HF or known exacerbation of chronic stable HF, and studies that included only subjects with specific conditions that may

123

influence BNP results (e.g., heart transplantation, obesity, hypertrophic cardiomyopathy, or valvular lesions). Comparators and outcomes We included any comparator method of diagnosing HF that did not use BNP or NT-proBNP. Since no gold standard diagnostic criteria exist in HF, sensitivity and specificity of BNP and NT-proBNP were calculated using whatever comparator methods or prediction scores were used in the included studies. Outcomes included test performance characteristics (i.e., sensitivity, specificity, positive and negative likelihood ratios [LR?, LR-], diagnostic odds ratios [DOR], and receiver operating characteristic [ROC] curves). We also examined the effects of various decision cutpoints and determinants (e.g., age, sex, and comorbidities) on the performance characteristics. We considered adverse events associated with the administration of the diagnostic tests if these events were presented in the articles. Data extraction Extracted data for all studies included general study characteristics (e.g., study design, patient characteristics, study end points), details of the patient population, and comorbidities. We also extracted blood sample type for natriuretic peptide measurement (plasma or serum), assay source (name), type of peptide assessed (BNP, NT-proBNP, or both), and specimen storage temperature (if applicable). Extracted outcomes included the type of analytical system, cutpoints (lowest and optimal, as chosen by study authors, and manufacturers’ suggested), and whether a specific outcome was primary or secondary. In addition, we extracted the location of care (i.e., the diagnostic data associated with primary care), information regarding the reference standard, and test performance characteristics. For the test performance characteristics, the extracted data included either primary data to allow us to calculate these characteristics on our own, or the summary estimates that were already presented by the authors. Adverse events were extracted if identified in the articles. In the case of multiple publications of the same study cohort, the primary study paper was considered for statistical analysis. All data were extracted separately for BNP and NT-proBNP. We used meta-analysis to obtain summary estimates of sensitivity and specificity. Details of our meta-analysis procedures are described in the overview paper. We also reported summary ROC (SROC) curves. The areas under the SROC curves were used to measure diagnostic performance of the tests [8]. DOR was calculated and pooled using the generalized linear mixed (GLM)

Heart Fail Rev (2014) 19:439–451

441

model approach to bivariate meta-analysis of sensitivity and specificity suggested by Chu and Cole [9].

analyzing NT-proBNP (n = 20) used the ElecsysÒ proBNP immunoassay.

Risk of bias, applicability assessment, and strength of evidence

Reference standard

We used the QUADAS-2 tool to assess risk of bias in four key domains: patient selection, index test(s), reference standard, and flow and timing. The questions in each domain were rated (low, high, unclear) in terms of risk of bias and concerns regarding applicability (for patient selection, index test(s), and reference standard only), with associated signaling questions to help with bias and applicability judgments. To grade the strength of evidence (SOE) [10, 11], we assessed two primary outcomes: sensitivity and specificity. Other diagnostic performance indicators [positive (PPV) and negative (NPV) predictive values, LR? and LR-, accuracy, and DOR] can be calculated from sensitivity and specificity if the prevalence of disease is known. As such, the conclusions regarding SOE for these performance indicators are unlikely to be different from those drawn for sensitivity and specificity. We graded the SOE in four domains; risk of bias (low, medium, or high) consistency (consistent, inconsistent, unknown, or not applicable), directness (direct or indirect), and precision of the evidence (precise or imprecise) [10, 11]. The overall SOE for each outcome was rated as high, moderate, low, or insufficient [10]. Further details regarding risk of bias SOE are presented in the overview and methods paper.

Results The search yielded 25,864 records identified from six bibliographic databases. An additional 35 records were identified from three gray literature sources: regulatory agency websites, clinical trial databases, and conference sources. After duplicates were removed, 16,893 records were screened at title and abstract level; 3,616 citations advanced to full-text screening. Twelve BNP articles met the inclusion criteria for primary care populations. Seven examined BNP only [12–18] and five examined both BNP and NT-proBNP [19–23]. Twenty studies evaluating NT-proBNP were included, with the majority focused on NT-proBNP alone (n = 15) [24– 38]. Analysis of BNP was performed by three different assays, the Triage-BNP assay was used in 10 studies [12– 20, 22], the ADVIA-CentaurÒ BNP assay in 1 study [21], and the Abbott AxSYMÒ BNP microparticle enzyme immunoassay (MEIA) in 1 study [23]. All studies

Eight studies that examined BNP (some also examined NTproBNP) based the diagnostic reference standard on the clinical judgment of at least one cardiologist [12–15, 20– 23]. Four of these eight studies used multiple cardiologists or other specialists [12–14, 23]. Four studies involving BNP included the Framingham criteria in their diagnosis [12, 13, 15, 17], while the remainder of the BNP studies used general practitioners [19], ECHO only [16], or did not report the reference standard [18]. For studies that analyzed NT-proBNP, 14 based the diagnostic reference standard solely on clinical judgment. Four of these 14 had a reference standard agreed upon by at least 2 physicians [19, 21, 23, 29], with eight other studies basing the final diagnosis on the opinion of a single physician [22, 24, 25, 28, 30, 33, 34, 36]. Three NT-proBNP studies based the final diagnosis of HF on both clinical judgment and the results of echocardiography [31, 32, 35], while two reported that the definitive diagnosis was ‘‘based on the Framingham criteria’’ [37, 38]. Two studies did not report the reference standard [26, 28]. Diagnostic performance The twelve studies evaluating BNP used several cutpoints ranging from 30 [12, 20] to 500 [12] pg/mL (ng/L), with reported sensitivities ranging from 25 [17] to 97 % [12], specificities from 23 [15] to 92 % [12], and AUCs of 0.62 [22] to 0.93 [21]. The 20 studies evaluating NT-proBNP also used several cutpoints, which ranged from 25 [25] to 6,180 [36] pg/mL or ng/L. Three studies [28, 31, 33] measured NT-proBNP in pmol/L. Reported sensitivities from all studies ranged from 44 [36] to 100 % [31, 32, 35, 38] and specificities ranged from 3 [35] to 97 % [26, 27] and AUCs ranged from 0.70 [37] to 0.98 [32]. Due to the use of several cutpoints in the literature, we selected three relevant cutpoints to create forest plots of sensitivities and specificities (Fig. 1): the lowest cutpoint presented in each article, the optimal cutpoint as chosen by each set of authors, and the manufacturers’ recommended cutpoint (the latter is likely to be the most commonly used in clinical practice). In the case of BNP, the manufacturers’ recommended cutpoint was 100 ± 5 pg/mL. For NTproBNP, the manufacturers’ recommended cutpoints were 150 pg/mL for patients younger than 75 years of age and 450 pg/mL for patients above 75 years of age. A summary of the performance characteristics for BNP and NT-proBNP using the three selected cutpoints is presented in Table 1.

