Reliability of Conventional Conventional echocardiography - - PDF document

SLIDE 1

3/12/2019 1

Reliability of Conventional Echocardiography Assessment of Pulmonary Hypertension in preterm infants

Hythem Nawaytou, MBBCh, MSc Assistant Professor of Pediatric Cardiology UCSF

Conventional echocardiography

• Subjective assessment e.g. RV size, septal flattening
• Measurements utilizing 2D images or Doppler (measurement of

velocity by echo) e.g RV area, TR jet

Reliability:

• the quality of being trustworthy or of performing consistently well
• the degree to which the result of a measurement, calculation, or specification

can be depended on to be accurate

Points of discussion

• Does my patient have pulmonary vascular disease? Are they a high

risk patient?

• How bad is their pulmonary vascular disease?
• How is the pulmonary vascular disease affecting the heart function?
• Did you look at the pulmonary veins?

Screening for pulmonary vascular disease

SLIDE 2

3/12/2019 2

How is pulmonary vascular disease defined?

Pulmonary Hypertension Pulmonary Vascular Resistance

Are they the same? Which one is better?

Pressure = Flow x Resistance

Flow is not a constant High pressure and low resistance e.g PDA and VSD Low pressure and high resistance e.g. RV failure Changes in resistance are not associated with proportional changes in pressure

Pressure definition versus resistance definition

PVR 3.2 ‐6WU Steurer et al, Pediatr. Pulmonol., in press 35% of the cohort would be classified differently between the two definitions

• PVR is a better determinant
• f mortality
• No added value to using a

pressure criteria Steurer et al, Pediatr. Pulmonol., in press

SLIDE 3

3/12/2019 3

Different definitions of PVD gives different results of echo reliability

5 10 15 20 25 30 35 40 45 50 55 60 65 70 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 M ean pulmonary artery pressure (mmHg) Indexed pulmonary vascular resistance (WU x m2 ) True negative False positive True positive False negative n=29 Agreement = 72% PVR only Definition

A

5 10 15 20 25 30 35 40 45 50 55 60 65 70 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean pulmonary artery pressure (mmHg) Indexed pulmonary vascular resistance (WU x m2 ) True negative False positive True positive False negative WHO Definition n=29 Agreement = 62%

B

POSITIVE ECHO TR jet velocity >2.9m/sec OR VSD or PDA systolic flow velocity estimating peak systolic PAP > 35mmHg OR Systolic septal flattening was present

Mourani, Pediatrics 2008;121:317-325

Subjectively RA enlargement RV dilation RV hypertrophy PA dilation Septal flattening

Accuracy of echo PAP estimation using PA pressure to define PVD

TR Jet Septal Flattening

Accuracy of echo PAP estimation using PVR to define PVD

Accuracy Sensitivity Specificity AUC Confidence interval Positive echocardiogram for PH 72 90.5 25 0.58 0.39‐0.76 Septal flattening 69 85.7 25 0.61 0.42‐0.79 Positive echocardiogram for PH excluding patients with PDA (n=18) 89 93.3 66.7 0.8 0.52‐0.94 Septal flattening excluding patients with PDA (n=18) 83 86.7 66.7 0.8 0.52‐0.94 Positive echocardiogram for PH excluding patients with PDA & VSD (n=14) 93 91.7 100 0.96 0.66‐0.99 Septal flattening excluding patients with PDA & VSD (n=14) 93 91.7 100 0.96 0.66‐0.99

POSITIVE ECHO TR jet velocity >2.9m/sec OR VSD or PDA systolic flow velocity estimating peak systolic PAP > 35mmHg OR Systolic septal flattening was present PVD DEFINED AS PVRi >3WU

What to do for patients with a shunt?

Shunt Direction of flow (N) P value PVRi median (Q1,Q2) PDA L to R (n=8) Bidirectional (n=2) R to L (n=1) 0.18 4.3 (2.5,5.5) 1.7 /3 8.9 VSD L to R (n=3) Bidirectional (n=1) R to L (n=1) 0.2 2.8‐4.6 12.6 8.9

A left to right shunt does not indicate absence of PVD

SLIDE 4

3/12/2019 4

Conclusion for screening PVD

• Echo is a good screening test for PH and elevated PVR, but be

prepared for some negative catheterization results

• A positive echo is reliable for elevated PVR in patients without PDA

and VSD.

