Conversations in Cardiology: The End of End-hole Left Ventriculography – A Consensus of Operators

Left ventriculography, once an integral part of every cardiac angiographic procedure, has been reconsidered as often unnecessary in light of high quality echocardiography for left ventricular (LV) functional assessment. In the 1980s, end-hole catheter ventriculography with a multipurpose or Sones type catheter was in decline and ultimately replaced by the pigtail multi-side hole catheter, because of the increased safety with less ventricular ectopy and near absence of LV perforation. However, the practice of end-hole ventriculography, often performed through a right Judkins catheter with a hand injection for operator convenience, has come under scrutiny as an inappropriate cath lab technique that some strongly advocate should be abandoned.

eft ventriculography, once an integral part of every cardiac angiographic procedure, has been reconsidered as often unnecessary in light of high quality echocardiography for left ventricular (LV) functional assessment. In the 1980s, end-hole catheter ventriculography with a multipurpose or Sones type catheter was in decline and ultimately replaced by the pigtail multi-side hole catheter, because of the increased safety with less ventricular ectopy and near absence of LV perforation. However, the practice of end-hole ventriculography, often performed through a right Judkins catheter with a hand injection for operator convenience, has come under scrutiny as an inappropriate cath lab technique that some strongly advocate should be abandoned.

In this ‘Conversation in Cardiology,’ our cath lab experts express their opinions and experience with left ventriculography and, for one of the first times in these pages, seem to be in agreement on the fact that LV angiography through an end-hole catheter is a practice that should be discontinued, for a variety of reasons. As with all of our “conversations,” the opinions expressed are those solely of the contributor. I thank them for making this exercise enjoyable and informative to help advance better practices for our patient outcomes in interventional cardiology.

Here is the question Dr. Steve Bailey of San Antonio, Texas, asked us: 

In doing some “quality reviews” internally and externally, I am seeing LV grams performed with hand injections through end-hole catheters that often poorly opacify the LV and are marred by ectopy, but are coded and billed. The EF [ejection fraction] for these same studies is also typically visually estimated rather than quantified using computer programs. Interestingly, in one community lab here in San Antonio, the cath techs did not even know how to do a quantitative LV EF assessment, as “we always just ask the doctor what it is.” As there are increasing questions regarding what constitutes “quality” for diagnostic studies, I wonder what the current standard is among our colleagues in their catheterization laboratory?

Question 1: Is it acceptable to do a hand injection to determine LV function? If so, when? What criteria are applied? 

Question 2: Is visual estimation acceptable or is quantitative assessment the standard of care?

Mort Kern, Long Beach, Calif.: Pertinent and practical questions. Here are my answers:

1. Never hand injection. If you can’t do a good job, get an echo. You shouldn’t bill for a “non-LV gram.”
2. Visual estimate is acceptable. We have not done quantitative assessment except for research study. Grading by 10% variance is probably common, i.e. <20, 20-30, 30-40, 40-50% EF, etc.  As a side comment, LVEDP [LV end diastolic pressure] for STEMI [ST-elevation myocardial infarction] and CHF [congestive heart failure] patients is still important. LVEDP pre contrast is better than post and the change in LVEDP (pre vs. post) has little importance clinically. I look forward to our colleagues’ comments.

Ajay Kirtane, New York City, New York: We don’t do hand injections for LV grams, and have outlawed end-hole injections for safety reasons. We do use visual estimation though (also, as I write this, my copy of Baim/Grossman is threatening to fall off the shelf and hit me on the head!).

We also routinely do LVEDP and LV grams in STEMI patients, PRIOR to PCI (to r/o mechanical complications, diagnose infarct territory, etc.). I have had many an argument (with some on this email list as well!) about that one, but the D2B [door-to-balloon] time lost by an experienced operator is minimal and every year or so we find a mechanical complication that would otherwise not have been diagnosed. We also titrate hydration to pre-LVEDP based upon the POSEIDON trial protocol.

Bonnie Weiner, Worchester, Mass.: We see this [practice] a lot doing external reviews for ACE [Accreditation for Cardiovascular Excellence] and call it out all the time as a quality gap. My answers are:

1. Never acceptable (yes, I know I should never say never, but really, if you don’t want to seriously do an LV gram, do an echo).
2. Visual estimation is the norm, but I agree it is severely flawed. There is no reason that staff cannot be trained to do this, would not add to physician time, but would certainly add to the quality of the assessment. Digital capability may have made the planimeter obsolete, but the concept is not.