123

442

Heart Fail Rev (2014) 19:439–451

Fig. 1 Summary forest plots showing sensitivity and specificity for BNP and NT-proBNP using the bivariate mixed-effect model. a Manufacturers suggested cutpoint (100 pg/mL) for studies using BNP. b Optimum cutpoint for studies using NT-proBNP

Table 1 Summary of test performance characteristics based on the manufacturer’s suggested, optimum and lowest cutpoints Test

BNP

Cutpoint

Assay type

Manufacturer Optimum

Number of studies

D

8 8

Lowest NT-proBNP

Test

BNP

NT-proBNP

Sensitivity

Specificity 2

Est

95 % CI

I

0.74 0.80

0.63, 0.84 0.71, 0.89

94.0 92.9

Est

95 % CI

I2

0.67 0.61

0.50, 0.85 0.43, 0.80

99.1 98.4

10

0.85

0.77, 0.92

95.8

0.54

0.42, 0.66

97.3

2

0.82

0.66, 0.98

86.7

0.58

0.54, 0.62

12.3

Optimum

11

0.86

0.79, 0.93

87.8

0.58

0.42, 0.75

99.0

Lowest

12

0.90

0.85, 0.95

84.7

0.50

0.41, 0.60

96.4

Manufacturer

E

LR?

LR-

Log DOR

AUC

Est

95 % CI

I2

Est

95 % CI

I2

Est

95 % CI

I2

Est

95 % CI

2.60

1.69, 4.00

96.9

0.38

0.23, 0.62

92.7

2.02

1.24, 2.80

90.2

0.80

0.71, 0.88

2.27

1.59, 3.24

96.1

0.30

0.16, 0.55

93.4

2.07

1.20, 2.94

90.9

0.80

0.71, 0.90

1.87

1.50, 2.34

94.1

0.22

0.11, 0.44

93.7

2.18

1.41, 2.95

87.9

0.81

0.73, 0.90

1.96

1.45, 2.66

87.7

0.29

0.10, 0.88

75.7

1.90

0.56, 3.25

78.9

–

–

2.18

1.81, 2.63

89.2

0.23

0.16, 0.34

75.5

2.50

1.87, 3.13

80.2

0.85

0.79, 0.90

1.87

1.59, 2.20

91.0

0.19

0.12, 0.29

73.1

2.38

1.86, 2.91

71.6

0.84

0.78, 0.89

AUCs were evaluated for groups with 4 or more studies Assay: D—Triage B-type test, E—Elecsys proBNP immunoassay AUC area under the curve, CI confidence interval, DOR diagnostic odds ratio, Est estimate, LR? positive likelihood, LR- negative likelihood

BNP The pooled estimate for sensitivity using the optimum cutpoint was 0.82 (95 % CI 0.69–0.90). All except a single study [17] (sensitivity = 0.25) had sensitivities greater than 0.80. The low sensitivity of the Barrios study may be due to a predominantly elderly population and high prevalence of diastolic HF. Specificities using the optimum cutpoint (which ranged from 30 to 147 pg/mL) were not as high as the sensitivities, with the overall specificity being 0.64 (95 % CI 0.45 to 0.79).

123

Pooling using the lowest cutpoint produced a slightly higher sensitivity of 0.89 (95 % CI 0.77–0.95) and a corresponding lower specificity of 0.54 (95 % CI 0.41–0.66). Eight studies used the manufacturer’s recommended cutpoint and we calculated an overall sensitivity of 0.76 (95 % CI 0.59–0.87), which was slightly lower than that for the optimal cutpoints. The corresponding specificity of 0.71 (95 % CI 0.52–0.85) was slightly higher than the specificity for the optimum cutpoint. Summary ROC analysis produced AUCs of 0.81 (95 % CI 0.77–0.84) for the optimum cutpoint, 0.76 (95 % CI

Heart Fail Rev (2014) 19:439–451

443

0.72–0.80) for the lowest cutpoint, and 0.80 (95 % CI 0.76–0.83) for the manufacturers’ suggested cutpoint.

