Introduction
Methods
Data sources and searchers
Study selection
Quality assessment
Diagnostic performance and data extraction
Data synthesis
Results
Search results
Quality assessment
Study characteristics
Study | Study design | Predictors | Outcome | Study population | Index group (n); sex (% female); age | Reference group (n) | Reference diagnosis |
---|---|---|---|---|---|---|---|
HFpEF studies | |||||||
Multivariable models and echocardiographic equations | |||||||
Thenappan USA [23] | Cross-sectional | Age, clinical data, echo, haemodynamics | PH-HFpEF | PH registry | PH-HFpEF (100); 82%; 64 ± 13 | PAH (522) | HF symptoms + LVEF > 50% + PWCP > 15 mmHg/LVEDP > 15 mmHg/PVR > 2.5 wood units |
Weber EU [11] | Cross-sectional | E/e′ + other echo, arterial function, clinical data | HFpEF | Subjects referred to RHC for suspected CAD | HFpEF (71); 33.8%; 67.7 ± 8.6 | Non-HFpEF (65) | LVEF > 50% + LVEDP > 16 mmHg + NT-proBNP > 220 pg/mL |
Cameron USA [24] | Cross-sectional | 2009 ASE/EAE guidelines + multivariable models | PH-HFpEF | Subjects enrolled in the PH program for the assessment of PH | PH + LVED > 15 mmHg (81); 67%; 62 (56–70) | PH + LVEDP ≤ 15 mmHg (80) | PASP > 25 mmHg + LVEDP > 15 mmHg |
Dokainish USA [17] | Cross-sectional | Echocardiographic equations | HFpEF | Subjects referred to LHC for clinical reasons | LVEF > 50% + LVEDP > 20 mmHg (69); 58%; 55.1 ± 8.5 | LVEF > 50% + LVEDP < 20 mmHg (53) | LVEF > 50% + LVEDP > 20 mmHg |
Dini EU [16] | Cross-sectional | Echocardiographic equations | HFpEF | HF subjects | HFpEF (55); 35%; 67 ± 12 | HFrEF (123) | LVEF > 50% + PCWP > 15 mmHg |
Reddy USA [18] | Cross-sectional | H2FPEF score | HFpEF | Subjects undergoing RHC for the evaluation of dyspnoea | HFpEF (267); 61%, 68 ± 11 | NCD (147); 59%, 56 ± 15 | LVEF > 50%, dyspnoea + PCWP at rest ≥ 15 mmHg or during exercise ≥ 25 mmHg |
Left ventricular strain and strain rate | |||||||
Kasner EU [12] | Case-control | Global strain rates and their ratios with early transmitral flow | HFpEF | HFpEF (21); 52%; 43–60 | Subjects with chest pain (12) | τ ≥ 48 ms and/or LVEDP ≥ 16 mmHg and/or stiffness constant β ≥ 0.015 mL/1 and/or stiffness b ≥ 0.19 mmHg/mL + HF symptoms + normal LVEF | |
Wang USA [14] | Case-control | Global longitudinal strain | HFpEF | DHF (20); 35%; 63 ± 11 | Healthy subjects (17) | LVEF > 50% + PCWP > 12 mmHg | |
Left atrial strain | |||||||
Kurt USA [13] | Case-control | E/E′/LA systolic strain (LA non-invasive stiffness) | HFpEF (DHF) | DHF (20);30%; 58 ± 16 | LVH + normal LVEF (19) | Clinical criteria + PCWP (ESC 2007 guidelines) | |
Lundberg EU [20] | Cross-sectional | LA global strain (LA-GS), TR Vmax, LAVi and E/e′ | HFpEF | Subjects referred to RHF for suspected HF | EF ≥ 50% (63) + abnormal LAP | Normal LAP (29) | Pulmonary artery wedge pressure (PAWPM) > 15 mmHg at rest or ≥ 23 mmHg during peak exercise |
