Clinical Editor's Corner

When Should You Discharge a PCI Patient From the Hospital?

Morton Kern, MD
Clinical Editor
Chief Cardiology, Long Beach Veterans Administration Hospital;
Associate Chief Cardiology, University California Irvine;
Professor of Medicine, UCI
Orange, California

Morton Kern, MD
Clinical Editor
Chief Cardiology, Long Beach Veterans Administration Hospital;
Associate Chief Cardiology, University California Irvine;
Professor of Medicine, UCI
Orange, California

One of my recent interventional fellows, now a new staff interventionalist at a nearby hospital, wants to know when to discharge uncomplicated patients after percutaneous coronary intervention (PCI). He mentions that the guidelines are not specific on this point. With radial PCI, our fear of femoral bleeding problems is gone. Why keep patients overnight? How long should we keep patients after uncomplicated PCI?

The SCAI Consensus

To start this discussion, in 2009, the SCAI Consensus Document regarding length of stay1 states, “The current standard of care for a patient who has undergone an uncomplicated elective PCI is an overnight stay in the facility performing PCI. This is based on the following concerns: (1) abrupt vessel closure and its resulting complications, (2) access site complications, and (3) management of comorbidities, such as renal insufficiency, diabetes, hypertension, congestive heart failure (CHF), etc. Stents, newer antithrombotic regimens, and arterial closure devices have not completely eliminated these concerns.”

It is also worthwhile to review our terminology:

  • Outpatient: a same-day procedure with the duration of supervised care seldom extending beyond 12 hours;
  • Admission for observation: < 24 hours after elective PCI, the patient is kept in the medical facility with monitoring and nursing care overnight, but is sent home < 24 hrs; 
  • Extended observation admission: > 24 hours, patients with comorbidities requiring extended monitoring and potential treatment).  
  • Inpatient admission would be the result of a complication of PCI needing a higher level of care than simple observation.

Clinical scenarios

The timing of patient release is individually determined by the specific clinical scenarios.1 For example, the simple, uncomplicated 1- or 2-vessel stent patient from a radial approach or in some cases, femoral access, can be discharged the same day (8 hours), but many practices use a 23-hour observation, because of concern about femoral bleeding or reimbursement issues. A more complicated patient with symptoms or transient hypotension deserves admission and monitoring with an understandable delay in discharge. 

Should you keep the PCI patient longer because of a troponin bump after the procedure without frank signs of myocardial infarction? This practice is variable and as noted by Dr. Klein below, may be associated with quality issues.

I asked these questions of my colleagues on a cath lab directors’ email list, and share their very interesting responses below. 

Personalized medicine

From Worchester, Mass.: Some of the answers are not directly related to medical issues. The patient’s age, distance from the hospital, EMS in the local community, social support system, and time of day are all issues. Would you send an uncomplicated PCI home at 8 pm if the case is done at noon? Many patients and their families would not find that “patient friendly.” Others would be fine with that. My bottom line — I don’t think access site is the issue; I think it is about “personalized” medicine.
Bonnie Weiner

From Harrisburg, Penn.: I was first author on the SCAI paper1, and most of the literature since 2009 is from a radial perspective. I am in a lab that is >50% radial. There is limited reason to keep people overnight, if stable from a procedural and an access standpoint. Even hand-held groins (mostly 6 Fr), no closure device, pulled 2 hours after bivalirudin stopped, 4-hour bed rest after hemostasis, and then ambulation/1-2 hour observation, go home.
Charles Chambers

Early radial discharge in Canada

From Quebec City, Quebec: Based on the Dutch previous experience, we used 4-6 hours [discharge time] after PCI completion in the EASY trial. With radial PCI, we take into account hemostasis in about 2 hours and 2 hours for observation to avoid rebleeding. [In the EASY study], there was only 1 patient who presented with a right coronary artery (RCA) re-occlusion 2 hours after uncomplicated PCI. [However, early] discharge requires some ‘discussion’ with nurses to discharge patients at 11 pm.
Olivier F. Bertrand

Early discharge economics

From Temple, Texas: Our lab (Scott & White) has undergone a substantial transformation over the past 18 months, from all elective PCI patients staying overnight to now about 50% going home the same day. This has mostly been driven by reimbursement issues, which are somewhat at the discretion of the agency that does Medicare billing. In Texas, there is no additional bump in reimbursement if the patient takes up a bed overnight. Our colleagues who run the hospital are “delighted” to have us send PCI patients home the same day, as it keeps a bed open for other incoming patients and doesn’t use hospital resources that are no longer reimbursed.  Nurse satisfaction in our cath holding area is actually higher, as they are no longer are babysitting stable PCI patients late into the evening, waiting for a bed to open up. Doing more radial PCIs has helped, but we also send stable PCI patients with femoral closure devices home. 