Risk of bias and applicability assessment and strength of evidence

NT-proBNP

The tests for publication bias, assessed using the GRADE tool, exposed no significant bias in the following conditions in our meta-analysis of BNP and NT-proBNP diagnostic use in primary care: (1) optimum cutpoint, (2) lowest cutpoint, and (3) manufacturers cutpoint. Risk of bias, assessed using the QUADAS-2, was low for the domains of reference standard, flow, and timing, with the majority of the studies showing a low risk of bias. In terms of patient selection, 58 percent of the studies had an unclear risk of bias. In the domain of index test, 33 percent of the studies had a high risk of bias. Despite the potential high risk of bias in the index test, the overall risk of bias was rated low. Summary plots of the risk of bias assessment are presented in Fig. 2. For assessment of the SOE, two primary outcomes were chosen: sensitivity and specificity. For all studies that presented sensitivity and specificity data (BNP n = 11 [12–17, 19–23]; NT-proBNP n = 17 [19–22, 24, 26–28, 30–38]), the SOE was examined for each of the relevant groups of cutpoints. The SOE using the GRADE tool for both BNP and NT-proBNP is presented in Table 3. For directness, the issue revolved around diagnostic accuracy, and sensitivity and specificity in primary care populations were assessed in this review. This domain was rated as direct, as these are concepts that are generally understood by clinicians and can be applied directly to the diagnosis of HF in similar clinical settings. Precision was assessed via the confidence intervals (CIs) around the summary estimates for sensitivity and specificity for BNP and NT-proBNP. The CIs were found to be imprecise; therefore, this domain was rated as imprecise. For the domain of consistency, the directions of the estimates were consistent and similar for BNP sensitivity, with the exception of a single study [17]. The directions of the sensitivity estimates were similarly consistent and the CIs are small for NT-proBNP, so this domain was rated as consistent for both BNP and NT-proBNP. However, the specificity was rated as inconsistent because of the wide range of estimates across studies for both BNP and NT-proBNP.

The optimal cutpoint chosen by the authors ranged from 87 to 424 pg/mL and produced a pooled sensitivity of 0.88 (95 % CI 0.81–0.93). Seven of the studies [19, 27, 30–32, 34, 36] showed sensitivities greater than 0.90. The pooled specificity of 0.58 (95 % CI 0.42–0.75) was not as high as the pooled sensitivity because the authors’ intent was to select cutpoints that optimized sensitivity. A single study by Stahrenberg et al. [37] had a sensitivity of 0.55 (95 % CI 0.44–0.65), which was likely due to a relatively high cutpoint of 220 pg/mL; the specificity in this study was 0.61 (95 % CI 0.47–0.74). The lowest cutpoint chosen by the authors produced an improved pooled sensitivity of 0.90 (95 % CI 0.85–0.95) when compared to the optimal cutpoint, with a limited decrease in pooled specificity, which was 0.50 (95 % CI 0.41–0.60). All but three of the 20 studies [22, 25, 37] had sensitivities greater than 0.90. Since only two studies presented data on the manufacturers’ recommended cutpoints for NT-proBNP, these data were not pooled in a meta-analysis. As with the summary plots, the SROC curves were developed based on the optimum and lowest cutpoints. The AUCs were 0.86 (95 % CI 0.82–0.88) for the optimum cutpoint and 0.82 (95 % CI 0.79–0.85) for the lowest cutpoint. Determinants of test performance We examined the effects of various determinants on the performance of BNP and NT-proBNP for diagnosing HF. The effects of age, gender, BMI, and renal failure were investigated by several included studies. A single study by Park et al. [21] examined the diagnostic properties of both BNP and NT-proBNP on gender, age 65 years and above versus age less than 65 years, BMI above and below 25 kg/m2, and renal function (creatinine clearance) with a cutoff of 65 mL/min as calculated by the Cockroft–Gault equation. Patients were then subcategorized by identification of decreased LVEF (\45 %) or advanced diastolic dysfunction (DD) with preserved LVEF ([45 %). Patients with decreased LVEF in all cases required a higher cutpoint relative to advanced DD patients to maintain optimal sensitivities and specificities. Fuat et al. [19] compared the AUC of males and females and did not find a significant difference (males 0.79, females 0.80). Five additional studies [22, 32–35] were identified that examined various determinants and are collectively summarized in Table 2.

Discussion Our systematic review indicates that both BNP and NTproBNP are useful diagnostic tools to identify patients with HF in primary care settings. Study populations in the included articles had signs and symptoms suggestive of HF or they had risk factors for HF without overt signs or symptoms of disease. Studies that used BNP or NT-proBNP to screen for HF in healthy populations were not considered

123

444

Heart Fail Rev (2014) 19:439–451

Table 2 Effect of various determinants on the sensitivity and specificity of BNP and NT-proBNP for diagnosis of HF Study

Determinant

End point

Cutpoint (pg/mL)

Sensitivity (%) (95 % CI)

Specificity (%) (95 % CI)

84

BNP Age Park et al. [21]

C65 years \65 years

Gender Park et al. [21]

LVEF \45

250

84

Advanced DD

236

84

84

LVEF \45

82

84

84

Advanced DD

70

83

83 79

LVEF \45

111

81

Advanced DD

99

80

80

LVEF \45

209

85

85

Advanced DD

166

85

85

\25 kg/m2

HF

100

89

38

25–30 kg/m2

HF

100

85

38

[30 kg/m2

HF

100

81

49

C25 kg/m2

LVEF \45

151

85

85

Advanced DD

82

80

80

Males Females

BMI Christenson et al. [22]

Park et al. [21]

\25 kg/m

2

LVEF \45

154

81

81

Advanced DD

140

83

83

C60 mL/min

LVEF \45

89

82

82

\60 mL/min

Advanced DD LVEF \45

70 264

83 78

82 78

Advanced DD

247

78

78

LVEF \45

1,446.00

82

81

Renal function (creatinine clearance) Park et al. [21]

NT-proBNP Age Park et al. [21]

C65 years \65 years

Shelton et al. [35]

Advanced DD

1,356

84

83

LVEF \45

379

84

84 82

Advanced DD

276

83

SR B75 years

MSHD

357

73.4 (47.3–79.3)

78.6 (51.3–84.2)

SR [75 years

MSHD

652

69.1 (43.0–79.0)

78.6 (47.7–87.8)

AF B75 years

MSHD

1,758

69.8 (58.3–92.7)

90.2 (63.2 to 96.9)

