[Editor's Note: Read Dr. Steve Goldberg's Letter to the Editor regarding this discussion and watch the webinar that took place on August 28, 2020 [video below]).
This “Conversation in Cardiology” focuses on how patients needing a patent foramen ovale (PFO) closure in a community hospital are best managed. It is a continuation of an older discussion about how the transmission of high quality, high volume lab techniques and structures occurs for smaller community hospitals without the same infrastructure as a major medical center. At one time, percutaneous coronary intervention (PCI) was thought to be too complex for small hospitals, particularly those without surgical backup. PCI is now a commodity procedure, done nearly universally across many types of medical centers, with reportedly good quality and safety. Of course, there are exceptions and, importantly, good judgment is critical for the decisions on how/where procedures are done for the highest complexity, highest risk patients.
Does PFO closure fall into this high risk/complexity category? The issue was raised by our colleague, Prashant Kaul, MD, Director, Cardiac Catheterization Laboratory, Piedmont Heart Institute, Atlanta, Georgia, who asks, “Should PFO closures be performed at community hospitals where there are no other structural/congenital procedures performed (currently or previously) and where there is no on-site cardiothoracic (CT) surgical backup? Is there any precedent for this nationally?”
For background, Dr. Kaul recaps the Society for Cardiovascular Angiography and Interventions (SCAI) expert consensus statement on the operator and institutional requirements for PFO closure for secondary prevention of paradoxical embolic stroke,1 which suggests that “the institution should have performed >100 structural/congenital catheter-based interventions in the year leading to program initiation. On an ongoing annual basis, the institution should perform at least 50 structural/congenital procedures, at least 25 of which involve atrial septal interventions and/or 12 specific to PFO device closure. There should be ready access to an active cardiothoracic surgical program. While it is preferred that this program exist on site, rapid transfer to a nearby facility which offers cardiothoracic surgery may be acceptable in some carefully considered circumstances.” Where does this put a small hospital’s program (and its operators) as it attempts to become bigger?
Mort Kern, Long Beach, California: If there is operator expertise with PFO/atrial septal defect (ASD) in a lab sophisticated enough to manage the procedure and complications (eg, expertise in intracardiac echo [ICE], standby mechanical circulatory support [MCS], and anesthesia support, etc.), then a full-service structural program is not required. Certainly, a higher volume and more experience is always better than doing the rare case. Regulating procedures by volume alone is a challenge. For example, if a high volume, very experienced operator moves to an Arizona community hospital and wishes to continue most of his procedures, should he be restricted because of the low volume setting? The same conversations were held years ago regarding complex PCI procedures in community hospitals, some without surgery on site. For full disclosure, at the Long Beach VA, we are a low volume lab and continue to offer PFO/ASD closures as indicated clinically.
Chris White, New Orleans, Louisiana: Why would a high volume, very experienced interventionalist move to a low volume community hospital and “want” to continue to perform “highly complex, low volume” procedures? [Was he or she hired to build a program? –MK]. Everything we know about complex interventional procedures tells us that volume and quality/safety go hand in hand. Not just for the operator, but for the support staff in the cath lab and critical care areas. These procedures (PFO) are not emergent. We are not talking about the pros and cons of primary ST-elevation myocardial infarction (STEMI) in rural hospitals. There is plenty of opportunity to safely transfer a patient to a high volume “hub”. This is not an economics argument, because the volume is so low…10-12 cases per year. This is an “ego” argument. What is best for the patient versus what is best for the hospital. I continue to believe that “high complexity, low volume procedures” (carotid stenting, structural/valve procedures) should be done at high volume “hubs”, not in the “spokes”.
Bonnie Weiner, Worchester, Massachusetts: I agree with Chris. It is not just the operator, but the entire experience and culture. Quality is not just about successful procedures, but about the whole environment. If a “high volume” operator chose to go to a “low volume” facility, there is likely a reason that we are unaware of. Not the best for the patient to have more complex, structural or other “low volume” procedures performed in those environments. [“…likely a reason we not aware of” sounds nefarious, but could such an operator have been hired to grow a program? -MK]
Neil Kleiman, Houston, Texas: The other side of this argument is that so-called PFOs can turn out to have other accompanying features that have to be recognized. Although most are technically easy to close, I don’t see a patient benefit to disseminating it beyond operators who have expertise in evaluating for hidden congenital anomalies.
