I have a 50 year old patient who had an inferior wall MI 2 and a half years back and is on Aspirin, Clopidogrel and Statins. He does not have hypertension or diabetes. A year after the MI, his cardiologist reported his LVEF to be 55-60% after an echo. A year and a half later, a repeat echo was performed (for routine purposes), and the cardiologist reported 50-55% LVEF and Grade 1 diastolic dysfunction. The patient is currently asymptomatic and is here for regular follow up. However he is a smoker but smokes only 1-2 cigarettes per day - he is on a plan to quit smoking gradually. I was wondering whether the decrease in LVEF is worrisome? Is LVEF a reliable measurement? Is it true that it undergoes day to day variations? Any help would be appreciated! Thanks!
Using echo is an easy tool to evaluate myocardial performance. One parameter you can measure is LVEF. Ejection fraction (EF) represents the volumetric fraction of blood pumped out of the ventricle every heart circle; it is stroke volume (SV) divided by end-diastolic volume (EDV). Several methods can be used to calculate EF. M-Mode measurement by Teichholz, volumetric measurement by Simpson-algorithm or subjective" eye-ball" method. If you use all these methods simultaneously you can have EF variety of up to 10-15%. Another point is inter-observer variety. Last but not least SV and EF is dependend on preload (remember Frank-Starlin curve), afterload and myocardial contractility which can be influenced by hormones,drugs or stress. In conclusion a slight difference(5-10%) you descriped in your question wouldn't bother me as long the difference is not generated by new regional wall motion disorder.
A change in LVEF of 5-10% is likely to represent a genuine fall in LVEF. Anything 10% is highly likely to represent a real decline. If you are uncertain, you could look for corroborating evidence of LV deterioration, such as increase in LV end-systolic volume index or LA volume index, decline in S' (peak systolic mitral annular tissue velocity) and decline in global longitudinal LV strain by speckle-tracking. Another important consideration, beyond the documentation of deteriorating LV function, is: how would such a finding change your management of the patient? I expect their pharmacotherapy and lifestyle interventions have been optimised. Would progressive coronary disease and silent ischaemia need exclusion? Or other causes of LV dysfunction, such as alcohol excess? They certainly do not qualify for primary prophylaxis of sudden death with an implantable defibrillator.
The caveat to the test-retest reproducibility figures I quoted in my previous answer is that it depends on the expertise of the measurer. High-end labs should know their coefficient of variation for LVEF, etc
On a general basis, an isolated 5% change in left ventricular EF would not worry me, especially when not accompanied by other changes in echo findings (for example, an elevation in LV filling pressure/worsened diastolic function, larger LA, higher sPAP,segmental abnormalities) and/or by symptoms. I feel that 5% is the average intraobserver variability for echocardiographers, some may have higher IOV, some others may have lower ones; it also depends critically on the patient's acoustic window, and by several confounding physiological and pharmacological variables. However, these issues become particularly compelling when looking at pts with LV dysfucntion who may or not be candidates for device therapies; in the case of chronic CAD with EF in the field of the 50s and no changes in symptoms or other clinical/instrumental data, I would feel it adequate just to give the patient a closer follow-up. In case of doubt, if available, BNP or NT-proBNP levels may aid.
The gold standard for EF is cardiac MRI , maybe it's more fair to say the most accurate instead of gold standard
Another important point is the type of echocardiograph that you use. If you use one that has real-time TD and automatic tracing of endocardial borders, you will probably retrieve high quality and highly reproducible information...
I would not consider a change of 5% on an Echo as significant. Echo EF's are usually estimated qualitatively and are least reliable when segmental wall motion abnormalities are present. As noted above the accuracy is very dependent on the experience of the techjnologist or cardiologist. While MR measurements are considered the most reliable the cost is probably not realistic for use in following most patients. Nuclear gated blood pool imaging is probably more reliable than echo with a smaller variation. It is considered the method of choice to follow chemotherapy patients for early cardiotoxicity by most oncologists. It has the advantage of not relying on geometric based models.
It is important to compare both exams. In addition the EF value depends on measurement echo method. Probably it worth to make a stress test or to assess EF by another method (for example MRI).
I think that litle change in EF Echo performed in different date not denote worse or improvement in ventricular function. I agree wuith Drs. Voigt and Janowitz, but I think unnecessary to perform other diagnosis method. Only in patients with very decreased EF a litle change in EF Echo is reason to performe other method for mesurement the EF to confirm.
With regard to ejection fraction there is a recognised intra-operator variability and also interoperator variability and this is quoted as being between 5 and 10%.
The search for the most accurate assessment of EF is ongoing however most people would agree that, when image quality allow, Simpson Biplane is most accurate.
The accuracy of the test is dependent on the experience of the operator and interpretor and on the technical quality of the study.
I would not agree with Dr Cocco's implication that where technicians perform scans and present the data to the physician that there is less guarantee of the accuracy.
What is important is the level of experience of the operator and their level of expertise. Accreditation from bodies such as the British Society of Echocardiograpy or the European Association of Echocardiography gives competence in the competency of the operator.
A drop of 5% should not be considered significant presuming that it does not cause the overall assessment of function to be go from one class ie normal to another ie mildly impaired. Therefore 55-60% going to 50-55% would not be considered significant, however 50-55% (normal) going to 45-50% (mildly impaired) would be.
To detremine contractility of LV precisely I look fro EF by 4 chamber view,2 chamber view,also I use FS and MAPSE.
