DOI: 10.1111/echo.

13350

O R I G I N A L I N V E S T I G AT I O N

Early detection of right ventricular dysfunction using

transthoracic echocardiography in ARDS: a more objective
approach

Subeer Kanwar Wadia M.D., M.Sc.|Trushil G. Shah M.D., M.B.B.S., M.Sc.|

Grady Hedstrom M.D.|Julie A. Kovach M.D., F.A.C.C., F.A.S.E.|Rajive Tandon M.D.

Rush University Medical Center, Chicago,

Illinois
Correspondence
Subeer Kanwar Wadia, M.D., M.Sc., Rush
University Medical Center, Chicago, IL, USA.
Email: Subeer_Wadia@rush.edu
Purpose: Right ventricular (RV) dysfunction is an independent predictor of morbidity
and mortality in acute respiratory distress syndrome (ARDS). Our goal was to describe
morphologic changes in the RV using objective measures on transthoracic echocardi-
ography (TTE) that occur following ARDS.
Methods: We retrospectively measured changes in the following RV parameters from

All of the authors had access to the data,
reviewed the manuscript, and approved this
submission.
a pre-ARDS TTE to an ARDS TTE: tricuspid annular plane systolic excursion (TAPSE),
myocardial performance index (MPI), fractional area change (FAC), systolic pulmonary
artery pressure (SPAP), peak tricuspid regurgitant (TR) velocity, and septal shift.
Results: Over 24months, 14 patients met inclusion/exclusion criteria. Mean TAPSE
decreased from 22.4mm pre-ARDS to 16.3mm during ARDS, P<.001. Mean MPI in-
creased from 0.19 to 0.38, P=.001. Mean FAC decreased from 60.8% to 41.2%,
P=.003. Peak TR velocity increased from 2.67m/s pre-ARDS to 3.31m/s during
ARDS, P=.02. SPAP and septal shift demonstrated trends but not statistically different
between pre-ARDS and ARDS states. TAPSE correlated with ARDS severity (PaO2/
FiO2 ratios), P=.004, and was lower among 30-day nonsurvivors compared with survi-
vors, P=.002.
Conclusions: Mild RV dysfunction is common after ARDS onset. RV morphologic
changes coupled with dysfunction can be detected noninvasively through TTE changes
with TAPSE, MPI, and FAC. Mild RV dysfunction by TAPSE is associated with ARDS
severity and mortality.

KEYWORDS
acute respiratory distress syndrome, dysfunction, right ventricle, TAPSE, transthoracic
echocardiography

1| INTRODUCTION
Since 1977, right ventricular (RV) dysfunction has been a known inde-
pendent risk factor for morbidity and mortality in acute respiratory dis-
14
tress syndrome (ARDS). Reducing RV afterload with RV-protective
ventilation and prone positioning have improved the rates of acute
cor pulmonale and mortality.59 Although pulmonary artery catheters
(PACs) were previously employed for right heart assessment in pa-
tients with ARDS, their use has diminished after the FACTT and PAC-
Man trials showed increased rates of complications without clinical
benefit.1014 The criticism against PACs leaves a gap in the accurate
evaluation of right heart function in patients with ARDS.
At the present time, no data have compared the quantified changes
that occur on noninvasive transthoracic echocardiography (TTE)

Echocardiography 2016; 16 wileyonlinelibrary.com/journal/echo 2016, Wiley Periodicals, Inc. | 1

