.................................................................................................................................................................................................................
.................................................................................................................................................................................................................
RESEARCH LETTERS
doi:10.1002/ejhf.1111
Online publish-ahead-of-print 4 January 2018
How accurate is clinical
assessment of neck veins in
the estimation of central
venous pressure in acute
heart failure? Insights from a
prospective study
Medical history and physical examination are
the primary diagnostic tools when assess-
ing emergency department (ED) patients
with suspected acute heart failure (AHF).
Among physical signs, positive hepato-jugular
refux (HJR) and jugular vein distention (JVD)
are considered to indicate elevated central
venous pressure (CVP). Both are major Fram-
ingham heart failure criteria and evaluated
with priority in dyspnoeic patients presenting
to the ED.
Unfortunately and in contrast to common
perception, the accuracy of clinical neck vein
assessment in estimating CVP in AHF patients
presenting to the ED is largely unknown. This
is a dilemma as treatment decisions are com-
monly based on estimating CVP from assess-
ing neck veins in AHF.
The recent development and validation of a
non-invasive forearm vein compression ultra-
sound technique to reliably measure CVP
1,2
allowed us to address this gap in knowledge
and to evaluate the diagnostic accuracy of clin-
ical neck vein examination for the detection of
elevated CVP in AHF patients at ED presen-
tation.
This sub-study of the Basics in Acute Short-
ness of Breath Evaluation Study ( ClinicalTrials
.gov identifer: NCT01831115) prospectively
enrolled adult AHF patients at the time of ED
presentation. Only patients with a fnal adjudi-
cated diagnosis of AHF were included in this
analysis. The study was approved by the local
ethics committee, and patients gave written
informed consent.
At the time of presentation a physical
examination was performed and doc-
umented by the treating ED physician
using the standardized case report form
used universally at the University Hos-
pital Basel. Findings of the examination
of the jugular veins were categorized
as: normal (HJR–/JVD–), distended
Figure 1 Box plots displaying central venous pressure (CVP) levels according to clinical
neck vein examination. Comparison between groups by Jonckheere–Terpstra test. HJR+,
patients with distended neck veins after provocation with maintained abdominal pressure;
JVD+, patients with distended neck veins without provocation; JVD–/HJR–, patients with
normal neck veins. HJR, hepato-jugular refux; JVD, jugular vein distention.
after provocation by HJR (HJR+), distended
at rest (JVD+). Physicians were blinded to
the measurements of CVP and were not
aware of the research aim to compare their
assessment to the actual measurement of
CVP. Immediately after physical examination
by the treating ED physician, CVP was mea-
sured by a vascular specialist as described
previously.
1,2
The vascular specialist perform-
ing compression sonography was blinded to
the clinician’s neck vein assessment and was
not directly involved in routine patient care.
Measurement of CVP was only recorded for
the purpose of this study and the clinical care
team remained blinded to its results.
Overall, 217 patients were included in
the analysis and CVP measurements were
performed within a median of 2.8 h after
ED presentation. The study population was
elderly (median age: 80 years). At presen-
tation, median B-type natriuretic peptide
(BNP) blood concentration was 1150 pg/mL
and median echocardiographic left ventric-
ular ejection fraction was 39%, with 51%
of patients suffering from heart failure with
reduced ejection fraction. Tricuspid annu-
lar plane systolic excursion in the overall
population was mildly impaired [16 mm,
interquartile range (IQR) 13–20].
ED physicians classifed neck veins as
normal in 67 (31%), HJR+ in 26 (12%), and
JVD+ in 124 (57%) patients. HJR+ and JVD+
patients differed only in the frequency of
peripheral oedema (71% vs. 65%, P = 0.02).
However, HJR+ or JVD+ patients more
frequently presented with other clinical signs
of AHF, tended to have higher BNP blood
concentrations, and more frequently had
already received outpatient diuretic therapy.
In a logistic regression analysis, only AHF
signs [peripheral oedema: odds ratio (OR)
3.48, P < 0.01; rales: OR 2.07, P = 0.02; con-
gestion on X-ray: OR 2.06, P = 0.02; lgBNP:
OR 2.29, P = 0.04; outpatient diuretics: OR
2.18, P = 0.02] and lower body weight (OR
0.98, P = 0.04) but not CVP (P = 0.71) were
signifcantly associated with HJR+ or JVD+
documentation by the ED physician.
Median CVP was 8.8 mmHg (IQR
5.9– 12.6). CVP values in the three groups
were identical [HJR–/JVD–: median 8.8 (IQR
6.7– 13.3) mmHg; HJR+: 7.4 (IQR 5.0– 13.5)
mmHg; JVD+: 9.6 (IQR 5.9-12.6) mmHg]
(P = 0.65) (Figure 1). Sensitivity and specifcity
of HJR+/JVD+ to detect elevated CVP was
poor. HJR+/JVD+ achieved a sensitivity of
68.5% and a specifcity of 28.2% to detect
CVP levels above 5 mmHg. Diagnostic accu-
racies for HJR+ (sensitivity 11.2%, specifcity
84.6%) and JVD+ (sensitivity 57.3%, specifcity
43.6%) were similarly poor for the detection
of CVP elevations above 5 mmHg. Sensitivity
© 2018 The Authors
European Journal of Heart Failure © 2018 European Society of Cardiology