Peter Ver Lee, Maine Medical Center, Portland, Maine: We never do it (hand injection), but I see the outside labs do it, with a left Judkins catheter! I have the techs do the LV gram, but often change it, as they seem to get a number that is too high. If the patient has had an echo, I don’t see the point in doing a ventriculogram. I might put a side hole or pigtail catheter in the LV for EDP, but not for pictures. But I think a 3D echo is much more accurate for wall motion and valve function than the best LV gram. 

Another thing that drives me crazy is outside docs who want to do an echo and a MIBI before transferring a patient to “the medical center,” obviously purely for economic reasons. Then there is no report and they are unreachable for their impressions of their tests. I end up repeating the echo and/or MIBI. Or doing a pressure wire and ignoring the results of the MIBI. Some general surgeons are even doing pericardiocentesis when the CT [computed tomography] scan “shows cardiac tamponade.” They transfer the patient to us with the drain in place. Only a matter of time before that one [hurts] them. It’s sad what financial pressures have done to physician ethics.

Fred Welt, Boston, Mass.: I agree with the sentiments expressed. This practice strikes me as transparently revenue-guided, as the information gleaned is almost never sufficient to add meaningfully to patient care.

Alan Jeremias, New York City, New York: Agree with Ajay [Kirtane] – end-hole catheters used only for LVEDP assessment (and subsequent titration of hydration), but never for LV grams. Having done both LV planimetry and visual EF assessment, I don’t think that there is a substantial quality difference between both methods (this is not QCA [quantitative coronary angiography]!).

Gregg Stone, New York City, New York: I couldn’t agree more. While some operators perform hand injections to reduce contrast use (e.g. in patients with chronic kidney disease at risk of contrast nephropathy), they are nearly always suboptimal and indeed, may be misleading, as too little contrast may not identify hypokinetic LV segments. PVCs [premature ventricular contractions] are frequent and also confound interpretation. And, at times, this technique can be seriously harmful, as hand injections are typically done through an end-hole catheter. I reviewed one case where an operator perforated the LV with a multipurpose catheter without realizing it and injected contrast directly into the pericardium. Hand injections (or any LV injection with a non-pigtail catheter, except under rare circumstances) should be considered class III in the guidelines.

The issue of LV quantification is less settled. While quantification is always preferred, a trained physician should be able to qualitatively assess LV global and regional function, at least in general categories (e.g. hyperdynamic, WNL [within normal limits], mild-moderately reduced, or severely reduced). I don’t believe that LVEF quantification is widespread as standard of care. However, operators should avoid visually reporting an exact LVEF, as their assessment will be often be quite inaccurate compared to a quantified value. In this regard, the methodology should be similar to what is done for a visually assessed diameter stenosis of a coronary lesion, for which general categories of narrowing rather than an exact diameter stenosis are reported.

Mort Kern, Long Beach, Calif.: Gregg, to your point, see the figures of an LV gram performed through a 5F Jacky catheter from the right radial approach. Fortunately, no harm was done as the myocardial staining did not penetrate into the pericardium, but this certainly could have been a catastrophe (Figures 1-2).

Lloyd W. Klein, Chicago, Ill.: Is it acceptable to do a hand injection to determine LV function? Never ever. Never use end-hole injections. When the Sones catheter was the standard, in the 1970s, it was all right. Not today. It should have gone out of favor with discos and platform shoes. Anyone who actually charges for this procedure seriously needs to attend an update.

Is visual estimation acceptable or is quantitative assessment the standard of care? The problem is that the single plane RAO [right anterior oblique] LV gram is not the most accurate quantitative method in the world. The biplane Dodge Sandler area length method and Simpson’s Rule are great, but no clinical lab is doing these. In many ways, I think an experienced angiographer can estimate an LVEF as well as an echo guy looking at 4 echo views and probably better. So, yes, it (visual estimation) is the standard of care.

Chris White, New Orleans, Louisiana: I would jump on the bandwagon in my crusade to stamp out “hand injections” for suboptimal LV grams. Either image the LV properly with a power injection, or rely on an echo. Visual estimation continues to be the standard of care, as it is in the echo lab.