 |
2 Wadia etal.
following onset of ARDS. The American Society of Echocardiography
15
(ASE) 2015 Chamber Guidelines recommend including quantitative
measures of RV function, yet TTE assessment of RV function in pa-
tients with ARDS remains primarily qualitative due to the limitations
in acoustic windows and the complex configuration of the RV cham-
16,17
ber. Although objective indices on TTE are being explored as re-
producible ways of assessing RV pressure or functional changes by
echocardiography,18,19 their utility in demonstrating change over time
and in producing prognostic information has not been validated in me-
chanically ventilated patients with ARDS. In this study, we sought to
identify objective parameters on TTE that can quantify change in RV
function in a patient pre-ARDS and acutely after ARDS onset. Such
data will provide clinicians with a simple and standardized tool to de-
tect early RV morphologic changes on TTE coupled with dysfunction
in patients with ARDS and will guide future interventions to improve
RV dysfunction in this population.
2| METHODS
2.1|Population
Over a 24-month period (between January 2013 to December
2014), we retrospectively analyzed patients admitted with ARDS in
the Medical Intensive Care Unit at Rush University Medical Center
in Chicago, IL. This study was performed in compliance with human
studies guidelines and approval from Institutional Review Board. The
Institutional Review Board waived the need for informed consent.
After querying the electronic medical record (EMR), patients were
selected if age was more than 18years, ARDS was the primary dis-
charge diagnosis, left ventricular (LV) ejection fraction was greater
than 50% (to avoid confounding effects of chronic left heart disease
on RV dysfunction), and TTE was performed while the patient was me-
chanically ventilated within 2weeks of ARDS admission. All patients
5
were ventilated according to the ARDSNet protocol. Only patients
with a prior echocardiogram predating ARDS within 1year of hospi-
talization for comparison were included. Echocardiograms had to be
considered good quality for imaging review and comparison. Patients
with a history of valvular heart disease, congestive heart failure, or
underlying pulmonary disease were excluded.
2.2|Data and echocardiograms
The data extracted for analysis included the following: baseline de-
mographics, medical history, indication for echocardiograms, ven-
tricular and valvular function by EMR report, arterial blood gases,
ventilator settings, and 30-day mortality from echocardiogram date.
PaO2/FiO2 (P/F) ratios were calculated using arterial blood gas and
ventilator settings on the day the echocardiogram was performed.
Each echocardiogram was initially reviewed by a cardiologist within
our echocardiography laboratory. These echocardiograms were sub-
sequently reevaluated by a single cardiologist (J.K.) to obtain values
for tricuspid annular plane systolic excursion (TAPSE), right ventricular
myocardial performance index (MPI), right ventricular fractional area
change (FAC), systolic pulmonary artery pressure (SPAP), peak tricus-
pid regurgitation (TR) velocity, and presence or absence of septal shift.
Tricuspid annular plane systolic excursion was measured using M-mode
from a standard apical four-chamber view. MPI was measured using tis-
sue Doppler imaging. FAC was obtained by tracing the RV endocardium
from a conventional LV-focused apical four-chamber view in systole and
in diastole (and presented as a percentage). SPAP was measured using es-
timated right atrial pressure [RAP] (based on inferior vena cava [IVC] size
and degree of collapse with sniff) and TR jet. A separate analysis was done
using only peak TR velocity (derived from the simplified Bernoulli equa-
tion). Septal shift was assessed by observing for D-shaped distortion of
the left ventricle during systole in the short-axis view. In accordance with
the 2010 ASE guidelines,20 the highest value measured by continuous-
wave Doppler in any of the standard two-dimensional views was used and
corroborated with the initial echocardiographic findings. The independent
reviewer (J.K.) was unaware of the patients clinical status (including de-
ceased status) or severity of ARDS during TTE measurements.
2.3|Statistics
Mean differences in RV function on TTE before ARDS and during
ARDS were compared using paired t-test and Fishers exact test.
Pearson correlation coefficient was used to associate RV indices with
the severity of ARDS (P/F ratio). Significance was defined at two-
sided P-value less than .05, and 95% confidence intervals were applied
when appropriate. Data analysis was carried out on SPSS Software
(IBM Corporation, Armonk, NY, USA), and graphs were made using
Microsoft Excel (Microsoft, Redmond, WA, USA).
3|RESULTS
Over 24months, 38 patients were treated for ARDS and had good-
quality echocardiograms done while intubated within 14days of pres-
entation. Of these 38 patients, 14 patients (nine women and five men)
had incidental echocardiograms pre-dating ARDS and were included
for our study. Mean patient age was 57.6years. Each patient had an
echocardiogram prior to ARDS (mean 88days prior to ARDS hospi-
talization) which was compared to an echocardiogram done during
ARDS (mean 7days after date of admission). All 14 patients received
treatment with stress-dose steroids and intravenous antibiotics. All pa-
tients were ventilated on pressure-regulated volume control using the
ARDSNet Protocol.5 10/14 patients (71%) had an infectious etiology as
the cause of ARDS, and 30-day mortality was 57% (8/14 patients died).
Descriptive characteristics are listed in Table1. Indications for previ-
ous TTE, causes of ARDS, causes of death, and the use of vasopressor
agents on the day of the ARDS echocardiograms are listed in Table2.
Among the TTE indices studied, all demonstrated changes consis-
tent with worsening RV function or pressure overload with onset of
ARDS (Fig.1). Mean TAPSE decreased from 22.4mm (CI 19.325.5)
before ARDS to 16.3mm (13.918.7) during ARDS, P<.001. Similarly,
RV MPI increased from 0.19 (0.130.25) to 0.38 (0.290.48), P=.001,
and FAC decreased from 60.8% (57.264.4) to 41.2% (32.450.0),
Wadia etal. |
3