Reddy USA [19] | Cross-sectional | LA reservoir, conduit and booster strain, LA reservoir strain/E/e′, LA reservoir strain/LAVI | HFpEF | Subjects undergoing RHC for dyspnoea | HFpEF (238), 62%, 68 ± 10 | NCD (125), 56%, 58 ± 14 | Clinical symptoms of HF + LVEF ≥ 50% + PCWP with rest ≥ 15 mmHg and/or exercise ≥ 25 mmHg |
Singh USA [21] | Cross-sectional | Peak LA strain | HFpEF | Subjects referred to LHC for various reasons (chest pain, ACS, etc.) | HFpEF (7) | LVDP <15 mmHg (25) | Pre-A-wave LVDP > 15 mmHg |
Telles AU [22] | Cross-sectional | LA global reservoir and LA pump strain | HFpEF | Subject referred to RHC for exertional dyspnoea | HFpEF (49), 71.4%, 69.4 ± 8.0 | NCD (22), 77.3%, 67.0 ± 9.9 | LVEF > 50%, dyspnoea + PCWP ≥ 15 mmHg at rest and/or ≥ 25 mmHg at maximal exertion |
Diastolic stress test markers | |||||||
Hammoudi EU [25] | Cross-sectional | Lateral and septal E/E′ at low-level exercise (25 and 50 W) | Early HFpEF | Subjects at high risk for HFpEF | LVEDP > 16 mmHg during exercise (34);23%, 64.8(55.2–73.4) | LVEDP < 16 mmHh (12) | LVEDP >16 mmHg |
Obokata USA [5] | Cross-sectional | ESC algorithm + exercise average E/E′ | HFpEF | Subjects referred to RHC for exertional dyspnoea | HFpEF (50); 54%; 70 ± 11 | NCD (24) | HF symptoms, LVEF ≥ 50%, PCWP at rest > 15 mmHg and/or with exercise ≥ 25 mmHg |
Single conventional echocardiography markers | |||||||
Nagueh USA [15] | Cross-sectional | Echo estimated RAP > 8 mmHg | HFpEF | Subjects with exertional dyspnoea enrolled in a multicentre study | HFpEF (50); 44%; 64 ± 9 | Non-HFpEF (79) | LVEF > 50% + PCWP > 12 mmHg |
Left ventricular diastolic dysfunction studies | |||||||
Goto Jap [27] | Cross-sectional | BNP > 22.4 pg/mL + E velocity < 7.4 cm/s | LVDD | Subjects referred to LHC for the evaluation of CAD | Isolated LVDD (91); 18.7%; 67.4 ± 8.2 | Normal diastolic function (189) | LVEF ≥ 50% + τ ≥ 48 ms |
Weber EU [29] | Case-control | LVETI, E/A, E′ and E/E′ | LVDD | Subjects with suspected CAD | LVDD (44), 50%, 65.7 (10.1) | Healthy controls (82), 28.1%, 55.6 (8.9) | LVEDP > 16 mmHg + LVEDVI < 102 mL/m2 + LVEF > 50% |
Bruch EU [26] | Cross-sectional | Tei index | IDD | Subjects referred to LHC or known/suspected CAD | HFpEF (29); 24%; 63 ± 9 | Normal echo (11) | LVEDP > 16 mmHg + LVEF > 45% |
Hayashi Jap [28] | Cross-sectional | Ratios of E wave to peak longitudinal strain (E/LS), E/A and E/E′ | LVDD | Subjects who underwent LHC for clinical diagnosis of cardiac diseases | LVEF > 50% (47), of whom 38 with τ ≥ 48 ms and 18 with LVMDP ≥ 12 mmHg | HFrEF (30) | Abnormal LV relaxation = τ ≥ 48 ms; LVMDP ≥ 12 mmHg |
Measures of diagnostic performance: HFpEF
Multivariable models
Study | Markers | Sensitivity | Specificity | AUC (95% CI) (+ p value) | PPV and PNV | Accuracy | LR + and LR- | NRI and IDI |
---|---|---|---|---|---|---|---|---|
HFpEF studies | ||||||||
Multivariable models and echocardiographic equations | ||||||||
Thenappan USA [23] | Age + WHO functional class, hypertension, obesity, DM, CAD, serum creatinine, diuretic, β-blocker, ACE inhibitors/ARBs + LVPWT, LA and RA enlargement | 0.935; (0.90–0.97) | ||||||