Barry Uretsky, Little Rock, Arkansas: I also agree with previous comments, but need to point out that some “end-hole” catheters also have side holes such as a multipurpose (originally developed for combined coronary and LV angiography, thus the name). These catheters are not verboten, but do emphasize the technical requirement of assuring that the catheter tip moves freely in the ventricle.

Bonnie Weiner, Worchester, Mass.: You are correct, Barry, but I think what we are really talking about here is the right Judkins in the LV and so we will do a miserable LV gram.

Kirk Garrett, New York City, New York: It looks like it’s unanimous. Another thing to consider is that Medicare stopped paying for “multiple looks” at LV function in stable patients some years ago. Most of our elective patients come with stress and/or echo info on LV function, and ACS [acute coronary syndrome] patients can’t avoid an echo even if they try. I get LVEDP in almost everybody, but I am slow to shoot LV grams unless there is no record of recent (within last 6 months) LV functional assessment or recent instability.  

By the way, reimbursement for coronaries with left heart cath (i.e. measuring LVEDP) is about the same as coronaries with left ventriculography.  

Joe Babb, East Carolina: I am presuming end-hole catheter means end-hole-only catheters. They should never be used for ventriculography. I agree with Bonnie that ‘never’ and ‘always’ are treacherous words, but ‘never’ seems right for this circumstance. A hand injection is a joke. I was taught that a quality LV gram mandates that I be able to draw end diastolic and end systolic silhouettes with certainty and not use guesswork. This requires power injection of an adequate volume of contrast. Volume restriction is another quality issue I have seen more commonly of late. Therefore, NO to end-hole only catheters, NO to hand injection. As to “drawing” versus eyeballing, I think people who have trained themselves by drawing can eyeball pretty well, but those who have only “eyeballed” are likely to produce some questionable LVEFs.

George Vetrovec, Richmond, Virginia: I agree that a single-hole LV gram is not useful. Visual assessment of LV function is not bad and often better than an echo estimate. Also, I still believe LV grams often provide better data than nuclear studies, which only show wall motion and nothing about valves, while echoes often do not well visualize certain important features such as apical obliteration and their usefulness is often hampered by “a poor window”. 

So, I believe for good patient care in the absence of renal insufficiency, a properly performed LV gram using a pigtail or other non-end hole catheter and injecting 20 ml or less of contrast for one second often gives important clinical information beyond a simple LVEF and is worthwhile, regardless of whether or not it is reimbursed. 

I am always disturbed when bureaucrats oversimplify, particularly in the interest in saving a buck rather allowing clinicians to make appropriate decisions.

Gus Pichard, Washington, D.C.: I agree with you all. I never use end-hole catheters for LV gram. I never perform hand injection for LV gram. I occasionally use low volume of contrast, but it is a power injection.

The Quality Challenge and Society for Cardiovascular Angiography and Interventions (SCAI) 

John Hirshfeld, Philadelphia, Penn.: I think that this issue is the tip of a larger iceberg that goes to the heart of what SCAI is about — namely, quality degradation through corner cutting. The opportunities to cut corners in diagnostic catheterization are legion and examples of poor quality studies are everywhere. I continue to feel that a diagnostic catheterization is a major clinical event in a patient’s life and a patient is entitled to a quality job that is completed expeditiously, safely and comfortably.

I think that SCAI has a responsibility to hold its members to high quality standards. This includes identifying corner cutting and not condoning it. The responsibility goes farther to those of us who participate in training programs, as we are responsible to set the standards that we expect our trainees to adhere to. 

Carl Tommaso, Chicago, Ill.: I have done many LV grams with a Sones or multipurpose catheter (end and side holes), but always with a power injector and never had a complication. But I have seen LV grams done with the Sones that have perforated the LV if catheter was not properly positioned. A hand injection of 10cc contrast does not give adequate opacification. If an LV gram is to be done, it should be done right. 