T A B L E 1 Descriptive characteristics impression of the reading cardiologist. Prior to ARDS, 0/14 patients
had any evidence of reduced RV function on TTE. During ARDS,
Patient Characteristics Mean ( SD) or frequency (%)
6/14 patients had either mild or moderate dysfunction (two pa-
Age (y) 57.6 (12.2)
tients with mild and four patients with moderate). The remaining
Body mass index 32.5 (14.2) 8/14 patients had normal RV function during ARDS. The difference
Echo prior to ARDS admission (d) 88 (106) between RV dysfunction pre-ARDS and ARDS was statistically signif-
Echo after ARDS admission (d) 7 (5) icant (0/14 patients with mild or moderate RV dysfunction pre-
pH 7.40 (0.09) ARDS vs 6/14 patients with mild or moderate dysfunction during
pO2 106 (27) ARDS, P=.03). Thus by visual estimation alone, mild-to-moderate RV
pCO2 47 (14) dysfunction was identified early after ARDS onset.
Bicarbonate 23 (7) As expected, the severity of ARDS (defined by the P/F ratios) had

PEEP (mL) 8 (4)

Tidal volume (mL/kg) 5.11 (1.8)
FiO2 (%) 58 (22)
PaO2/FiO2 ratio 179 (89)
30-d mortality 8/14 (57%)
P=.003. SPAP values were above normal reference ranges at base-
line but still increased from 40.2mm Hg (28.851.6) pre-ARDS to
50.2mm Hg (41.159.3) during ARDS, P=.08. Peak TR velocity in-
creased from 2.67m/s (2.362.97) at baseline to 3.31m/s (3.063.53)
during ARDS, P=.02. Septal shift was present in six patients during
ARDS compared with only one before ARDS, P=.06.
Changes in RV function pre-ARDS and during ARDS were also
analyzed using the EMR echocardiogram reports. These reports
were based on traditional subjective approach of visually estimat-
ing RV function (qualitative analysis). Function was categorized as
normal, mildly, moderately, or markedly reduced based on the
a significant association with mortality. Mean P/F ratio among survi-
vors was higher at 254 (171336) compared with 135 (91180) in
nonsurvivors, P<.001. Of the indices studied, only TAPSE was signifi-
cantly associated with increased mortality and severity of ARDS. Mean
TAPSE was lower among 30-day nonsurvivors at 14.0mm (11.716.4)
compared with 19.6mm (17.321.9) among survivors, P=.002 (Fig.2).
There was a significant positive relationship between TAPSE and P/F
ratios, showing more severe ARDS is associated with lower TAPSE,
r(10)=.76, P=.004 (Fig.3). These results remained significant after ad-
justing for PEEP, pH, and pCO2.
4|DISCUSSION
Our study showed acute RV dysfunctional changes occurring within
patients after onset of ARDS. We were able to quantify changes in
RV morphology using objective markers on TTE. The longitudinal de-
terioration in RV function with onset of ARDS was best illustrated by

T A B L E 2 Indication for initial echocardiogram, causes of ARDS, Causes of Death, and Use of Vasopressor Agents on Day of
Echocardiograms

VasoPressor on
day of ARDS
Patient Indication for prior TTE Cause of ARDS Cause of death (if applicable) TTE