We try to have our elective procedures done by noon and then have PCI patients stay until approximately 6 pm. If they live too far away, the hospital will actually pay for the hotel (across the street from the hospital), as it is still far less expensive than taking a bed. The only patients who stay are those with complications during PCI (small branch vessel occlusion, groin issues, etc.) or those we feel need extra hydration. I write this keeping my fingers crossed, but we have had no bad experiences as yet. We’ve had a few patients come back the next day, more because they are “worried” about something minor (usually the groin), but nothing serious has happened.
Greg Dehmer

From Chicago, Illinois: Each center can individualize their own needs and comfort level. At least for practices in the U.S., [I believe] the “standard of care” is to observe OP-PCI overnight either as a 23-hr observation (or inpatient admission). I think we all have to be careful when designing “same-day discharge” programs that we acknowledge that it is a voluntary, “pilot,” or observational program. We need to track outcomes and attempt to report the data so that we can alter the standard if the data suggests it.

My colleagues and I have developed a same-day PCI discharge program at the University Hospital and at the VA. From a financial standpoint, the VA was easy — keeping patients costs money. At the University, I found out some interesting information. This may be specific to our state or our institution. Government insurers (Medicare/Medicaid) do not pay an incremental fee for observation time. There is a flat fee for OP-PCI regardless of 8-, 10-, or 23-hour observation. Private insurers do reimburse the medical center a small amount per hour of observation time. Therefore, there is a small (~6%) loss in reimbursement between discharging a patient at 8 hours versus 23 hours. It was clear that opening up a bed for another patient was a priority.

The upshot to me is that same-day discharge for PCI is safe in selected patients who are part of a larger program with phone follow up, protocols, and order sets. In order to be safe and successful, it must be a team approach, and each physician cannot make it up as s/he goes along.
Adhir Shroff

From Chicago, Illinois: We do 23-hour observation routinely. Why? Because in Illinois, that maximizes the reimbursement and minimizes the perceived medicolegal exposure. There is no compelling medical reason for most patients to stay after 12 hours. There is a medicolegal issue in sending someone home “early” who has a complication (i.e., bleeding, stent closure), so the “natural” tendency is to keep people in as long as possible. Regrettably, most hospitals are unwilling to develop a dedicated center for observation and want to utilize the inpatient resources as much as possible. There are local administrative and political questions at play here, but not medical ones. There are no clear standards, and the SCAI document1 leaves a lot to judgment. There is no definite standard practice in this regard.
Lloyd W. Klein

From Seattle, Washington: We have a similar issue in Washington due to local coverage decisions and reimbursement, so they are billed accordingly. As Lloyd points out, billing as an observation/outpatient and how long the patient is actually in a bed are not clinically tied together, so our patients go home the next day.
Larry Dean

Early discharge criteria

From New York City: For Medicare patients, reimbursement is made at outpatient rates unless certain criteria are met, no matter how long the patients are in the hospital. Private insurers haven’t all followed suit yet, but they will. The CMS criteria that qualify a patient for inpatient reimbursement rates are a little vague (of course) but I went through them to create a list of justifications for inpatient status. I assembled the list into three categories:  patient factors (frail, elderly, fall risk, etc.), procedure factors (complicated procedure, high contrast use, thrombus in lesion, etc.) and presentation factors (acute coronary syndrome, electrical/hemodynamic instability, etc.). We built that list into our Philips cath report program. The list is reviewed when the staff and fellow do the preliminary report at the end of the case, and if a criterion is met, the patient is kept as an inpatient. Otherwise, everybody is outpatient, regardless of access site.

We dismiss outpatient PCI the same day, even at night, as long as they’ve had about 8 hours of observation (some patients prefer to go home at 10 pm rather than spend the night in the hospital — I would).  The rest stay overnight on outpatient status and are dismissed early the next day.
Kirk Garratt

Perclose and discharge

From Los Angeles, California: At UCLA, we keep patients overnight after PCI, and we have not had any bleeds in the patients who were otherwise stable. We looked into the feasibility of discharging patients the same day. The economic benefit depends on the insurance carrier. Since it is illegal to direct care based on the economic status or insurance of a patient, we found that it would not be financially beneficial to discharge on the same day of PCI all patients who met the criteria. The downside is obviously the potential legal risk of a patient bleeding or closing down an artery after they went home. I use a Perclose on all the patients and our hematoma rate has dropped dramatically. We also do radial PCI, but I’m not impressed that the complication rate with those is any less than from the femoral artery. So the final answer for us was, when it becomes economically advantageous, we are ready to initiate a program of same-day discharge.
Jonathan Tobis

Geography-related discharge timing

From Portland, Maine: We do mostly radial PCI now (65%), having switched over during this past year. I looked into this [early discharge]. There are several obstacles, not the least of which is the geography of our referral area. Many patients live 2-4 hours away; I don’t want someone on the road for 4 hours after a PCI. I am convinced that this is an area that is ready for change though. I’m impressed with how little complications there are with radial approach. Even the surgeons notice this; the patients don’t drop their hemoglobin much after a radial cath. Even the best femoral approach with a Perclosed groin can hide blood.

