On September 26, 2018, JAMA Cardiology published an important paper1 reporting the safety and reduced costs of same-day discharge (SDD) for elective percutaneous coronary intervention (PCI). The potential cost savings for this approach per case was about $5000 and if applied across the country would likely be in excess of $500 million.
At the VA Long Beach, Dr. Arnold Seto, Chief of Cardiology, and I implemented SDD several years ago. Our default for PCI discharge is the same day, mostly prompted by our radial success rates. We are still reticent to send most of our femoral patients home the same day because of our irrational fear of late bleeding, which almost never occurs or is not reported in the literature. Probably like many others, we err on the side of being conservative, which is exactly what this latest paper says we don’t have to be concerned about. We discharge everyone we can after elective PCI (~40%), but are still limited by our VA patient population (5-10% homeless, up to 30% living alone).
Based on our experience and the recent paper by Amin et al1, I asked for words of wisdom from our cath lab expert group regarding their experience with same-day discharge at their center. What are the obstacles to SDD? Who are the champions and why are we not doing it more?
What Does the SDD Data Show?
Amin et al1 examined 672,470 elective PCIs performed in 493 U.S. hospitals between 2006 and 2015. SDD rates ranged from 0 to 83% with a large variation among sites, in excess of 300% between hospitals. This variation occurred despite identical clinical profiles between SDD and non-SDD patients and was unchanged over the 9-year study period. As we might expect, rates of radial access (8.9% vs 3.4%), use of fractional flow reserve (FFR) (5.7% vs 2.8%), and intravascular ultrasound (IVUS) (11.5% vs 9.7%) were higher in the SDD patients. SDD hospitals were more likely to use low-molecular-weight heparin for their procedures (1.7% vs 14.2%) or glycoprotein IIb/IIIa inhibitors (10.4% vs 20.2%). Importantly, patients discharged the same day after an uncomplicated PCI did not have higher risk of death, bleeding, acute kidney injury, or acute myocardial infarction (MI) at 30 days, 90 days, or 1 year compared with the non-SDD patients. Associated with SDD, the cost savings were estimated at $5,128 per procedure due to supply and room-and-board costs.
It is also worth noting that a very similar outcome was reported in the American Journal of Cardiology (September 2018, online) by Kwok et al2, in further support of SDD safety and cost effectiveness. They looked at 324,345 patients in the United States who underwent an uncomplicated elective PCI between 2010 and 2014. Length of stay ranged from 0 (same-day discharge) to ≥3 days. At 30 days, rates of unplanned readmission were lowest among SDD or 1-day stay (4.8% and 4.7%, respectively) and were highest for those with ≥2-day hospitalization (6.4% and 9.4%, respectively). Similarly, costs ranged from $15,063 for same-day discharge to $14,693 for a 1-day stay, $18,136 for a 2-day stay, and $24,336 for ≥3-day stay. The likelihood of readmission was not significantly higher for SDD or a 1-day stay, but it was greater for patients with ≥2 days or ≥3 days.
What Are You Doing at Your Hospital?
Jim Blankenship, Danville, Pennsylvania: We started SDD in 2007 for a relatively restricted set of patients, all with femoral access. We started routine radial access about 2011, which gave SDD a big boost, and since then, we have liberalized discharge criteria. We send home nearly every stable outpatient with an uncomplicated procedure, who remains stable for 5 hours after the procedure, and wants to go home. I never insist on someone going home when the patient or family is uncomfortable. We send home about 80% of outpatient PCIs.
I call every patient 24-48 hours post discharge. For over 400 patients since 2007, only 1 was readmitted within 24 hours for non-cardiac chest pain. When asked, about 98% say they were glad to go home the same day rather than stay overnight. The typical comment is “I enjoyed sleeping in my own bed.”
Barriers to SDD in our institution are the lingering concern over adverse events during the night after discharge and the fact that the advanced practitioners go home at 5pm, leaving the PCI patients in the recovery suite under the care of RNs and fellows. No one really wants to put in the time/effort to do a final discharge evaluation after about 4:30pm and it is easier to let the patients stay overnight.