David Rizik, Scottsdale, Arizona: My sentiments on these questions:

1. If you’re going to do an LV gram, perform a quality and safe study, not a “hand” injection though an end-hole catheter. As others have suggested, it has the aroma of being a financially motivated practice.
2. For AMI [acute myocardial infarction] patients, we keep an echo in the ER. If there is concern regarding mechanical complications of MI, I would choose the info and “safety” from an echo over an end-hole catheter. Does it slow D2B? Not even for a second. There’s always enough time to do a quick-look echo while cath lab is setting up.
3. For elective cases, as cath lab director, I have discouraged the practice of routine LV grams, especially in this era of routine echocardiography. Nearly everyone has had what seems of a surfeit of non-invasive studies by the time they get to the lab. That said, we have a few who insist on hand injected LV grams through an end-hole catheter, irrespective of the clinical circumstances (i.e. cardiogenic shock at 2 in the morning, where every second counts). Of what diagnostic benefit is a hand injection in a patient admitted for mitral valve repair who has had multiple echoes in the previous year?

With recent emphasis on appropriately using our diagnostic and therapeutic technologies safely and effectively, I support a moratorium on the practice.

Charles Chambers, Hershey, Penn.: I will share a story with you from a founder of SCAI and the Lab Survey committee, which reflects, I believe, the drift that has occurred to current practice, but the “heart” of the societies’ concerns for quality imaging.

As part of the SCAI Lab Survey Site Review, the two reviewers look at cines randomly from the lab.  I have done several of these with Bill Sheldon, SCAI trustee for life, who has been there from the beginning with SCAI. Regretfully, but almost consistently, there is an operator who, as routine practice, “slips” the JR4 into the ventricle and hand injects the LV gram with the expected non-diagnostic images, with wasted radiation and contrast. This consistently generates a 5-minute tirade from Bill as to how can this be allowed, what has happened to current operators, and so forth. We note this in the report, encouraging that this practice stop. 

The sad part to this is that when we return for follow-up, which is common for low-volume sites, despite our recommendations, there is no change.

Mort and Steve’s interest here should become a statement. Not just for billing issues, but for quality. If an operator is going to “administer” radiation and contrast, it should be for diagnostic quality or not given at all. Hand-injection JR4 LV grams are not diagnostic quality, and as “routine practice” for some operators, should not continue.

Mitch Krucoff, Duke University, Durham, North Carolina: This may be the first ‘conversation’ that has produced consensus, e.g. that 1) [There is a]  safety issue with end-hole catheters for LV grams, and 2) That hand injection is inadequate for LV chamber opacification and thus, not a satisfactory modality for assessing LV function. For experienced (e.g. old) angiographers who have lived from planimetry to the digital age, I have no qualms that a visual estimate of ejection fraction is both reasonable and reliable for clinical practice, and I do it all the time, recording my visual assessment in the procedure and/or progress note. In fact, with some new techs or fellows in the lab, we have made them go back and re-draw their “digital” calculations when they obviously missed something and the digital value is clearly wrong.

One other minor point. For the transradialists among us who use “universal” TRI [transradial intervention]-style catheters (like the TIG), the long-nose tips that typify these instruments also have a single side hole, ostensibly so that the LV gram, right, and left coronary injections can all be performed without a catheter exchange. Unlike the Sones [catheter], these catheter tips are angled and do not sit easily in the left ventricle. I do not think a single side hole is sufficient to avoid risk of traumatic power injections. Assuming we all agree that hand injections unacceptable, I think radialists should also be reasonable and avoid using universal catheters for LV grams (note: see case Figures 1-2).

John Bittl, Ocala, Florida:

1. Hand injections are inappropriate for LV grams and should not be used. All catheters used for power-injected LV grams should have side holes. A helpful tip for small-stature patients with low eGFR [estimated glomerular filtration rate] is to inject 20 ml/sec for a total of 15 ml at 900 psi through 4 Fr [French], 5 Fr or 6 Fr pigtail catheters. This approach allows edges to be defined for computer-assisted measurement of EF, but larger volumes up to 30 ml may be required for larger patients.
2. Visual assessment is appropriate after the cardiologist has calibrated the accuracy of his or her measurement of EF with a computer-based method (or planimetry) in several hundred cases. 

Ajay Kirtane, New York City, New York: I totally agree, and just this weekend at the AIM-RADIAL meeting, saw an unbelievably scary angiogram (sent to Ian Gilchrist) of exactly what wrong can happen with this increasingly common practice from the transradial approach. Even if using a universal catheter, we exchange out for the pigtail from a transradial approach.