AF [75 years

MSHD

1,764

68.9 (38.7–87.8)

60.6 (43.9–97.2)

Gender Mikkelsen et al. [32]

Male

HF

85

95 (83–99)

71 (55–84

Female

HF

110

98 (87–100)

88 (71–97)

Nielsen et al. [33]

Male Female

HF HF

93 144

96 94

67 69

Olofsson et al. [34]

Male B79 years Female B79 years Male [79 years Female [79 years

Park et al. [21]

123

Males

HF

200

90

56

HF

300

80

78

HF

200

79

64

HF

300

64

76

HF

200

100

50

HF

300

100

50

HF

200

100

22

HF

300

83

61

LVEF \45

510

81

81

Heart Fail Rev (2014) 19:439–451

445

Table 2 continued Study

Determinant

Females

End point

Cutpoint (pg/mL)

Sensitivity (%) (95 % CI)

Specificity (%) (95 % CI)

Advanced DD

431

83

81

LVEF \45

1,678

87

87

Advanced DD

860

85

85

HF

Age specific*

88

50 51

BMI Christenson et al. [22]

\25 kg/m2 25–30 kg/m

Park et al. [21]

2

HF

68

[30 kg/m2

HF

69

64

C25 kg/m2

LVEF \45

85

87

771

Advanced DD

309

80

80

LVEF \45

830

81

81

Advanced DD

682

81

81

C60 mL/min

LVEF \45

418

84

84

\60 mL/min

Advanced DD LVEF \45

276 1,981

83 78

82 78

Advanced DD

1,733

78

76

\25 kg/m2 Renal function (creatinine clearance) Park et al. [21]

* NT-proBNP age-specific cutpoints = 450 pg/mL for age \50 years, 900 pg/mL for 50–75 years, and 1800 pg/mL for >75 years

in this particular review. These broad inclusion criteria were chosen to represent the diverse patient populations seen in primary care. Therefore, our conclusions should be applicable to the majority of primary care practices. The results obtained in this review agree with a recent meta-analysis using individual patient-level data. This meta-analysis reported that both BNP and NT-proBNP had high sensitivities for diagnosing HF [39]. Overall, we found similar pooled sensitivities when we compared our summary results across the three groups of cutpoints, i.e., the authors’ self-defined optimal cutpoints, the lowest cutpoints, and the manufacturers’ recommended cutpoints. However, the pooled specificities were substantially lower in all three cases, which suggest the peptides have better utility for ruling out the presence of HF. In the case of BNP, we pooled studies that reported results for the manufacturers’ suggested cutpoint of 100 pg/mL. This analysis is particularly important because many laboratories will likely use the 100 pg/mL cutpoint due to the expense and complexity of developing in-house cutpoints. The majority of the studies with the 100 pg/mL cutpoint had sensitivities above 80 %. Two studies showed strikingly low sensitivities: Barrios et al. [17] and Murtagh et al. [18] had substantially lower sensitivities and correspondingly higher specificities for identifying patients with HF. The study population used by Barrios et al. contained predominantly elderly patients, and HF was defined according to the Framingham criteria. Sixty percent of patients had DD with preserved LVEF and only 2.8 % had a reduced LVEF. The authors suggested that the reduced

sensitivity in their study relative to other studies was due to the high proportion of diastolic HF. Murtagh et al. [18] showed a sensitivity of 45 % for the detection of systolic HF, defined as LVEF \50 %. The patients included in Murtagh et al. had at least one risk factor for HF, but they were asymptomatic at the time of BNP measurement. The reduced sensitivity in Murtagh et al. may be due to the lack of symptoms, suggesting less severe HF, while the other pooled studies included patients with either dyspnea or clinically suspected HF. Although caution should be exercised when drawing conclusions from a single study, the lower sensitivity may suggest decreased sensitivity for detection of HF in asymptomatic populations. For NT-proBNP, only two studies [22, 27] examined the manufacturers’ suggested cutpoints of 125 pg/mL for patients \75 years and 450 pg/mL for patients [75 years. The sensitivities were somewhat different; however, the specificities were similar. Gustafsson et al. [27] used an LVEF of \40 % to identify patients with HF and calculated sensitivity and specificity of 91 and 60 %, respectively. Christenson et al. [22] used cardiologist adjudication, an LVEF \40 %, and other signs, symptoms, and objective markers to diagnose HF; they found sensitivities ranging from 68 to 88 % and specificities ranging from 50 to 64 % for patients grouped by BMI (Table 2). The more rigorous reference standard may account for the lower sensitivity in the Christenson report. When the effect of various determinants on BNP and NT-proBNP were examined, we found that values for both peptides increased with age [21, 35] and declining renal

123

446

Heart Fail Rev (2014) 19:439–451

Fig. 2 QUADAS-2 assessment of risk of bias and applicability for BNP and NT-proBNP. a Risk of bias for BNP studies; b applicability for BNP studies; c risk of bias for NT-proBNP studies; d applicability of NT-proBNP studies

function [21]. A single study looked at the age effect on BNP and demonstrated that a higher cutpoint is required in patients greater than 65 years to maintain an optimal specificity compared with younger patients [21]. A similar age-related increase in NT-proBNP was seen in the same study, with a higher cutpoint required to maintain an optimal sensitivity [21]. A pooled analysis performed by Hildebrandt et al. showed similar results by demonstrating that higher cutpoints are required to maintain equivalent diagnostic accuracy as age increases [40]. The effect of BMI on BNP and NT-proBNP was investigated in two studies [21, 22]. Both studies showed a negative correlation between BMI and BNP or NT-proBNP, with decreasing sensitivities for diagnosing HF. However, no BMI-specific cutpoints were suggested in the included articles. Two studies investigated the effect of gender on BNP; both Fuat et al. [19] and Park et al. [21] did not identify any significant effects of gender on BNP. Five studies [19, 21, 32–34] examined the effect of gender on NT-proBNP, and although the authors identified different optimal cutpoints for males and females, no clear conclusions could be drawn regarding optimal cutpoints. Decreased renal function, measured by creatinine clearance (\60 mL/min), increased both BNP and NT-