Regarding costs, we presented an abstract (never published) based on charges (we couldn’t get cost data) that showed decreased charges. Of course, charge data is of very limited value. Others have noted the marginal costs of keeping patients overnight are not large. Our recovery suite is open anyway, staffed with night shift nurses. The bed that is vacated by a SDD patient is NOT filled by another patient overnight. The marginal costs are basically the cost of the patient’s breakfast, their morning meds, and the electricity they use watching TV. Of course, if they were occupying an inpatient bed, the story would be quite different.
Adhir Shroff, Chicago, Illinois: We have been doing SDD for several years now and it has been a very popular program. I would say that the focus is identifying stable patients, successful procedures, and a structured process. This was the thrust of the Society for Cardiovascular Angiography and Interventions (SCAI) length of stay (LOS) paper.3 We have liberalized the early discharge criteria over the years as we have gained experience.
Bernie Meier, Berne, Switzerland: Same-day discharge (and admission) has been common at our center for more than 10 years for approximately 30% of PCI patients (femoral or radial, no or multiple stents including left main [LM]), 20% of left atrial appendage (LAA) occlusion patients, and 100% of patent foramen ovale (PFO), and atrial septal defect (ASD) closure patients. All transcatheter aortic valve replacement (TAVR) and MitraClip (Abbott Vascular) patients so far have stayed at least 1 night. TAVR has been done in Germany on outpatients.
Lloyd Klein, Chicago, Illinois, and Sonoma, California: Traditionally, the fear has been that a complication, such as bleeding or stent closure, might occur in the early hours after discharge. As we know, there is little data to support that fear, and some data that suggests it doesn’t happen. There is also a medical-legal concern around the fear of late events. With the radial approach, one would think that the likelihood is small. To do [SDD] correctly, you need dedicated personnel and holding area space. Some hospitals don’t have space/personnel or won’t invest in that process. Other obstacles include whether the clinical situation can be explained to the patient well enough in that [shorter] period of time (e.g., need for dual antiplatelet therapy [DAPT], follow-up, etc.), and how far away from the treating center the patient lives in case something does “go south” (i.e., a complication occurs). In Illinois, same-day admission has been interpreted to be 23 hours, so with careful planning, the patient can be kept overnight with no financial incentive to move faster.
James Tcheng, Raleigh, North Carolina: Great topic. The SDD analysis was good, but the conclusion and recommendations were bad, particularly for the uninitiated health care administrator. I’m fully supportive of SDD where appropriate, but I would NOT use the manuscript as carte blanche, broad brushstroke justification for increasing SDD. The authors (almost completely) miss the largest confounder — MD wisdom and decision making in terms of who is suitable for SDD. There is only 1 sentence that attempts to directly address this in the limitations section: “Second, angiographic details and procedural complexity are not captured in our data and there is potential for unmeasured confounding.” Instead, the analyses in the manuscript1 are indirect statistical arguments that the analysis did not miss confounders, rather than trying to actually measure and adjust for (the obvious) confounders. Furthermore, how do you adjust for what MDs are trained best to do … aggregate large amounts of varied and disparate information/MD wisdom?
Paul Teirstein, La Jolla, California: We are at about 80% SDD at Scripps. We keep unprotected left main PCIs overnight since a complication can result in rapid death. Also, if the patient has nobody at home, they stay overnight. Use of Impella (Abiomed) has to be associated with being an “in patient” for proper billing. Other than those two conditions, if the patient looks good and wants to go home or to a hotel, they go.
Malcolm Bell, Rochester, Minnesota: We have been doing SDD for elective PCI for a few years now. Initial barriers were a ‘resistance to change’ mentality, convincing people that it was safe, and not listing a lot of exclusion/inclusion criteria where you could find one (unproven) reason not to proceed. Also, changing the mindset of referring physicians (and patients) and instructing them that their patients will go home or stay locally as the default so that the patient is not surprised or concerned when suddenly someone says “you are OK to go home now.” Also, there are risks of staying in hospital, too — falls, medication errors, infection, delirium, etc. The establishment of a SDD program required a lot of hands-on encouragement, but is now our accepted standard practice. Our eligibility criteria:
- Elective PCI;
- Radial or femoral access (<7 French);
- Stay within radius of 100 miles;
- Adequate social support.