The variability of echo visually-derived EF is greater than the variation you have observed in your patient making this change irrelevant. While I strongly support CMR as the 'gold-standard' it is worth mentioning that directional changes in Echo EF over time is of value but the time it takes to detect such a relevant finding isn't too clinically valuable except in the asymptomatic patient. Theoretically EF, if it truly measured innate contractile function (and could be measured accurately), should not undergo 'day-to-day' changes. However, in that it is load dependent it will reflect changes in preload and afterload and can undergo hour-to hour changes (albeit not clinically relevant).
Echocardiography is an imaging modality. Performance of Echocardiography is a technical skill. Interpretation of the Echocardiogram is a professional skill that is ideally done by the clinician who is also seeing the patient. The clinician who is trained in cardiology (ie. cardiologist) is in my opinion the best one to do that. Anyone else (cardiac technician, general physician, radiologist) can also be trained to interpret echocardiograms, but they would lack the clinical experience to place the imaging findings in context. I am not referring to clear-cut guidelines which everyone can read and agree with, but to individual patients, who often do not fit in the cookie-cutter moulds that we would like to pigeon-hole them in. As a physician, I have to ultimately recommend a therapeutic plan, and then continue to see the patient who expects to get better. Quoting guidelines to patients is cold comfort if they do not improve. In this sense, I agree with Guiseppe Cocco that Echocardiography and Echocardiology are two distinct entities.
I would agree with the comments above but I think (and I would as a technician) that you can have excellent technicians who can both take excellent quality scans but also interpret the results well and present the findings in a very clear way which answers the clinical question. One problem I have in daily practice is that the question or clinical background given on a request for can be poor which limits my ability to answer that particular question.
I think that it is very important to correlate the clinical scenrio with the echo result and I would always say that if the result does not match the clinical scenario to come and look at the images and talk to the performing sonographer.
With regards to EF, one of the limitations is that all traditional measures of EF (M-mode, Simpsons) assess mostly radial function. It is not uniformally accepted to assess longitudinal function, which can be the first sign of systolic function. We used to measure MAPSE (mitral annular planar systolic excursion) but how many centres stil do it and how do you combine it with traditional measures of EF
Timing of aortic valve opening and closing is critical (hopefully non-regurgitating), while mItral valve prolapse could greatly mess up the answer. Echo is highly inaccurate, especially for whole ventricular function, but then every mode has its drawback as recent reviews show. If the patient is arrhythmic then you are really in trouble. The dynamics of forces in the wrenching heart are very complex so that wall abnormality is only a gross index of malfunction. Don't forget the heart's neuro-muscular adaptation to regional malfunction that may make the reading of the summed ejection fraction quite weird. So, in the end, correlation of different perspectives is your best indicant of ventricular function as ejection fraction is as very simplified notion of what the heart does to spit blood out of the left ventricle. I would ask is Doppler studies have ever been studied as indicative of ventricular forces? Has your patient exhibited ant left atrial and aortic root abnormalities?
Stress-anything would give more useful results that the patient lying around inactive. I believe that you can't be sure what that means unless you stress the patient before and after activity training. All best are off if your patient suffers from diabetes for some years or microvasculitis or any inflammatory disorder. By the way, I'd like to do a whole blood viscosity study as you would be surprised how often that's a factor!
Lastly, radiologists and cardiologists don't always see the same thing, as can be acutely seen in neurosurgery. Time is your best friend in this case, worry your worst enemy, especially if you are the patient!
Supervised exercise may changer the values over time as animal data showed. The advantage of echo is that you can do it and do it often so you can get a sense of how the whole thing sums up. Let's face it, cardiac Dx is a crude thing so the more data you get the more a sudden change means something.
....permit me to add that the diastolic dysfunction is a bigger muddle than thwe systolic. As Dr. Yao seems to suggest, cardiac imaging is new so even though we "see" things, we're still swimming in pea soup. I still believe in high Tesla MRI BOLD and other type oxygen studies. Therein lies the future that may be there now. Smoker is sort of like a diabetic microvascularly speaking. If you stop him from smoking and get him exercising the muscle mass south of his umbilicus, not north, you will have done much for his survival in my view.
to simply answer your question: I would assume that the differences in EF are neglibigle. Even if you have the same rater and excellent exercise conditions it is almost impossible (and useless) to determine if EF is 50-55% or 55-60%. Though by definition EF > 55% is mildly decreased. Be Aware that diastolic function also seems to be issue and that stroke volume also depends on the diameter of the left ventricle.
An echocardiographic-derived EF change of 5% via one way or another is within the standard deviation of the reader and as such does not represent an absolute change nor a clinical endpoint. In the best studies, interobserver and intraobserver studies have variabilites b/w 8-13% and that is with contrast and harmonics. Obviously, the variability is worse without those additional tools. Now, over time, directional changes do have value such that upon repeat imaging, ideally intepreted by the same reader, continue to show a depressed or worsening EF, you have a higher reliability that this initial observation is valid. I would recommend that, if such a confidence is required, cardiac MRI generally has a variability with inter- and intraobserver reliability b/w 1.5 and 3% in good centers with full 3D volumetrics analysis (generally takes 10-15 minutes maximum of scanning and analysis time total).
I have dilated cardiomyopathy with biventriclaar hypertrophy and right bundle branch block and my ejection fraction on echo was
ECHOCARDIOGRAPHY
User Dependant tool with INTER OBSERVER Variability.
EF by ECHO
Simpson's Volume Method more ACCURATE than Teich & other MMODE methods.
cONSIDERING your patient 50-55% Vz 55-60% is EQUIVOCAL given the 55% Intersection. The 5% difference probably 2° INTER OPERATER\OBSERVER.
OtherCauses of ↓ning EF
Worseming Infarction
aGE
I wonder how much more reliable is stressed nucleotide study and how has global longitudinal stress speckles study improved long term predictive power. The latter, I believe, is a matter of software.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Infectious Diseases