123

proBNP concentrations; however, the effect was more pronounced with NT-proBNP [21]. This differential effect is likely due to reduced clearance of NT-proBNP, as NTproBNP is cleared by the kidneys [41], while BNP is not [42]. Patients who present with risk factors or symptoms of HF can have different underlying pathologies, either systolic or DD. Diastolic HF is distinct from systolic HF in the fact that patients have a preserved (normal or near normal) ejection fraction (EF) rather than a reduced EF. If patients are categorized based on EF alone, a significant portion of HF patients could be missed as the prevalence of diastolic HF is estimated to be similar to systolic HF [43]. Natriuretic peptides have the ability to identify both systolic and diastolic HF. Two studies divided their HF populations into those with and those without a reduced EF [21, 31] and measured the diagnostic accuracy of BNP and/or NTproBNP. In both cases, diagnostic cutpoints needed to be reduced to maintain sensitivity, suggesting in patients with diastolic HF lower cutpoints may be required; however, the authors of these studies did not make any specific recommendations with regard to cutpoints. The diagnostic accuracy of BNP and NT-proBNP was directly compared by five studies that were included in our review (Table 4) [19–23]. Two studies showed marginally

Lowest

Optimal

NT-proBNP

Manufacturer

Lowest

Optimal

BNP

Decision cutpoint

Case series (n = 9), cohort (n = 1)

Case series (n = 9), cohort (n = 1)

Specificity

Case series (n = 9), cohort (n = 1), unknown (n = 1)

Specificity

Sensitivity

Case series (n = 9), cohort (n = 1), unknown (n = 1)

Case series (n = 7), cohort (n = 1)

Specificity

Sensitivity

Case series (n = 7), cohort (n = 1)

Case series (n = 9), cohort (n = 1)

Specificity

Sensitivity

Case series (n = 9), cohort (n = 1)

Case series (n = 7), cohort (n = 1)

Specificity

Sensitivity

Case series (n = 7), cohort (n = 1)

Study design

Sensitivity

Outcome

n = 3,439

n = 3,439

n = 3,321

n = 3,321

n = 3,089

n = 3,089

n = 3,439

n = 3,439

n = 2,319

n = 2,319

# of patients

Low

Low

Low

Low

Low

Low

Low

Low

Low

Low

Risk of bias

Inconsistent

Consistent

Inconsistent

Consistent

Inconsistent

Consistent

Inconsistent

Consistent

Inconsistent

Consistent

Consistency

Direct

Direct

Direct

Direct

Direct

Direct

Direct

Direct

Direct

Direct

Directness

Imprecise

Imprecise

Imprecise

Imprecise

Imprecise

Imprecise

Imprecise

Imprecise

Imprecise

Imprecise

Precision

No evidence

No evidence

No evidence

No evidence

No evidence

No evidence

No evidence

No evidence

No evidence

No evidence

Publication bias

Moderate

Moderate

Moderate

High

Moderate

High

Moderate

High

Moderate

High

Evidence for outcome

Table 3 GRADE SOE estimates of primary outcomes, sensitivity, and specificity based on the investigated decision cutpoints for BNP and NT-proBNP

Moderate

Moderate

Moderate

High

Moderate

High

Moderate

High

Moderate

High

Overall GRADE

(0.41–0.60)

0.50

(0.85–0.95)

0.90

(0.42–0.75)

0.58

(0.79–0.93)

0.86

(0.50–0.85)

0.67

0.73 (0.63–0.84)

(0.42–0.66)

0.54

(0.77–0.92)

0.84

(0.43–0.80)

0.61

(0.71–0.89)

0.8

Effect size

Heart Fail Rev (2014) 19:439–451 447

123

448 Table 4 Diagnostic accuracy summary of studies directly comparing the diagnostic performance of BNP and NTproBNP

Heart Fail Rev (2014) 19:439–451

Study

Study population

Comparator

Assay

Manufacturer

AUC (95 % CI)

Fuat et al. [19]

Suspected HF (n = 297)

LVSD, EF \40 %

BNP

BiositeTriage

0.79

NTproBNP

Roche Elecsys

0.81

BNP

BiositeTriage

0.84 (0.79–0.89)

NTproBNP

Roche Elecsys

0.85 (0.81–0.90)

BNP

ADVIACentaur

0.909

Zaphiriou et al. [20]

Park et al. [21]

Suspected HF (n = 306)

Dyspnea or chest discomfort (n = 1,032)

Diagnosis of HF

LVEF \45 % Advanced DD ADD, LVEF [ 45 % LVEF \45 % Advanced DD

NTproBNP

Roche Elecsys

ADD, LVEF [45 % Christenson et al. [22]

\25 kg/m2

Dyspnea (n = 675)

Intermediate risk of HF (n = 111)

higher AUC for NT-proBNP, while the others showed marginally lower AUC. A single study compared two BNP assays (AxSYM & Centaur) to one NT-proBNP (Roche Elecsys) and produced similar AUC ranging from 0.76 to 0.81 [29]. Regardless of the assay or peptide, the overall diagnostic accuracy, as represented by the AUC, is functionally equivalent.