1 Chest pain (not ACS) Osteomyelitis with sepsis Refractory shock No

2 Syncope Bacteremia No
3 Dyspnea at rest Pneumonia Progressive hypoxemia/shock Yes
4 Dyspnea on exertion Idiopathic No
5 Orthostatic hypotension Pancreatitis No
6 Dyspnea on exertion Pneumonia No
7 Elevated troponin (not ACS) Idiopathic PEA arrest No
8 MSSA bacteremia Chemical pneumonitis VT/VF No
9 Bilateral leg edema Pneumonia Progressive hypoxemia/shock No
10 Dyspnea at rest Rhinovirus/enterovirus Progressive hypoxemia/shock Yes
11 Chest pain (not ACS) Bacteremia No
12 ACS/myocardial infarction Pneumonia Yes
13 Syncope Spontaneous bacterial peritonitis with sepsis Refractory shock Yes
14 Bilateral leg edema Parainfluenza Progressive hypoxemia/shock No

ACS, acute coronary syndrome; MSSA, Methicillin-sensitive Staphylococcus aureus; PEA, pulseless electrical activity; VT/VF, ventricular tachycardia/
ventricular fibrillation
 |
4 Wadia etal.
F I G U R E 1 Box plots showing changes in (A) TAPSE, P<.001;
(B) MPI, P=.001; (C) FAC, P=.003; (D) SPAP, P=.08; and (E) Peak TR
Velocity, P=.02; with onset of ARDS ( = mean)
TAPSE, MPI, RV FAC, and peak TR velocity, which were measured at
an average 7days after ARDS admission and compared to an echocar-
diogram pre-dating ARDS (average 88days prior to hospitalization).
RV function (as indicated by TAPSE) was related to ARDS severity (as
indicated by P/F ratio) and mortality. The relationship between RV
dysfunction measured by TAPSE and ARDS severity or mortality was
independent of PEEP, pH, and pCO2.
F I G U R E 2 Box plot of TAPSE among 30-day survivors and non-
survivors, P=.002 ( = mean)
F I G U R E 3 Correlation between TAPSE and P/F ratio, r(10)=.76,
P=.004 ( = patient not on vasopressor agents; x=patient on
vasopressor at time of TTE)
In our study, 8/14 patients showed no evidence of RV dysfunc-
tion when evaluated qualitatively on TTE. However, more direct and
subtle changes in RV dysfunction were identified using TAPSE, MPI,
and FAC. These trends were clinically significant, as the reductions in
TAPSE were associated with higher severity of ARDS and increased
30-day mortality. Our findings of increased mortality associated with
subtle RV morphologic changes in a small population of only fourteen
patients suggest even mild RV dysfunction is clinically relevant and
leads to poorer outcomes.
Tricuspid annular plane systolic excursion, MPI, and FAC are di-
rect assessments of RV function20 and are more suitable in detect-
ing early changes in RV physiology in ARDS compared with SPAP
or septal shift. TAPSE provides an assessment of the base-to-apex
shortening of the RV and reflects longitudinal contraction of the
ventricle. MPI is a ratio of the isovolumetric and ejection times of
the RV and provides direct assessment of RV diastolic and systolic
physiology. With RV dysfunction, isovolumetric time is prolonged
and ejection time is shortened. Therefore, the MPI ratio increases
with a dysfunctional RV. Finally, RV FAC is measured by tracing
the RV endocardium during systole and diastole, and quantifying
change in area through the cardiac cycle as a percentage. With
higher pulmonary pressures, the RV will be unable to contract
Wadia etal.
effectively and FAC will be reduced. These three modalities di-
rectly evaluate RV physiology and have proven to correlate with
radionuclide and magnetic resonance-
derived evaluation of RV
function.21,22
In contrast to TAPSE, MPI, and FAC, SPAP and septal shift indi-
rectly evaluate RV function by identifying pressure-associated se-
quelae on the RV. SPAP measurements are based on the estimations
of RAP and TR gradients. However, RAP estimation using IVC col-
lapsibility is unreliable in ventilator-dependent patients.23 When our
analysis was performed using only peak TR velocity, we noted a sig-
nificant increase in right-sided pressures following ARDS onset. Our
population also had elevated SPAP in the baseline state (40.2mm Hg).
The significance of the elevated SPAP at baseline is unclear, potentially
reflecting the range of obesity seen in our population (mean BMI 32.5
with standard deviation 14.2). While no studies have yet established
a causal relationship between obesity and pulmonary hypertension (in
