E/e′ ratio
E/e′ · Mitral E (PW at leaflet tips) divided by tissue Doppler e′ from the same study; report the average
Normal range & thresholds
| Site | Normal | Abnormal (elevated LAP) |
|---|---|---|
| Average | <8 | >14 |
| Septal | <10 | ≥15 |
| Lateral | <8 | ≥13 |
The British Society of Echocardiography publishes an alternative algorithm; where two schemes disagree, say which one was applied.[99]
The 2025 two-step algorithm: Step 1 assess the three primary variables simultaneously — e′, E/e′, peak TR velocity. Concordant results classify LAP directly. Step 2, for discordant or incomplete cases, adjudicate with any one of: LA reservoir strain ≤18%, pulmonary vein S/D ≤0.67, LAVI >34 mL/m². The ‘indeterminate’ category has been removed.[2]
Pathophysiology
E is driven by the atrioventricular pressure gradient, which depends both on atrial pressure and on the ventricle’s ability to relax and suck. e′ isolates the relaxation half of that. Dividing one by the other therefore corrects E for the relaxation that produced part of it, leaving a residual that tracks atrial pressure. The logic only holds when relaxation is genuinely impaired and the annulus is free to move — which is exactly why the ratio fails in young normal hearts, in constrictive pericarditis, in significant primary mitral regurgitation with a normal ventricle, and wherever the annulus is calcified, ringed or prosthetic.[2]
Raised by
- Elevated left atrial pressure from any cause — HFpEF, HFrEF, hypertensive heart disease, ischaemia, restrictive and infiltrative disease, HCM
- Volume overload and renal failure
- Mitral stenosis (E is stenotic, not filling-driven — the ratio is not interpretable)
- Falsely raised: mitral annular calcification, mitral annuloplasty ring or prosthesis, prior mitral surgery, and regional infarction at the sampled annulus — all lower e′ mechanically
Lowered by
- Normal filling pressures
- Young age and athletic conditioning (high e′)
- Constrictive pericarditis — the preserved medial e′ produces a low E/e′ despite grossly elevated filling pressures: the classic false negative[57]
- Primary mitral regurgitation with a normal ejection fraction — the guideline states plainly that E/e′ is not useful; use IVRT, Ar−A duration and IVRT/TE-e′ instead[2]
- Over-diuresis
Technique & pitfalls
- Measure E and e′ on the same study, at the same heart rate; the ratio of numbers from different visits is meaningless.
- Use the average of septal and lateral where both are valid; use lateral alone in LBBB, RV pacing, post-cardiac surgery and pulmonary hypertension; use septal alone with caution in lateral wall infarction.[2]
- In pulmonary hypertension the guideline specifies the lateral ratio, with 8–13 recognised as weakly associated with LAP.[2]
- Do not compute E/e′ at all in mitral stenosis, in a mechanical mitral prosthesis, or in significant mitral annular calcification — report why instead.
- State in the report which sites were used and which algorithm was applied.
Pseudo-change & artefact
- Anything that tethers the annulus falsely raises the ratio — annular calcification, rings, prostheses, prior surgery. This is the commonest false-positive diastolic report in clinical practice.
- Constriction and primary MR with normal EF falsely lower it.
- Diuresis: the ratio falls within hours — a ‘normal’ E/e′ after treatment does not mean the diastolic disease has gone.
- Tachycardia with E–A fusion; ectopy; atrial fibrillation cycle-length variation.
- Angle error in e′; sample volume off the annulus.
- Using a single site in a ventricle with regional disease.
Treatment thresholds
- HFpEF diagnosis: E/e′ ≥15 is a major functional criterion in HFA-PEFF, and the 2025 ASE guideline devotes a dedicated section to HFpEF diagnosis; a confirmed diagnosis brings SGLT2 inhibitors (class I in the 2023 ESC focused update) into play.[2,12,82,101]
- Diastolic stress echocardiography: exercise average E/e′ ≥15 with peak TR velocity >2.8–3.4 m/s confirms exercise-induced elevation of filling pressure and converts unexplained dyspnoea into a diagnosis of HFpEF.[105,106]
- Decongestion: a persistently elevated E/e′ at discharge predicts readmission and is used as an objective target for diuretic titration — the closest thing to a therapeutic threshold this parameter has.
- Hypertrophic cardiomyopathy: average E/e′ >14, with Ar−A ≥30 ms, TR velocity >2.8 m/s and LAVI >34 mL/m², indicates elevated filling pressures and supports symptom-directed therapy.[2,13]
- Heart transplant: average E/e′ <7 normal, 7–14 needs E/SRIVR, >14 elevated — a transplant-specific algorithm.[2]
Next step
- Next: apply the full two-step algorithm; do not report a filling pressure on E/e′ alone.
- Then: if the primaries are discordant, measure LA reservoir strain, pulmonary vein S/D and LAVI.[2]
- Then: if resting values are normal or grey-zone and the patient is breathless, do a diastolic stress echo; add natriuretic peptides and the H2FPEF score.[102,105,106]
- Then: when management genuinely hinges on the answer — unexplained dyspnoea, discordant non-invasive results, pre-transplant or pre-device evaluation — proceed to invasive haemodynamics with exercise. It remains the reference standard, and the guideline says so.[2]
Drugs
- Loop diuretics — the fastest and largest reduction; hours to days.
- SGLT2 inhibitors reduce E/e′ and improve symptoms in HFpEF and HFrEF.[82]
- ARNI lowers filling pressures and E/e′ alongside reverse remodelling.[68]
- MRAs, nitrates and other preload reducers lower it acutely.
- Beta-blockers and rate control lengthen diastole, lower E and improve the ratio in tachycardia and in HCM; mavacamten improves it in obstructive HCM.[72]
- Raised by: intravenous fluids, transfusion, NSAIDs, high-dose steroids, and any agent causing sodium retention.
Reversibility
The most rapidly reversible parameter in this document. E/e′ falls within hours of intravenous diuresis and can normalise entirely during a single heart failure admission, because the numerator is a pressure and pressures move fast. That is a trap as much as a therapeutic marker: the ventricle is unchanged. The denominator (e′) barely moves in the same period, so a normalised E/e′ with a still-low e′ is decongestion, not recovery — report both, always.[2]