• Echo will never rule out PVD in preterm infants because we don’t

catheterize patients with negative echoes, hence we don’t know the accuracy of a negative study.

Severity of pulmonary vascular disease

Ability of echocardiography (ECHO)-estimated sPAP to predict the severity of PH determined with cardiac catheterization (CATH).

• Mourani. Pediatrics 2008;121:317-325

11% UNDERDIAGNOSIS 11% OVERDIAGNOSIS A B

Groh , J Am Soc Echocardiogr (2014) 27(2) 163–71 Ge, Int J Cardiol (1993) 40(1):35–43) Ge, Am Heart J (1992) 124(1):176–82) Ge, Clin. Cardiol. 15, 818‐824 (1992)

TR Jet = 40mmHg PI Jet = 50mmHg PDA = 30mmHg

Reliability of Doppler estimation of severity of pulmonary hypertension Limits of agreement are wide (up to 40‐ 50 mmHg) Wider limits at higher pressures

SLIDE 5

3/12/2019 5

Kappa Kappa Kappa Free assessment of right ventricle (RV) Decrease categories of reporting Consensus criteria of measurable variables only RV pressure

0.3

RV pressure

0.58

RV pressure

0.65

Normal

0.17

< 50% systemic

0.4

< 50% systemic

0.67

<50% systemic

0.29

50% systemic

0.11

50% systemic

0.44

50% systemic

0.53

>50% systemic

0.33

Systemic

0.36

Systemic

0.89

Systemic

0.77

Suprasystemic

0.45

Inter‐rater agreement: an echocardiographer’s curse

• Inter‐rater agreement improves by reporting on less categories and agreeing about criteria used.
• Strong agreement between echocardiographers as to what constitutes systemic

and supra‐systemic pressure on an echo.

Accuracy

Rater2 | 1 3 6 | Total ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ 1 | 2 3 0 | 5 3 | 1 7 0 | 8 6 | 0 2 5 | 7 ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ Total | 3 12 5 | 20 Rater1 | 1 3 6 | Total ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ 1 | 2 2 0 | 4 3 | 1 8 0 | 9 6 | 0 2 5 | 7 ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ Total | 3 12 5 | 20 Rater3 | 1 3 6 | Total ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ 1 | 3 4 0 | 7 3 | 0 6 0 | 6 6 | 0 2 5 | 7 ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ Total | 3 12 5 | 20 Rater4 | 1 3 6 | Total ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ 1 | 2 1 0 | 3 3 | 1 10 0 | 11 6 | 0 1 5 | 6 ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ Total | 3 12 5 | 20 Rater5 | 1 3 6 | Total ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ 1 | 3 5 0 | 8 3 | 0 7 2 | 9 6 | 0 0 2 | 2 ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ Total | 3 12 4 | 19 Rater6 | 1 3 6 | Total ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ 1 | 2 3 0 | 5 3 | 1 8 0 | 9 6 | 0 1 5 | 6 ‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐+‐‐‐‐‐‐‐‐‐‐ Total | 3 12 5 | 20

Cath derived RV/SBP ratio Cath derived RV/SBP ratio Echo derived RV/SBP ratio 1 = Normal or <50% 3 = 50% or >50% 6 = systemic or suprasystemic Echocardiographers are quite accurate in grading systemic or supra‐systemic PA pressure

Conclusion on severity of PH

• Echo is unreliable in estimating PH severity.
• Limiting the categories of PH reported on and the number of

echocardiographers in the lab reporting on PH patients maybe desirable.

• Echocardiographer agree on systemic or supra‐systemic PA pressure

and are usually accurate. (for once)

• The reliability of echo in determining degree of PVR elevation in

preterm infants is under studied but probably unreliable

Are assessments of RV dimension and function reliable

SLIDE 6

3/12/2019 6

RV dimensions and function

• There are no MRI –echo correlation data in preterm BPD patients.
• There are no pressure volume loop or high fidelity cardiac

catheterization‐ echo correlation data in preterm BPD patients.