Patients get a green light to go if there are no significant unresolved PCI complications or their dismissal is done by a certain evening hour (nursing driven). We don’t do SDD in pre-TAVR PCI patients. A follow-up phone call next morning by a cath lab nurse to the patient has made the program highly successful without any significant events. Patients love it.
Gurpreet Sandhu, Rochester, Minnesota: As Malcolm [Bell] pointed out, our nurses have played a major role in making SDD successful. Any elective PCI completed by 5-5:30pm followed by the standard observation after sedation and vascular access can get most people home by 9:00pm. It’s also worth noting that we have the same discharge guidelines for both radial and femoral access, and we do not routinely use closure devices.
Pinak Shah, Boston, Massachusetts: Our SDD is used for elective PCI, PFO/simple ASD, and peripheral cases. SDD has been the standard of care at Brigham and Women’s Hospital for the last 9 years for both radial and femoral cases. We will hold on to patients if there is an untoward event in the lab or if the patient doesn’t have support at home. We ensure that patients with chronic kidney disease (CKD) get their renal function checked in 48 hours as outpatients. One of our physician assistants makes a call to the patient to ensure they are all right and answer any questions. We have had two unexpected returns to the emergency department — one for recurrent angina that was likely due to another lesion that was not treated at the index procedure and one for cholecystitis.
Our major issue has been timing. If the case happens late in the day, we don’t like the idea of sending people home very late at night. It is a challenge for us to calculate the economic impact of SDD here, mostly because of opaque finances, which I am sure most of us deal with. Nonetheless, the patients love it [SDD] and it keeps beds open for people who need them.
Sam Butman, Cottonwood, Arizona: From this rural community hospital perspective, the average age of PCI patients is 10 years older than those I saw when I was at University of Arizona, Tucson. These often-elderly folks live 20 miles away to the north and west of us; however, SDD is the norm, but perhaps less common for a number of reasons. The femoral approach PCI patient I did this week had the puncture sealed (i.e., use of a vascular closure device) and spent the night because he lives 20 miles away and had his opposite knee arthroscopy done the day before. I do keep patients after PCI if it makes clinical sense for the individual and err on the conservative side. However, it is never just an automatic, although 90% go home the same day. Some of my colleagues have a lower SDD rate due to higher use of femoral approach in most of their procedures.
Fortunately, no administrative pressures exist whatsoever. Use of the radial approach makes distance to travel after the procedure a non-issue. You asked who the SDD champions are. Oddly enough, there are no champions, just the institutional knowledge that SDD has been safe and effective with essentially no returns for complications when sent home that day.
Habib Samady, Atlanta, Georgia: For the past 4-5 years, we have been sending the majority of our elective PCIs home the same day. I had performed a literature review and had consulted with Sunil Rao at the start of our program. Our criteria for SDD include:
- Achieving hemostasis by 5pm (radial or femoral);
- Having no very high risk features or complications during PCI;
- Living within 30 miles (if they live further away, we encourage them to stay at a local hotel);
- Having adequate social support;
- Being free of new cath-related issues at the time of discharge.
The remaining obstacles include patient preference and staff shortages for late shift. However, the comfort level among staff and physicians has improved with this experience. The financial impact of overnight stay vs SDD for the patients is complex and variable based on their insurance. The patients incur the cost of the hotel stay.
Molly Szerlip, Plano, Texas: We at the Heart Hospital have a 60% same-day discharge rate for PCI. The rate would be higher, but we have some older operators who are still not comfortable with [SDD], though this is changing. SDD does save the hospital money, especially in utilization and bed costs. Barriers to SDD are mostly education. We have seen that radial access as well as femoral access makes it safe to send patients home, especially when a femoral puncture closure device is used. We have moved to being less conservative with no increase in rehospitalization, morbidity, or mortality. We also send balloon valvuloplasty as well as many structural procedure patients home the same day.
Jeffrey Popma, Boston, Massachusetts: As a more senior physician in our group, I was completely wrong in using my years of [femoral] experience to object to SDD. As my younger partners have taught me, SDD can be performed in the vast majority of uncomplicated outpatient PCIs (we are at approximately 70% at the Beth Israel Deaconess Medical Center). I used to call the patients at night to check on them so I could sleep at night, but I usually just woke them up from their restful sleep. A follow-up call the next day is more than sufficient. We are even using SDD at one of our outlying hospitals without surgery on site for our elective PCIs. So I guess I am a convert, having once been a resister. Evolution is a key survival technique in interventional cardiology.
