Table 2 Accuracy of LAP measured by non-invasive and invasive techniques in the non-critically ill
From: Bedside assessment of left atrial pressure in critical care: a multifaceted gem
Studies | Population | Methods | Measurement | Main findings | Exclusion criteria |
|---|---|---|---|---|---|
Non-critical care studies evaluating PAOP and invasive LVEDP | |||||
Sato et al. [16] | Elective cardiac catheterisation | N = 79 Retrospective subgroup analysis of those undergoing simultaneous LHC and RHC | PAOP during RHC (method not specified) versus post-A wave LVEDP during LHC | Strong correlation, r = 0.82 p < 0.001 | ACS, AF, mitral valve surgery, mitral valve disease (stenosis, severe regurgitation or severe MAC), severe AR, prior heart transplantation. Heart rate > 100, any change in diuretic, vasodilator or antihypertensive treatment between cardiac catheterization and echocardiography |
Hemnes et al. [17] | PH | N = 2270 Retrospective, single-centre study over 16 yrs in patients referred for simultaneous RHC and LHC | Digitised mean PAOP during RHC and ‘manually measured’ LVDEP during LHC | Mean difference − 1.6 mmHg IQR − 15 to 12 mmHg Modest correlation by linear regression r2 = 0.36, p < 0.001 In those with PH (n = 1,331) mean difference 0.3 mmHg IQR − 14 to 14mmHG, less correlation r2 = 0.27, p < 0.001 | Any patient deemed to have ‘extreme critical illness’. Acute decompensation, shock, vital signs suggesting imminent death) or cardiac-related critical illness (hypertensive crisis) |
Halpern et al. [18] | PH | N = 11,523 Retrospective, single centre. Patients referred for simultaneous RHC and LHC data over a 10-year period | Mean LAP during RHC versus simultaneously measured LVEDP during LHC | Moderate discrimination between patients with high vs normal LVEDP AUROC = 0.84; 95% CI 0.81 to 0.86 PAOP poorly calibrated to LVEDP (Bland–Altman limits of agreement, − 15.2 to 9.5 mm Hg N = 3926 with mean PAP greater than 25 mm Hg. 14.8% with a PAOP < 15 mm Hg of which 310 (53.5%) were misclassified, having an invasive LVEDP > 15 mm Hg | Mitral stenosis or HR > 130 bpm |
Mascherbauer et al. [19] | HFpEF | N = 152 Prospective simultaneous RHC and LHC | Digitised mean PAOP over 8 cardiac cycles during RHC LVEDP ‘manually measured’ during LHC | Modest pressure difference 2.0 ± 4.4 mmHg between PAOP and LVEDP |  > Moderate valvular heart disease, congenital heart disease, significant coronary artery disease requiring PCI or CABG. Severe congenital abnormalities of the lungs, thorax, or diaphragm, COPD with a forced expiratory volume in 1 s (FEV1) < 50% |
Non-critical care studies evaluating echo Doppler and invasive LVEDP or PAOP | |||||
Lancelloti et al. [25] | Patients with and without heart failure (25% had an EF < 50%, 53% had coronary artery disease) clinically requiring coronary angiogram | N = 159 Prospective multicentre, 9 centres in Europe | Echo estimate of LVFP using 2016 recommendations (E/A, E/e′, left atrial volume index, tricuspid regurgitation jet velocity) within 30 min of LHC measured LVEDP (elevated defined as ≥ 15 mm Hg and measured as the mean LVEDP averaged over 3 consecutive cycles) | 65% of patients with normal non-invasive estimate of LVFP had normal LVEDP. 79% of those with elevated non-invasive LVFP had elevated invasive LVEDP Sensitivity 75%, specificity 74%, PPV 39%, NPV 93%, AUC 0.78 | ACS, > mild valvular heart disease, valvular prosthesis, MAC, previous MI involving basal septum and/or basal lateral wall, AF/severe arrhythmias precluding Doppler analysis, LBBB, PPM HCM, pericardial disease, inadequate echocardiographic imaging or any administration of diuretics or vasodilators within the day prior the hemodynamic evaluation |
Balaney et al. [26] | ‘Clinically indicated LHC’ | N = 90. Prospective, single centre 9 patients ‘indeterminate’, total n = 81 | Non-invasive estimate of LVFP using 2016 recommendations versus invasive LVEDP (pre-A pressure at end expiration with LHC) | Sensitivity (of the detection of elevated LVFP) 0.69, specificity 0.81, PPV 0.77, NPV 0.74, accuracy 0.75 | Hemodynamically unstable, AF, > moderate mitral regurgitation, > moderate MAC, mitral stenosis, heart transplantation, sinus tachycardia, prosthetic valves |
Nauta et al. [27] | HFpEF | Systematic review of 9 studies Comparison of E/e′ to invasively measured ‘LVFP’ | Five studies used PAOP and four studies used LVEDP as invasive reference. Invasive measurements were simultaneous or directly after echo in seven out of nine studies | Meta-analysis using a random-effects model yielded a pooled r correlation coefficient of 0.56 | 101 full test articles assessed |