the absence of obstructive sleep apnea or anorexigenic drug use), we
do not know the presence or absence of sleep apnea or anorexigenic
drug use in our study group.24 The elevated SPAP in the baseline state
represents a limitation of the study, but may concede the utility of
measuring SPAP in an obese population.
We have previously shown in the larger cohort of 38 patients
that TAPSE, MPI, and septal shift correlated with ARDS severity
and death, and of these indices, TAPSE was an independent pre-
dictor of mortality.25 The data from this study expand on our pre-
vious results by showing that patients will show trends in TAPSE,
MPI, and FAC toward RV dysfunction from a pre-ARDS baseline
to an acute ARDS state. Given the high likelihood of detecting RV
changes on average 7days after ARDS onset, our results advocate
screening for RV dysfunction early in the course of ARDS with TTE
and measurements of TAPSE, MPI, and FAC. The availability of the
objective indices on two-dimensional echocardiography makes
them appealing for routine use at the bedside and provides clini-
cians with a tool to identify and to quantify early RV dysfunction
in acute ARDS.
The following limitations are recognized in our study. As specific
echocardiographic views adequate for measurement of TAPSE, MPI,
and FAC were not routinely obtained at our institution before 2013
and only patients with two complete echocardiograms for compari-
son were included, our study size of fourteen patients was small. As
the study was retrospective, we could not conclude a causal relation-
ship between mild RV dysfunction and mortality, only an association.
RV FAC was measured using conventional LV-focused apical four-
chamber view rather than RV-focused apical views. In this retrospec-
tive study, echocardiograms were recorded before publication of the
2015 ASE Chamber Quantification guidelines. For consistency, in all
instances conventional LV-focused views were selected for tracing RV
area. Finally, as mentioned previously, patients had an elevated SPAP
(40.2mm Hg) in the baseline state. The elevated SPAP in the base-
line state may either reflect the poor utility of this measurement or
preexisting pulmonary vascular disease, which worsened with onset
of ARDS. Epidemiologic studies are needed to assess mortality rates
in patients with underlying pulmonary vascular disease who develop
|
5
ARDS, as these patients may be more likely to succumb to the destruc-
tive effects of ARDS.
5|CONCLUSIONS
Our study shows (1) mild RV dysfunction is reliably identified in ARDS
when measured using TAPSE, MPI, and FAC on TTE (confirming ASE
15
guideline recommendations to include reporting of quantitative
measures of RV dysfunction), (2) RV dysfunction as measured by re-
ductions in TAPSE is associated with severity of ARDS and thirty-day
mortality, and (3) TTE using objective measurements might be useful
to evaluate RV function prospectively in patients with ARDS. Future
studies should be undertaken to determine the prognosis and poten-
tial for progressing to acute cor pulmonale in patients with ARDS who
demonstrate early echocardiographic changes on TTE.
AC KNOW L ED G M ENT
The authors report no financial acknowledgments for this study.
REFERENCES
1. Zapol W, Snider M. Pulmonary hypertension in severe acute respira-
tory failure. N Engl J Med. 1977;296:476480.
2. Monchi M, Bellenfant F, Cariou A, et al. Early predictive factors of
survival in the acute respiratory distress syndrome. A multivariate
analysis. Am J Respir Crit Care Med. 1998;158:10761081.
3. Scherrer-Crosbie M, Streckenbach SC, Zapol W. Acute cor pulmonale
in acute respiratory distress syndrome: a dreaded complication of the
past? Crit Care Med. 2001;29:16411642.
4. Vieillard-Baron A. Is right ventricular function the one that matters in
ARDS patients? Definitely yes Intensive Care Med. 2009;35:46.
5. The Acute Respiratory Distress Syndrome Network. Ventilation with
lower tidal volumes as compared with traditional tidal volumes for
acute lung injury and the acute respiratory distress syndrome. N Engl
J Med. 2000;342:13011308.
6. Vieillard-Baron A, Matthay M, Teboul JL, et al. Experts opinion on
management of hemodynamics in ARDS patients: focus on the effects