David J. Cohen, Kansas City, Missouri: The key to same-day discharge for PCI is having or developing the appropriate infrastructure to make it relatively seamless — in particular, ancillary staff who can evaluate the patients and accomplish all the appropriate goals of care (education, discharge prescriptions/appointments/etc., follow-up surveillance) to ensure that it can be done smoothly. Also, don’t expect to achieve real cost savings of anything like those in the Amin paper.1 This is not a critique of the study itself. Amit was my fellow, so I know he does quality research — it’s just an issue with cost accounting systems. The cost savings he has calculated are “fully loaded”, meaning that they assume you can reduce all sorts of things like nursing staff, administrative overhead, etc., which may be possible in the long run, but not in the short-to-intermediate term, when many of those costs are fixed. The best economic argument for SDD is actually the ability to put another paying customer in that same bed. If you can do that, it’s definitely a win-win. If you can’t, the true savings to the hospital are likely to be smaller.
Sunil Rao, Raleigh, North Carolina: I rarely comment on these conversations, but reading the full-throated support of SDD warms the cockles of my heart. In 2011 when we published on this, we got lots of pushback and even some hate email (really).
Duane Pinto, Boston, Massachusetts: As a co-author on the Amin paper1, I thought I’d add a few words. SDD requires a commitment from the operators, nurses, and the administration to do it safely and the process needs evolve in scope over time. As others have noted, we had a nurse do a follow-up call, but that fell by the wayside, because no one was having problems next day. There were mostly clinical questions and these can be handled by the hotline for patients to call (staffed 24:7). We keep the holding area open to 11pm so that some cases can be discharged late with a ride. I shared Jeff [Popma’s] and Binny [Pinak Shah’s] worry about late discharge, but patients want to sleep in their beds and [don’t mind] leaving late. We have an early TR Band (Terumo) removal protocol that also helps with LOS in the holding area and the patients are “primed” for discharge by the intake nurses. We evolved from just “easy cases” to femoral closure cases, peripherals, and even a few bilateral 8 French chronic total occlusions, as well as the other types of structural cases. It took years to get here. We are glad our senior guy (Dr. Popma) saw the light.
The Bottom Line
Whether the operator selects radial or femoral access, same-day discharge is safe in most moderate- to low-risk elective PCI patients. Patients prefer sleeping in their own beds at home. Beyond the fixed costs of hospital operations, reductions in bed day costs and supplies can result in substantial hospital savings. Given the >1 million annual procedures in the U.S., pretty soon we will be talking about real money savings related to better patient care and procedural satisfaction.
- Amin AP, Duane Pinto House JA, et al. Association of same-day discharge after elective percutaneous coronary intervention in the United States with costs and outcomes. JAMA Cardiol. 2018 Sep 26. doi: 10.1001/jamacardio.2018.3029. [Epub ahead of print]
- Kwok CS, Rao SV, Gilchrist Sr IC, et al. Relation of length of stay to unplanned readmissions for patients who undergo elective percutaneous coronary intervention. Am J Cardiol. 2018. doi: https://doi.org/10.1016/j.amjcard.2018.09.028. [Epub ahead of print]
- Dickens CA, Shroff A. Developing a same-day discharge program – how to identify appropriate patients for safe and efficient discharge. Cardiac Interventions Today. 2012 Jan/Feb; 19-25. Available online at https://citoday.com/2012/02/developing-a-same-day-discharge-program/.
- Seto AH, Shroff A, Abu-Fadel M, et al. Length of stay following percutaneous coronary intervention: An expert consensus document update from the society for cardiovascular angiography and interventions. Catheter Cardiovasc Interv. 2018 Oct 1;92(4):717-731. doi: 10.1002/ccd.27637.
Disclosure: Dr. Kern reports he is a consultant for Abiomed, Abbott Vascular, Philips Volcano, ACIST Medical, Opsens Inc., and Heartflow Inc.