Steve Bailey, Shreveport, Louisiana: Best practice for PFO closure involves imaging (echocardiography, transcranial Doppler [TCD], computed tomography, magnetic resonance imaging [MRI]) that can illustrate and elucidate the issues surrounding the reason PFO closure is being considered, and be able to discern other congenital etiologies, etc. The evaluation should occur with a neurology colleague, which may not be occurring in many communities. Optimally, the cath lab staff will be active participants (ICE or rarely, echo). Better patient selection, device selection, plan of care, and outcomes are related to the expertise of the entire care team. While we may have done a lot of PFOs, the outcomes will be determined by our entire care team. These procedures are elective and do not require prolonged hospitalization or recovery time away from home, decreasing the urgency/need for local care. While complications are unusual (embolization, perforation, etc.) early recognition and treatment while stabilizing the patient and awaiting transfer to a site prepared for higher level of care may stress a lab that is unused to the higher level of care. They also may not have support devices in their center.
In centers who have focused on this, PFO without surgery backup works well. I do not believe that most community hospital sites are recruiting experienced operators. Early career, low volume operators who are inserted into teams with little experience is the more likely scenario. Any benefit seems to accrue to the operator, not to the patient from this scenario.
Lloyd Klein, Sonoma, California: There are community hospitals and there are community hospitals. Some are large and vibrant; others are quaint, to be kind. I’ve worked in some that had such programs, even helped start a PFO program at one, and it’s done very well. Others, that would be a truly nutty idea. I think it depends on the hospital and what its resources are. I would eschew any “broad brush” answer to this question. Specifically, the absence of CT surgery or other structural procedures would not be a major concern, but the ability for the operator and staff to deal with tamponade, perforations, etc., would have to be clearly established.
Jonathan Tobis, Los Angeles, California: The writing committee for the SCAI statement1 had a moderate amount of discussion about this issue. Our concerns were similar to those expressed above. We tried to weigh the right of operators to have access to new procedures versus the right of patients to be treated by an experienced physician in a protective environment. PFO closure is relatively simple, but there are some unusual situations. ASD is more complicated and is best reserved for trained individuals who can cope with an embolized device. It is difficult to make broad guidelines that will accommodate every situation.
Neil Kleiman, Houston Methodist, Texas: Do operators have “rights” to new procedures? [This is an entire conversation in itself. -MK]
Larry Dean, University of Washington, Seattle, Washington: Basically, why do this [low volume, high complex procedure in small hospital]? I’d also point out that it’s not only hospitals that might want this, but there is also a revenue stream to the provider.
Ken Rosenfeld, Boston, Massachusetts: These questions apply not just to PFOs, but almost every higher end or specialized procedure we do, and the same principles should apply. The big question is how the paradigm for the procedure should be monitored and enforced. For example, who will distinguish between the smaller volume place like Mort’s which, as Steve says, has focused on these procedures, versus the smaller (or larger) hospital which has not. Who will evaluate and how?
This issue is not unique to cardiology…in almost every specialty, there are higher-end procedures that require an infrastructure and multi-specialty colleagues for support. This struggle [around who and where the procedure should be performed] exists for them, too…which hospitals, which operators, etc. It’s hard to know the best solution.
View From Abroad:
Bernhard Meier, Bern, Switzerland: With due respect, I like to bring in a different aspect. The most important thing about PFO closure is that it be performed more frequently. After all, it may well be the procedure in interventional cardiology with the best net yield. Hence, I have always supported the performance of PFO closure in every catheterization laboratory. It is not more difficult or dangerous than diagnostic cardiac catheterization (Figure 1) and saves myocardial infarctions, strokes, and lives, among other things. Small ASDs are as easy to close as PFOs, but they are less clinically relevant and thus less important.
View From a Community Hospital Operator:
Sam Butman, Cottonwood, Arizona: I read the opinions, and fully understand the logic and background to each, given my having worked in an ivory tower for 21 years and now in the “boonies” for 13 years at what is a very small community hospital (as written earlier by an esteemed colleague above…we are quaint). However, we were one of the first in the country to do primary STEMI PCI without on-site surgical support (50 miles away, but helicopter supported). We still have not sent a STEMI or elective PCI emergently away (knock on wood)!
There is expertise with PFO/ASD by the operator in this not as sophisticated lab, but this operator does not do ASD cases here for reasons others have suggested and because he knows that someday he might have to spend some “extra time” trying to snare a larger device that may have “moved” (but so far, never). The same conversations were held regarding complex PCI procedures in community hospitals, some without surgery on site. We do those, but we do not do elective left main or Rotablator (Boston Scientific) cases, and can happily still reference our lack of helicopter use to date. I agree with Dr. Meier that “PFO may well be the procedure in interventional cardiology with the best net yield. It is not more difficult or dangerous than diagnostic cardiac catheterization in experienced hands.”
Times have changed and as another opined, “I like our current de-centralized approach that puts the authority and burden of approving/maintaining privileges on the hospitals. This lets the hospital determine what’s best for the community it serves.”
Being here in the hinterlands has been a roller coaster ride of administrative and personnel changes, but the cardiovascular care has been more than fine. We all watch what we are doing and more important, so do others. That is a view from actually being out there.