Limitations This review examined the evidence for the use of BNP and NT-proBNP in the diagnosis of HF. However, the effect of BNP and NT-proBNP as part of ‘‘test panels’’ or in combination with other diagnostic algorithms was not investigated. The effect of heterogeneity among the studies on the overall estimates of diagnostic performance was not

123

0.893 0.879 0.796

BNP

BiositeTriage

0.78 (0.71–0.084)

25–30 kg/m2

0.62 (0.54–0.70)

[30 kg/m2

0.72 (0.66–0.79)

\25 kg/m2

Kelder et al. [23]

0.897 0.806

NTproBNP

Roche Elecsys

0.77 (0.70–0.84)

25–30 kg/m2

0.64 (0.56–0.20)

[30 kg/m2

0.71 (0.65–0.77)

Diagnosis of HF

BNPAxSYM

Abbott AxSYM

0.81 (0.73–0.87)

BNPCentaur

ADVIACentaur

0.80 (0.73–0.86)

NTproBNP

Roche Elecsys

0.76 (0.69–0.82)

investigated. Mastandrea et al. [44] examined factors that can contribute to heterogeneity of meta-analyses of studies using BNP and NT-proBNP. He examined 98 samples from 67 studies (52 samples/41 studies of BNP, 46 samples/24 studies of NT-proBNP) and found that disease severity, disease prevalence, and the reference test were factors that contributed to heterogeneity for BNP. Whereas disease severity is an intrinsic factor in the pathology of the disease, prevalence and the reference test were considered true elements of heterogeneity. For NT-proBNP, Mastandrea et al. were unable to identify factors contributing to heterogeneity. One study [45] for BNP used the echocardiogram as the sole criterion for the reference test in the diagnosis of HF. All others used a combination of signs, symptoms, objective criteria (e.g., X-ray, electrocardiogram, echocardiogram), and diagnostic scorecards [e.g., Framingham,

Heart Fail Rev (2014) 19:439–451

Boston, National Health And Nutritional Examination Survey (NHANES)]. Similarly, for NT-proBNP, one study [46] used echocardiogram as the sole diagnostic criterion. All others used the same global criteria as BNP. The lack of a single ‘‘gold standard’’ for the diagnosis of HF necessitates the use of the clinical diagnosis and thus contributes to the variability of results seen across the different studies.

449

2.

3.

Conclusions In primary care settings, the majority of patients do not present to general practitioners with obvious symptoms of HF. Patients often present with limited symptoms or subclinical disease. Identification of patients at risk of developing HF or those with subclinical or limited symptoms is critical because effective treatments for HF do exist. Furthermore, the condition will progress without treatment in undiagnosed patients, thereby increasing the cost to the healthcare system and decreasing patient quality of life. Both BNP and NT-proBNP have good diagnostic performance in primary care settings for identifying patients who are either at risk of developing HF or who have fewer symptoms or less severe signs suggestive of HF. Using the manufacturers’ suggested cutpoint, BNP can effectively be used to rule out the presence of HF in primary care settings. In the case of NT-proBNP, limited evidence exists for using the manufacturers’ suggested cutpoint. We rated the SOE for sensitivity as high and specificity as moderate. Overall, we rated the SOE as high, as we feel it is unlikely further studies will change the conclusions presented here.

4.

5.

6.

7.

8.

9.

10. Acknowledgments This manuscript is based on the research conducted by the McMaster Evidence-based Practice Center (EPC) under contract to the Agency for Healthcare Research and Quality (AHRQ), Rockville, MD (Contract No. HHSA 290 2007-10060-I). The findings and conclusions in this paper are those of the authors, who are responsible for its content, and do not necessarily represent the views of the Agency for Healthcare Research and Quality. No statement herein should be construed as an official position of the Agency for Healthcare Research and Quality or of the U.S. Department of Health and Human Services. Parminder Raina holds a Tier 1 Canada Research Chair in Geroscience and the Raymond and Margaret Labarge Chair in Research and Knowledge Application for Optimal Aging. Conflict of interest Ronald A. Booth, Stephen A. Hill, Andrew Don-Wauchope, P. Lina Santaguida, Mark Oremus, Robert McKelvie, Cynthia Balion, Judy A. Brown, Usman Ali, Amy Bustamam, Nazmul Sohel, and Parminder Raina have no conflicts of interest or financial ties to disclose.

References 1. (1992) Effect of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left

11.

12.

13.

14.