Then there are the insurance and legal issues, as others have discussed. If these weren’t an issue, I would favor sending patients that lived near Bangor home the same day after radial PCI, but keep femoral patients overnight. So, eventually, we will be sending radial PCI patients home.  We might send our radial PCI to a nearby hotel, which the hospital would pay for, rather than have the patient occupy an expensive bed.
Peter VerLee

Question: Should we keep the patient longer because of troponin (Tn) bump without frank myocardial infarction?

From New York City: We check troponins on those that stay overnight and will keep them longer if it’s significantly elevated. We don’t check those going home the same day.
Kirk N. Garratt

From Chicago, Illinois: We routinely measure Tn post PCI at 12 hours post procedure. Of course, it takes a measurement 12 hours later to find a small infarctlet, so if we rush the patient out, we won’t find these. If we try to cut corners and measure them at 8 hours and send the patient home, we will a) miss cases and b) have to deal with discharged patients with enzyme bumps. Does that matter? For those who do not routinely measuring them: See the NCDR paper from about two years ago that showed that those institutions which measure enzymes routinely post PCI are the high quality ones. This is referenced in the second part of the recent series of SCAI quality papers we wrote.

That said, the routine measurement of Tn after PCI presents us with a logical conundrum.  If you do not measure them, you will by definition, have fewer “complications,” since you won’t be finding the 18% of patients with enzyme leaks. Hence, there is a bias not to find them. Besides, as the argument goes, as long as the case went well, what will you do differently? Well, there is data that shows that those patients who do have enzyme bumps over 3x normal have a diminished prognosis. On the other hand, you are not going to treat the patient differently per se.

The reasons for troponin leaks are many and varied, related to both the procedure and the patient. The occurrence and the incidence of leaks in an operator’s experience is NOT a measure of quality. However, the centers that do collect this data are the ones who do the most cases and have the best mortality rates.

So the answer to your question is that measuring Tn enzymes routinely does give information about the patient and about operator quality, although not necessarily in the ways most centers interpret or act on them.
Lloyd W. Klein

From Chicago: Regarding troponins post PCI, we do not check Tn for our same-day PCI patients. If something occurs during the cases that would prompt us to want to check a Tn, then the patient is excluded from the same day discharge program. We currently only do same-day discharge for transradial PCI because we have had problems with groin bleeds during the night following the PCI (but there is literature to support transfemoral PCI discharge in selected patients). 
Adhir Shroff

My bottom line

At the Long Beach Veterans Administration Hospital, after an uncomplicated radial PCI, we discharge the patient 8 hours later. Our experience with radial cath as our access of first choice has shown that the bleeding complications have become non-existent and permits us, with high confidence, to discharge the patient after brief observation, hydration and time for conscious sedation to wear off. I do not have the confidence yet that all femoral PCI patients (even with closure devices) can go home early. 

At the University California Irvine cath lab, for both radial and femoral procedures, we use a 23-hour admission and discharge the patient in the morning; I think this is mostly because of habit and possibly my notion of reimbursement favoring 23 hours rather than outpatient PCI procedure. We will soon be moving to outpatient radial PCI early discharge protocols. Lastly, since we do not usually treat asymptomatic small Tn elevations, we are not routinely measuring Tn post uncomplicated PCI.

I wanted to add a final word about discharge or any medical care related to insurance or reimbursement. Dr. Chambers expressed his opinion, which I share, “[regarding reimbursement and hospital discharge]…do what is best for the patient and at the end of the day, it is easier to live with one’s self and it will ultimately be best for all, and you won’t have to remember why you justified what. To paraphrase Mark Twain, the more you tell the truth, the less you have to remember…

I hope this discussion moves your lab toward considering when it is best for the patient and the lab to discharge the uncomplicated PCI patient.


  1. Chambers CE, Dehmer GJ, Cox DA, Harrington RA, Babb JD, Popma JJ, Turco MA, Weiner BH, Tommaso CL; Society for Cardiovascular Angiography and Interventions. Defining the length of stay following percutaneous coronary intervention: an expert consensus document from the Society for Cardiovascular Angiography and Interventions. Endorsed by the American College of Cardiology Foundation. Catheter Cardiovasc Interv 2009 Jun 1;73(7):847–858.
  2. Carere RG, Webb JG, Buller CE, et al. Suture closure of femoral arterial puncture sites after coronary angioplasty followed by same-day discharge. Am Heart J 2000;139:52–58.
  3. Bertrand OF, De Larochellière R, Rodés-Cabau J, et al. A randomized study comparing same-day home discharge and abciximab bolus only to overnight hospitalization and abciximab bolus and infusion after transradial coronary stent implantation. Circulation 2006;114:2636–2643.
  4. Heyde GS, Koch KT, de Winter RJ, et al. Randomized trial comparing same-day discharge with overnight hospital stay after percutaneous coronary intervention: Results of the Elective PCI in Outpatient Study (EPOS). Circulation 2007;115:2299–2306.