David Kandzari, Atlanta, Georgia: To clarify, however, the question proposed to the group is not the scenario of an experienced operator transitioning to the ‘quaint’ community hospital, but instead more specifically, the situation of 1 or 2 skilled interventional cardiologists who formerly did PFO closure in their training, but now want to offer it at their hospital without surgical backup. At present, the cases are sent to the central main hospital, a quaternary referral center for complex coronary and structural cases. The physicians are willing to retrain with proctorship, but want to offer the procedure locally. The situation raises the juxtaposition of procedural volume, experience and providing local care to residents. I hope that provides some perspective on the reason for the question.
Prashant Kaul, Atlanta, Georgia: From the perspective of the writing committee of the consensus document,1 the suggestion that “the institution should have performed >100 structural/congenital catheter based interventions in the year leading up to program initiation” would seem to preclude starting a new program in a community hospital that has not previously performed any structural or congenital interventions. Was this the intent of the writing committee?
Jonathan Tobis, Los Angeles, California: Eric Horlick was the chair of the writing committee, so he should probably represent what the consensus opinion was, but my take is that the committee did not recommend PFO closure be done by anybody in any setting. The hospital needed to demonstrate expertise in dealing with left heart procedures and that is where the number of cases of structural or congenital heart disease came from. It is a demonstration of experienced personnel and laboratory.
Chris White, New Orleans, Louisiana: Sam, David, and Lloyd, thanks for your perspectives. No one doubts that the experienced MD has the skill and expertise to do these procedures successfully. The question is whether the “juice is worth the squeeze” when there are high volume referral centers available for elective referrals. Maybe PFO closure is not seen as “highly complex”, but certainly it is a very low volume procedure. Regarding performing high complexity, low volume procedures in rural community hospitals, what would you recommend? Would you recommend a finishing transcatheter aortic valve replacement (TAVR) fellow set up a TAVR program in your hospital to do 12-15 TAVRs a year? What about doing 5 MitraClips (Abbott Vascular) per year? Where does it end?
(continues below video)
WATCH: August 28, 2020 Discussion with Drs Sam Butman and Lloyd Klein.
Mort Kern, Long Beach, California: For the record, I had no one person in mind for the discussion about low volume centers doing complex procedures. For PCI, the battle to retain centralized control was ceded to the community hospitals, but we all believe that there are procedures well beyond community level ‘complexity’ requiring a full service interventional/structural program to ensure safety. Our low volume program moves patients to the university lab for just such scenarios.
Sam, I apologize if you thought I was talking about you; I was actually talking about me. I moved to California (2005) to exactly that scenario, then moved into the university/VA relatively low volume programs. I’d love to do more complex cases at the VA, but logic prevents it.
Many opinions in this and other Conversations in Cardiology come from my ‘list’ of cath lab experts which is comprised mostly of “ivory tower” operators. Community operators, like Sam, are underrepresented. I also think that the interventional community has always struggled to address the “town and gown” aspect of technical and clinical dominance between the average practitioner and the ivory tower operators. Many ivory tower guys, including me, have an inherent bias known as “illusory superiority”. For example, when asked “are you an average driver?”, the majority of people consider themselves to be well above average drivers, but this is a statistical impossibility. IC guys also believe they’re better than they are. Have you ever met one who wasn’t the best IC guy around (particularly me [JK...])?
How does a small program become bigger? Get a well-experienced ivory tower guy to move from a bigger center to a small center and carefully build.
Lastly, if a family member needed a complex procedure or TAVR, wouldn’t you recommend the most experienced and most successful place for the best result? This is not in debate. How do we then explain continuing any small program that is not full service? Local centers supply a great need and services to semi-remote regions. It’s not the need but rather the level of complexity (and safety) that is the question raised here.
Sam Butman, Cottonwood, Arizona: TAVRs, clips, and complex procedures, even including complex PCI, should definitely be done at high volume centers. One might be surprised to learn, as I have been, how many people really do not want to go to the “city” when they live in the boonies. It really is something. Trust me, it’s not worth the few RVUs for the time involved.
Bonnie Weiner, Worchester, Massachusetts: Since I have and do live in both worlds, I was the lone voice at my facility (even though we have CT surgery) to suggest that we shouldn’t do complex structural procedures, since there are multiple high volume facilities nearby (no, we are not in the boonies but many in Boston think we are). I am all about quality and providing the best care to the patients, and not necessarily what is thought to be in the hospital’s or individual operator’s (including me) best interest. There is no “one size fits all” in this situation. I, like Chris, do not see the value to the patient to have low volume procedures at a facility that rarely performs them. Again, it’s not just the operator, but the entire system of care at that facility is important. We all think we do a good job, but my experience has been that even under the best of circumstances, there are opportunities for improvement. It is in these areas where those opportunities may be most important that certain procedures may be identified which shouldn’t be performed.