of mechanical ventilation. Intensive Care Med. 2016;42:739749.
7. Repesse X, Charron C, Vieillard-Baron A. Acute cor pulmonale in ARDS:
rationale for protecting the right ventricle. Chest. 2015;147:259265.
8. Jozwiak M, Teboul JL, Anguel N, etal. Beneficial hemodynamic effects
of prone positioning in patients with acute respiratory distress syn-
drome. Am J Respir Crit Care Med. 2013;188:14281433.
9. Jozwiak M, Monnet T, Teboul JL. Optimizing the circulation in the
prone patient. Curr Opin Crit Care. 2016;22:239245.
10. Osman D, Monnet X, Castelain V, et al. Incidence and prognostic
value of right ventricular failure in acute respiratory distress syn-
drome. Intensive Care Med. 2009;35:6976.
11. Repesse X, Charron C, Viellard-Baron A. Right ventricular failure in
acute lung injury and acute respiratory distress syndrome. Minerva
Anestesiol. 2012;78:941948.
12. Harvey S, Harrison DA, Singer M, etal. PAC-Man study collaboration.
Assessment of the clinical effectiveness of pulmonary artery cathe-
ters in management of patients in intensive care (PAC-Man): a ran-
domized controlled trial. Lancet. 2005;366:472477.
13. National Heart, Lung, and Blood Institute Acute Respiratory Distress
Syndrome (ARDS) Clinical Trials Network, Wheeler AP, Bernard
 |
6 Wadia etal.
GR, Thompson BT, et al. Pulmonary-Artery versus Central Venous
Catheter to guide treatment of acute lung injury. N Engl J Med.
2006;354:22132224.
14. Richard C, Warszawski J, Anguel N, et al. Early use of pulmonary
artery catheter and outcomes in patients with shock and acute re-
spiratory distress syndrome: a randomized controlled trial. JAMA.
2003;290:27132720.
15. Lang RM, Badano LP, Mor-Avi V, etal. Recommendations for cardiac
chamber quantification by echocardiography in adults: an update from
the American Society of Echocardiography and the European Society
of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015;28:139.
16. Vieillard-Baron A, Prin S, Chergui K, Dubourg O, Jardin F. Echo-Doppler
demonstration of acute cor pulmonale at the bedside in the medical
intensive care unit. Am J Respir Crit Care Med. 2002;166:13101319.
17. Lazzeri C, Cianchi G, Bonizzoli M, Batacchi S, Peris A, Gensini GF. The
potential role and limitations of echocardiography in acute respiratory
distress syndrome. Ther Adv Respir Dis. 2016;10:136148.
18. Fichet J, Moreau L, Genee O, et al. Feasibility of right ventricular
longitudinal systolic function evaluation with transthoracic echocar-
diographic indices derived from tricuspid annular motion: a prelimi-
nary study in acute respiratory distress syndrome. Echocardiography.
2012;29:513521.
19. Lazzeri C, Bonizzoli M, Cozzolino M, et al. Serial measurements of
troponin and echocardiography in patients with moderate-to-severe
acute respiratory distress syndrome. J Crit Care. 2016;33:132136.
20. Rudski LG, Lai WW, Afilalo J, et al. Guidelines for the echocardio-
graphic assessment of the right heart in adults: a report from the
American Society of Echocardiography. J Am Soc Echocardiogr.
2010;23:685713.
21. Anavekar NS, Gerson D, Skali H, Kwong RY, Yucel EK, Solomon
SD. Two-dimensional assessment of right ventricular function:
an echocardiographic-MRI correlative study. Echocardiography.
2007;24:452456.
22. Miller D, Farah MG, Liner A, Fox K, Schluchter M, Hoit BD. The rela-
tion between quantitative right ventricular ejection fraction and indi-
ces of tricuspid annular motion and myocardial performance. J Am Soc
Echocardiogr. 2004;17:443.
23. Fisher MR, Forfia PR, Chamera E, etal. Accuracy of Doppler echocar-
diography in the hemodynamic assessment of pulmonary hyperten-
sion. Am J Respir Crit Care Med. 2009;179:615621.
24. Dela Cruz CS, Matthay RA. Role of obesity in cardiomyopathy and
pulmonary hypertension. Clin Chest Med. 2009;30:509523.
25. Shah TG, Wadia SK, Kovach J, Fogg L, Tandon R. Echocardiographic
parameters of right ventricular function predict mortality in acute
respiratory distress syndrome: a pilot study. Pulm Circ. 2016;
6:155160.
How to cite this article: Wadia, S. K., Shah, T. G., Hedstrom, G.,
Kovach, J. A. and Tandon, R. (2016), Early detection of right
ventricular dysfunction using transthoracic echocardiography in
ARDS: a more objective approach. Echocardiography, 0: 16.
doi: 10.1111/echo.13350