ventricular ejection fractions. The SOLVD Investigattors. N Engl J Med 327:685–691. doi:10.1056/NEJM199209033271003 Balion C, Santaguida PL, Hill S, Worster A, McQueen M, Oremus M, McKelvie R, Booker L, Fagbemi J, Reichert S, Raina P (2006) Testing for BNP and NT-proBNP in the diagnosis and prognosis of heart failure. AHRQ, Rockville. AHRQ Publication No. 06-E014 Jessup M, Abraham WT, Casey DE, Feldman AM, Francis GS, Ganiats TG et al (2009) 2009 focused update: ACCF/AHA Guidelines for the diagnosis and management of heart failure in adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circ Cardiovasc Qual Outcomes 119:1977–2016. doi:10.1161/CIRCULATIONAHA.109. 192064 McKelvie RS, Moe GW, Ezekowitz JA, Heckman GA, Costigan J, Ducharme A et al (2013) The 2012 Canadian Cardiovascular Society heart failure management guidelines update: focus on acute and chronic heart failure. Can J Cardiol 29:168–181. doi:10.1016/j.cjca.2012.10.007 Scottish Intercollegiate Guidelines Network (SIGN) (2007) Management of chronic heart failure. 95. http://www.sign.ac.uk/ guidelines/fulltext/95/index.html National Clinical Guideline Centre (2010) Chronic heart failure: management of chronic heart failure in adults in primary and secondary care. National Institute for Health and Clinical Excellence, Manchester. NICE Clinical Guideline 108. http:// www.nice.org.uk/nicemedia/live/13099/50517/50517.pdf American Academy of Family Physicians (2011) Primary care. American Academy of Family Physicians, Leawood. http://www. aafp.org/online/en/home/policy/policies/p/primarycare.html# Parsys0002 Hanley JA, McNeil BJ (1982) The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 143:29–36 Chu H, Cole SR (2006) Bivariate meta-analysis of sensitivity and specificity with sparse data: a generalized linear mixed model approach. J Clin Epidemiol 59:1331–1332. doi:10.1016/j.jclinepi. 2006.06.011 (2014) Methods guide for effectiveness and comparative effectiveness reviews. Agency for Healthcare Research and Quality, Rockville. AHRQ Publication No. 10(14)-EHC063-EF. http:// www.effectivehealthcare.ahrq.gov/search-for-guides-reviewsand-reports/?pageaction=displayproduct&mp=1&productID=318 (2012) Methods guide for medical test reviews. Agency for Healthcare Research and Quality (AHRQ), Rockville. AHRQ Publication No. 12-EHC017. http://effectivehealthcare.ahrq.gov/ index.cfm/search-for-guides-reviews-and-reports/?pageaction= displayproduct&productid=558 Aspromonte N, Feola M, Scardovi AB, Coletta C, D’Eri A, Giovinazzo P et al (2006) Early diagnosis of congestive heart failure: clinical utility of B-type natriuretic peptide testing associated with Doppler echocardiography. J Cardiovasc Med (Hagerstown) 7:406–413. doi:10.2459/01.JCM.0000228690.40452.d3 Arques S, Roux E, Sbragia P, Ambrosi P, Taieb L, Pieri B et al (2005) Accuracy of tissue Doppler echocardiography in the emergency diagnosis of decompensated heart failure with preserved left ventricular systolic function: comparison with B-type natriuretic peptide measurement. Echocardiography 22:657–664. doi:10.1111/j.1540-8175.2005.40076.x Jeyaseelan S, Goudie BM, Pringle SD, Donnan PT, Sullivan FM, Struthers AD (2007) A critical re-appraisal of different ways of selecting ambulatory patients with suspected heart failure for echocardiography. Eur J Heart Fail 9:55–61. doi:10.1016/j. ejheart.2006.04.003

123

450 15. Mak G, Ryder M, Murphy NF, O’Loughlin C, McCaffrey D, Ledwidge M et al (2008) Diagnosis of new onset heart failure in the community: the importance of a shared-care approach and judicious use of BNP. Ir J Med Sci 177:197–203. doi:10.1007/ s11845-008-0186-3 16. Macabasco-O’Connell A, Meymandi S, Bryg R (2010) B-type natriuretic peptide (BNP) is useful in detecting asymptomatic left ventricular dysfunction in low-income, uninsured patients. Biol Res Nurs 11:280–287. doi:10.1177/1099800409339625 17. Barrios V, Llisterri JL, Escobar C, Alfaro P, Colado F, Ridocci F et al (2011) Clinical applicability of B-type natriuretic peptide in patients with suspected heart failure in primary care in Spain: the PANAMA study. Expert Rev Cardiovasc Ther 9:579–585. doi:10.1586/erc.11.28 18. Murtagh G, Dawkins IR, O’Connell R, Badabhagni M, Patel A, Tallon E et al (2012) Screening to prevent heart failure (STOPHF): expanding the focus beyond asymptomatic left ventricular systolic dysfunction. Eur J Heart Fail 14:480–486. doi:10.1093/ eurjhf/hfs030 19. Fuat A, Murphy JJ, Hungin AP, Curry J, Mehrzad AA, Hetherington A et al (2006) The diagnostic accuracy and utility of a B-type natriuretic peptide test in a community population of patients with suspected heart failure. Br J Gen Pract 56:327–333 20. Zaphiriou A, Robb S, Murray-Thomas T, Mendez G, Fox K, McDonagh T et al (2005) The diagnostic accuracy of plasma BNP and NTproBNP in patients referred from primary care with suspected heart failure: results of the UK natriuretic peptide study. Eur J Heart Fail 7:537–541. doi:10.1016/j.ejheart.2005.01. 022 21. Park HJ, Baek SH, Jang SW, Kim DB, Shin DI, Shin WS et al (2010) Direct comparison of B-type natriuretic peptide and N-terminal pro-BNP for assessment of cardiac function in a large population of symptomatic patients. Int J Cardiol 140:336–343. doi:10.1016/j.ijcard.2008.11.107 22. Christenson RH, Azzazy HM, Duh SH, Maynard S, Seliger SL, deFilippi CR (2010) Impact of increased body mass index on accuracy of B-type natriuretic peptide (BNP) and N-terminal proBNP for diagnosis of decompensated heart failure and prediction of all-cause mortality. Clin Chem 56:633–641. doi:10. 1373/clinchem.2009.129742 23. Kelder JC, Cramer MJ, Verweij WM, Grobbee DE, Hoes AW (2011) Clinical utility of three B-type natriuretic peptide assays for the initial diagnostic assessment of new slow-onset heart failure. J Card Fail 17:729–734. doi:10.1016/j.cardfail.2011.04. 013 24. Goode KM, Clark AL, Bristow JA, Sykes KB, Cleland JG (2007) Screening for left ventricular systolic dysfunction in high-risk patients in primary-care: a cost–benefit analysis. Eur J Heart Fail 9:1186–1195. doi:10.1016/j.ejheart.2007.10.003 25. Goode KM, Clark AL, Cleland JG (2008) Ruling out heart failure in primary-care: the cost-benefit of pre-screening using NTproBNP and QRS width. Int J Cardiol 130:426–437. doi:10.1016/ j.ijcard.2007.08.131 26. Gustafsson F, Badskjaer J, Hansen FS, Poulsen AH, Hildebrandt P (2003) Value of N-terminal proBNP in the diagnosis of left ventricular systolic dysfunction in primary care patients referred for echocardiography. Heart Drug 3:141–146. doi:10.1159/ 000073839 27. Gustafsson F, Steensgaard-Hansen F, Badskjaer J, Poulsen AH, Corell P, Hildebrandt P (2005) Diagnostic and prognostic performance of N-terminal ProBNP in primary care patients with suspected heart failure. J Card Fail 11:S15–S20. doi:10.1016/j. cardfail.2005.04.022 28. Hobbs FD, Davis RC, Roalfe AK, Hare R, Davies MK (2004) Reliability of N-terminal proBNP assay in diagnosis of left

123

Heart Fail Rev (2014) 19:439–451

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41. 42.