Lloyd Klein, Sonoma, California: Coronary PCI is a mainstay of treatment procedures. It is being done at most community hospitals highly successfully; when a complex PCI arises beyond local capabilities, it is sent to a larger center. This is a sign of its incredible success. Many patients don’t want to go to the biggest places or the university hospitals, even if convenient; they may not feel like they are treated kindly or individually at a big, impersonal place. People choose their care for many reasons beyond physician experience, and as long as whom they choose is competent, it is how things work.
Valve cases remain in the phase of development just beyond pioneering, but still where lots of places are in a learning curve. They should only be done in large centers right now. I am less certain about where PFO cases stand; someplace in-between, though. The operator and staff must be fully trained, be experienced, and able to manage complications that are sure to arise at times. IF it can be done safely, all procedures will eventually filter into the community. Why are we training fellows to do them? Why are we studying how to make them better and simpler? I highly respect the argument that too few cases suggest limitations in delivered care, but I reject the idea that the operator with the most cases has the best results or provides the best patient experience. That leaves a wide-open gray zone. If Sam or Mort feel that they can do these cases safely, and have a record to back that up, our profession ought to support them.
Chris White, New Orleans, Louisiana: I completely agree that this topic is broad and complicated, and I’m not advocating for one solution for all, but a recognition that “in general”, higher volume centers achieve better outcomes with complex procedures. Carotid stenting is a great example, with lots of evidence to support the correlation of experience/volume of operators and hospitals with outcomes. There are other examples that inform my opinion about performing complex procedures at low volume venues.
CABG is a great example, where the revenue from even 30 cases per year was important for some of our community hospitals and as we tried to centralize this “commodity” procedure to the medical center, I got a ton of pushback. None of those CEOs (or their family members) would have had their CABG in their own institution, where the nurses were much less experienced, the anesthesiologists were less experienced, etc. This is why I don’t trust hospital administrators to regulate or “police” their institutional quality. Their conflict of interest is too great.
Our last 3 TAVR fellows have not been able to find satisfying jobs in high volume (150 cases per year) TAVR centers, but have landed in low volume hospitals lucky to do 2 procedures a month. A finishing TAVR fellow in a low volume program is a recipe for disaster. They are incentivized by compensation and desperate for professional satisfaction to do more cases. That leads to taking more risk or perhaps doing procedures that shouldn’t be done. We are training more structural/valve interventionalists than we need and sadly, no one is explaining this to the prospective fellows.
Kirk Garratt, Newark, Delaware: Our electrophysiology (EP) colleagues have fought against having procedural volume targets built into credentialing and certification (including MOC) requirements for some of the reasons discussed. For the most part, they’ve been successful. Good EP operators flourish and the bad ones are contained, regardless of volumes. We’re better at addressing bad operators in America than we’ve been in the past, helped along by hefty penalties if hospital quality measures aren’t so good.
I like our current de-centralized approach that puts the authority and burden of approving/maintaining privileges on the hospitals, not the state or, worse, the federal government. This lets the hospital determine what’s best for the community it serves. Hospitals can be motivated by greed and make bad choices, but one really great consequence of payment reform is that all hospitals are driven to manage quality. They have everything they need: consensus around needed safety rails (facility requirements, ability to handle emergencies, etc.), statements that define outcome expectations, resources to audit labs when trouble is suspected, and organizations to help remediate problems.
The Bottom Line:
Here are my take-home messages:
- Patient safety is always the first and foremost consideration regarding where a procedure should be done.
- Patients value the convenience and perceived expertise of the community hospital.
- PFOs are not as high in complexity for most patients as some PCIs and certainly are lower than TAVR/MitraClip procedures. For any procedure, but particularly PFO closure, the entire interventional and support services of the program, and not just the operators’ experiences, must be included in determining whether a procedure can be done in a small hospital without on-site surgical backup. Local policy determines what can and can’t be done at their hospital.
I hope this discussion brought some insight on how and where PFO closures can be done.
Disclosures: Dr. Morton Kern reports he is a consultant for Abiomed, Abbott Vascular, Philips Volcano, ACIST Medical, and Opsens Inc.
Dr. Kern can be contacted at email@example.com.
On Twitter @drmortkern
- Horlick E (Chair), Kavinsky CJ (Co-Chair), Amin Z, Boudoulas KD, et al. SCAI expert consensus statement on operator and institutional requirements for PFO closure for secondary prevention of paradoxical embolic stroke. The American Academy of Neurology affirms the value of this statement as an educational tool for neurologists. Catheter Cardiovasc Interv. 2019; 93(5): 859-874. doi:10.1002/ccd.28111