Weber EU [11] | E/e′ + aortic PP + age + ACE-I/ARB + β-blocker + NO-donator | 90.09% | 0.952 (0.894–0.983 (p = 0.0002) | Echo + aortic PP: 32.9% | ||||
Cameron USA [24] | E/A, E/e′, LA diameter (1.5 x LA diameter) + (1.7 x E/A) + (1.1 x E/e′ septal) | 68% | 63% | 0.7 (0.62–0.68) | 63% and 65% | 1.7 and 0.5 | ||
Dokainish USA [17] | 1) PASP + LAVI)/2 > 30 2) (E + LAVI)/2 > 57 | 1) 72% 2) 73% | 1) 80% 2) 81% | 1) 0.84 (p < 0.001) 2) 0.82 (p < 0.001) | ||||
Dini EU [16] | CART model (EDT < 150 ms + AR dur-A dur > 30 ms + E/e′ > 13 + LAVI > 40 mL/m2 + E/Vp > 2) | 87% | 90% | 92% and 84% | 88% | |||
Reddy USA [18] | H2FPEF score: obesity + AF + age > 60 years, treatment with ≥ 2 antihypertensive drugs + E/e′ > 9 + and PASP > 35 mmHg | 76% | 78% | 0.841 (0.798–0.876), p < 0.0001 | 3.49–0.31 | |||
Left ventricular strain and strain rate | ||||||||
Kasner EU [12] | SRE, SRIVR, E′/A′, E/SRE, E/SRIVR and E/e′ lat | SRE = 0.55, SRIVR = 0.70, e′/A′ = 0.72, E/SRE = 0.75, E/SRIVR = 0.80, E/e′lat = 0.83 | ||||||
Wang USA [14] | GLSl < − 16% | 95% | 95% | 0.98 | ||||
Left atrial strain | ||||||||
Kurt USA [13] | LA non-invasive stiffness index > 0.99 mmHg | 85% | 78% | 0.85 (0.72–0.98) | ||||
Lundberg, EU [20] | 1) Rest LA GS (LA-GS, − 21%) 2) Stress LA GS | 1) 93% 2) LA-GS 92% | 1) LA-GS 77% 2) LA-GS 88% | 1) 0.87 (p < 0.001) 2) 0.93 (p < 0.001) | ||||
Reddy, 2019 USA [19] | 1) LA reservoir strain (< − 24.5%) 2) LA conduit strain (< − 18.4%) 3) LA reservoir strain/E/e′ (< 3) 4) LA reservoir strain/LAVI | 1) 56% 2) 64% 3) 65% 4) 58% | 1) 94% 2) 63% 3) 78% 4) 85% | 1) 0.719 (0.664–0.767), p < 0.0001 2) − 0.071 (− 0.102 to − 0.040) (vs reservoir strain), p < 0.0001 3) + 0.053 (+ 0.019 to 0.088) vs reservoir strain, p = 0.003 4) + 0.032 (+ 0.016 to 0.001) vs reservoir strain, p = 0.04 | ||||
Singh USA [21] | Peak LA strain (< − 20 mmHg) | 71% | 92% | 83% and 92% | 91% | |||
Telles Au [22] | 1) LA global reservoir (< − 32.2%) 2) LA pump strain (< − 15.5%) (AF subjects excluded) | 1) 90% 2) 94% | 1) 74% 2) 80% | 1) 0.85 (0.76–0.95), p < 0.001 2) 0.88 (0.77–0.98) p < 0.001 | 1) NRI 12% 2) NRI 14% (vs ESC) | |||
Diastolic stress test markers | ||||||||
Hammoudi Eu [25] | Ex septal E/é at 25 W ≥ 8 | 71% | 83% | 0.79 (0.67–0.92) (p < 0.0001) | ||||
Obokata USA [5] | 1) ESC + Ex E/e′ > 14 2) ESC + 20 W Ex E/E′ > 14 | 1) 90% 2) 80% | 1) 71% 2) 88% | 1) 0.80 (0.68–0.89) (p < 0.05 vs ESC) 2) 0.84 (0.73–0.91) (p < 0.05 vs ESC) | 1) 87% and 77% 2) 93% and 68% | 1) 3.1 and 0.1 2) 6.7 and 0.2 | ||
Single conventional echocardiography markers | ||||||||
Nagueh USA [15] | RAP > 8 mmHg | 76% | 89% | 80% and 87% | 85% | |||
Diastolic dysfunction studies | ||||||||
Goto Jap [27] | BNP > 22.4 pg/mL + E velocity < 7.4 cm/s | 44% | 86.8% | 61.5% and 76.3% | ||||
Weber EU [29] | LVETI (427.1 ms) | 70% | 82% | 0.81 (0.72–0.89), p < 0.0001 | 76% | |||
Bruch EU [26] | Tei index > 0.49 | 37% | 86% | 0.61 ± 0.08 | ||||
Hayashi Jap [28] | E wave/peak longitudinal strain (E/LS) > 680 cm/s | 72% | 88% | 0.80 |