43.

ventricular systolic dysfunction within representative and high risk populations. Heart 90:866–870. doi:10.1136/hrt.2003.014258 Kelder JC, Cramer MJ, van Wijngaarden J, van Tooren R, Mosterd A, Moons KG et al (2011) The diagnostic value of physical examination and additional testing in primary care patients with suspected heart failure. Circ Heart Fail 124:2865–2873. doi:10. 1161/CIRCULATIONAHA.111.019216 Koschack J, Scherer M, Luers C, Kochen MM, Wetzel D, Kleta S et al (2008) Natriuretic peptide vs. clinical information for diagnosis of left ventricular systolic dysfunction in primary care. BMC Fam Pract 9:14. doi:10.1186/1471-2296-9-14 Lim TK, Dwivedi G, Hayat S, Collinson PO, Senior R (2007) Cost effectiveness of the B type natriuretic peptide, electrocardiography, and portable echocardiography for the assessment of patients from the community with suspected heart failure. Echocardiography 24:228–236. doi:10.1111/j.1540-8175.2007. 00398.x Mikkelsen KV, Bie P, Moller JE, Videbaek L, Villadsen HD, Haghfelt T (2006) Neurohormonal activation and diagnostic value of cardiac peptides in patients with suspected mild heart failure. Int J Cardiol 110:324–333. doi:10.1016/j.ijcard.2005.08. 003 Nielsen LS, Svanegaard J, Klitgaard NA, Egeblad H (2004) N-terminal pro-brain natriuretic peptide for discriminating between cardiac and non-cardiac dyspnoea. Eur J Heart Fail 6:63–70. doi:10.1016/j.ejheart.2003.10.003 Olofsson M, Boman K (2010) Usefulness of natriuretic peptides in primary health care: an exploratory study in elderly patients. Scand J Prim Health Care 28:29–35. doi:10.3109/ 02813430903345074 Shelton RJ, Clark AL, Goode K, Rigby AS, Cleland JG (2006) The diagnostic utility of N-terminal pro-B-type natriuretic peptide for the detection of major structural heart disease in patients with atrial fibrillation. Eur Heart J 27:2353–2361. doi:10.1093/ eurheartj/ehl233 Sivakumar R, Wellsted D, Parker K, Lynch M, Ghosh P, Khan SA (2006) Utility of N terminal pro brain natriuretic peptide in elderly patients. Postgrad Med J 82:220–223. doi:10.1136/pgmj. 2005.037051 Stahrenberg R, Edelmann F, Mende M, Kockska¨mper A, Du¨ngen HD, Lu¨ers C et al (2010) The novel biomarker growth differentiation factor 15 in heart failure with normal ejection fraction. Eur J Heart Fail 12:1309–1316. doi:10.1093/eurjhf/hfq151 Valle R, Aspromonte N, Barro S, Canali C, Carbonieri E, Ceci V et al (2005) The NT-proBNP assay identifies very elderly nursing home residents suffering from pre-clinical heart failure. Eur J Heart Fail 7:542–551. doi:10.1016/j.ejheart.2004.07.005 Mant J, Doust J, Roalfe A, Barton P, Cowie M, Glasziou P et al (2009) Systematic review and individual patient data meta-analysis of diagnosis of heart failure, with modelling of implications of different diagnostic strategies in primary care. Health Technol Assess 13:1–207. doi:10.3310/hta13320 Hildebrandt P, Collinson PO, Doughty RN, Fuat A, Gaze DC, Gustafsson F et al (2010) Age-dependent values of N-terminal pro-B-type natriuretic peptide are superior to a single cut-point for ruling out suspected systolic dysfunction in primary care. Eur Heart J 31:1881–1889. doi:10.1093/eurheartj/ehq163 Hall C (2005) NT-ProBNP: the mechanism behind the marker. J Card Fail 11:S81–S83. doi:10.1016/j.cardfail.2005.04.019 Charles CJ, Espiner EA, Richards AM (1993) Cardiovascular actions of ANF: contributions of renal, neurohumoral, and hemodynamic factors in sheep. Am J Physiol 264:R533–R538 Bursi F, Weston SA, Redfield MM, Jacobsen SJ, Pakhomov S, Nkomo VT et al (2006) Systolic and diastolic heart failure in the community. JAMA 296:2209–2216. doi:10.1001/jama.296.18. 2209

Heart Fail Rev (2014) 19:439–451 44. Mastandrea P (2008) Some heterogeneity factors affecting the B-type natriuretic peptides outcome: a meta-analysis. Clin Chem Lab Med 46:1687–1695. doi:10.1515/CCLM.2008.348 45. Choi S, Park D, Lee S, Hong Y, Kim S, Lee J (2007) Cut-off values of B-type natriuretic peptide for the diagnosis of congestive heart failure in patients with dyspnoea visiting emergency departments: a study on Korean patients visiting emergency

451 departments. Emerg Med J 24:343–347. doi:10.1136/emj.2006. 041368 46. Oh J, Kang SM, Hong N, Choi JW, Lee SH, Park S et al (2009) Relation between red cell distribution width with echocardiographic parameters in patients with acute heart failure. J Card Fail 15:517–522. doi:10.1016/j.cardfail.2009.01.002

123