• In the absence of gold standard data to report accuracy, I will test

reliability using consistency between readers

• Correlation between echo‐cardiac catheterization measurements of

cardiac output and diastolic pressures

Kappa Kappa Kappa Free assessment of right ventricle (RV) Decrease categories of reporting Consensus criteria of measurable variables only RV dilation

0.38

RV dilation

0.62

RV dilation

0.7

Normal

0.55

Normal/mild

0.62

No significant

0.7

Mild

0.32

Moderate

0.35

Moderate/severe Significant Severe

0.34

RV hypertrophy

0.2

RV hypertrophy

0.43

RV hypertrophy

0.22

Normal

0.28

Normal/mild

0.43

No significant

0.22

Mild

0.06

Moderate

0.24

Moderate/severe Significant Severe

0.29

RV dysfunction

0.37

RV dysfunction

0.4

RV dysfunction

0.13

Normal

0.59

Normal/mild

0.59

No significant

0.13

Mild

0.19

Moderate

0.13

Moderate

0.13

Present Severe

0.39

Severe

0.39

Severe

none

Inter‐rater agreement

• Inter‐rater agreement can be improved by limiting reporting categories and possibly measuring the RV dimensions
• Measurements don’t always help

Correlations between echo & cath derived cardiac

• utput

RV LV RV

No correlation between non‐simultaneous echo‐cath correlation in BPD patients. Wurzer et al, Shock. 2016 September ; 46(3): 249–253.

Conventional echo is unreliable in assessment of RV or LV output

Moderate correlation between RV output estimated by echo and thermodilution in older children but with very large limits

• f agreement. (‐3L/min to +6L/min)

Khemani et al (n=13) : LAP 10.8 ± 3.1 Streuer et al (n=29): LAP 9 ± 3.1 / RVEDP 8.7 ± 3.2 Mourani et al (n=31): PCWP 10 ± 3 Del Cerro et al (n=14): PCWP 10.6 ± 2.3, LVEDP 11.7 ± 2.8 Berman et al (n=9): normal left heart

Correlations between echo surrogates & cath measured LA pressure / PCWP/ RVEDP

Echo Cath Spearman correlation p value Mitral E wave velocity (m/sec) 0.8 (0.74, 1.0) LA pressure/PCWP 0.23 Mitral E/A ratio 1.0 (0.83, 1.1) LA pressure/PCWP 0.37 Mitral SDI 1.7 (1.5,1.9) LA pressure/PCWP 0.53 Lateral mitral e’ (cm/sec) 8 (5,10) LA pressure/PCWP 0.76 Mitral E/e’ 10.7 (8.4, 13.5) LA pressure/PCWP 0.62 tricuspid E wave velocity (m/sec) 0.9 (0.55,1.1) RVEDP 0.19 Tricuspid E/A ratio 0.8 (0.7, 1.0) RVEDP 0.62 Tricuspid SDI 1.8 (1.6, 2.0) RVEDP 0.85 Lateral tricuspid e’ (cm/sec) 14 (9,20) RVEDP 0.3 Tricuspid E/e’ 5.8 (5.6, 6.9) RVEDP 0.5

13/29 fused mitral inflows 17/29 fused tricuspid inflows

Khemani et al, Pediatrics 2007;120;1260 Streuer et al, Pediatr Pulmonol, in press

• Mourani. Pediatrics 2008;121:317‐325

Del Cerro et al, Pediatr Pulmonol. 2014; 49:49–59 Berman et al, Pediatrics 1982;70(5):708‐12

Despite abnormal echocardiographic indices, The diastolic pressures on cath are not elevated And there is no correlation between diastolic pressure on cath and echo indices of diastolic dysfunction

SLIDE 7

3/12/2019 7

Reliability of echo in diagnosing pulmonary vein stenosis

• Many criteria indicating no good criteria
• Only 56% were diagnosed by echo

Mahgoub et al, Pediatric Pulmonology. 2017;52:1063–1070

Criteria of PVS in Literature

Turbulent flow Monophasic flow Reversal of A wave Flow not reaching baseline Flow not reversing in late diastole Gradient / velocity Peak Velocity > 1.5 m/s Peak Velocity > 1.6 m/s Mean velocity > 2 m/s Mean pressure >5 mmHg Discrepancy between branch PA size

Thanks