Bernard De Bruyne, Aalst, Belgium: In Aalst, we are probably a bit old fashioned. [Regarding] LV angiogram, in each and every case [it’s performed] in biplane with a pigtail and 45 mL of contrast, except if GFR<35, mechanical aortic prosthesis, or deep cardiogenic shock. The main advantage over echo is the fact of obtaining the complete information by the same operator, at the same time, the same place, that very place where the decision about revascularization has to be taken.  

The EF is always quantitatively (LV volumes are often more important than the mere EF) done by the nurses.  

The LV angiogram [is also done] at the beginning of primary PCI. It takes 3 minutes, which are outweighed by several advantages: 1) immediate risk stratification (LVEDP, EF, MR [mitral regurgitation], etc.); 2) exclusion of massive MR, VSD [ventricular septal defect], Takotsubo, myocarditis); 3) indication about the culprit vessel, 4) evaluation of the need for IABP [intra-aortic balloon pump] or other devices; 5) all-in-one approach (no need for echo); 6) documentation of baseline LV function.   

Gregg Stone, New York City, New York: This [opinion] is not old fashioned, Bernard. I agree that your description is the absolute best clinical practice in all regards (except I always perform an LV gram in true cardiogenic shock, after a hemodynamic support device is in place, but before PCI — this is when VSD, papillary muscle rupture, etc., are most likely to be seen — unless a high quality echo was performed first). The baseline LV gram before coronary angiography and PCI is critically important, as first taught [to me] by Geoff Hartzler. In addition to the reasons you’ve outlined, it also provides critical insight as to why the BP [blood pressure] is falling during coronary angiography/PCI, as it so often does – whether due to hypovolemia with normal or near normal LV function, or decreased cardiac contractility with contrast injection and/or ischemia if baseline LV function is severely depressed. As you state, the 3 minutes is more than compensated for by the additional information the LV gram and measurement of LVEDP provide. 

[On another topic] This could be another discussion train, but we have to stop worshipping at the altar of DBT [D2B time] — as we’ve now seen from NCDR [American College of Cardiology’s National Cardiovascular Data Registry] and at least 3 other studies, incrementally decreasing DBT in the last several years by as much as 30 minutes has not resulted in further reductions in mortality. We have to pay attention to every aspect of the care of primary PCI patients, from pre-cath stabilization, to discussion with the patient to assess when DAPT [dual antiplatelet therapy] compliance is likely (and therefore DES [drug-eluting stent] is appropriate), to loading (absent contraindications) with statins, IV beta-blockers, potent ADP antagonists (soon to be cangrelor), preventing bleeding with bivalirudin +/- radial intervention in expert hands, to best PCI technique (e.g. IVUS [intravascular ultrasound] when useful), consideration of appropriate staging strategies, post-PCI use of defibrillator vests when appropriate followed by ICDs/CRT [implantable cardioverter defibrillators/cardiac resynchronization therapy], etc.

Malcolm Bell, Rochester, Minn.: I could not agree more, Gregg! It seems we have a generation of operators who either don’t know how to perform a good quality LV gram (as illustrated by this discussion) or are afraid to in emergency situations (incredibly useful in shock cases), without much justification. My only 2 cents is that I vary contrast volume and rate of injection depending on size of ventricle and what information you are looking for, and keeping catheter off the septum to avoid PVCs.

Mladen Vidovitch, Chicago Ill.: I can imagine that with the use of power injectors such as ACIST the temptation to perform a power injection through a universal radial catheter (TIG, Jacky or similar) is too great to be ignored.  I strongly believe a pigtail is the way to go rather than an end-hole catheter — the small side holes on TIG/Jacky are just not sufficient.

The bottom line

Mort Kern, Long Beach, Calif.: This ‘conversation’ expresses our current opinion on using an end-hole catheter with a hand injection and calling it an LV gram. In agreement with my colleagues, I think the time has come to stop this practice and perform a good quality LV gram when indicated, and with the right technique for safety and accurate interpretation.

References

1. Kern MJ, ed. The Cardiac Catheterization Handbook. 5th ed. Elsevier: Philadelphia, Pennsylvania; 2011: 1-456. 
2. Schwartz BG, Burstein S, MD, Economides C, Kloner RA, Shavelle DM, Mayeda GS. Review of vascular closure devices. J Invasive Cardiol. 2010 Dec;22(12): 599-607.