Editor's note: Dr. Rosen is Medical Director of the Inpatient Specialty Program (ISP) Hospitalist service at Cedars-Sinai Medical Center. He also assists with the operation of Cedars-Sinai's innovative Procedure Center, which provides numerous procedural services for inpatients and outpatients. This Center and Dr. Rosen's work have been featured in articles in the Wall Street Journal and elsewhere. In this conversation, we explore the value of a dedicated procedure center and the emergence of specialized physicians to staff them ("proceduralists") and the challenges inherent in teaching novices how to perform risky procedures without harming patient safety.

Whose idea was the procedure service, and what forces motivated you to think about doing this?

Dr. Bradley T. Rosen: The idea behind Procedure Center came in 1990, and it preceded any grand plan to create a new field. The driving force behind it was Mark Ault, MD, our division director in general internal medicine. There was a need, primarily on the faculty teaching service, to not only provide more teaching and guidance for the residents in doing procedures, but also to get some procedures (for example, PICC [peripherally inserted central catheter] lines) done in a more timely fashion. What started with him and a colleague doing these catheters, a couple of cases a week with the help of a phlebotomy nurse, slowly grew into a service. Philip Ng, MD, medical director of the Procedure Center, was recruited in 1994 to accommodate the growing volume of procedures. For any procedure, by virtue of doing more of them, you get better, learn from your experiences, and translate that back to the bedside. You end up providing better patient care and having lower complication rates. Over time, this was not just a service provided to teaching physicians, but private physicians who also practice at Cedars started to request these physicians to do central lines, PICCs, and dialysis catheters. Years later, we began doing some of the other medical procedures such as paracentesis, thoracentesis, lumbar punctures, and then ICU [intensive care unit] procedures such as percutaneous tracheostomies. We've slowly added to our "menu" over the years based on the needs of the referring physicians and the patients in the hospital.

RW: Some places that have read about your work are going to try to begin a service that looks like yours. What would you see as the main obstacles for them to overcome—financial, political, logistical—in developing a service like this?

BR: They will have to face all of those: political, financial, and technical, as well as learning the procedures. The biggest challenge that whoever is looking to start this needs to consider is where is the greatest need at their institution for a procedure service? In other words, are there specific procedures that nobody is taking ownership over, and as a result there are delays in the procedure being performed? For example, PICC lines are relatively low reimbursing and relatively simple to do, so interventional radiologists aren't excited about putting them at the top of their list, and not all facilities have PICC nurses to be able to do these procedures. So those lines sometimes wait until the end of the day and get pushed off to the next day. But obviously those can make a big impact on length of stay and quality of care.

I don't think that it makes sense to try to replicate in total what has evolved organically at Cedars-Sinai over the past 17 years. It has to develop slowly at any institution. The first thing is to look at what the need is. Second, people do need to at least look at the numbers, because at the front end there is an investment in equipment, such as a portable ultrasound. Then there's staffing, supplies, and scheduling infrastructure, and making sure that people know how to reach you. Making the case that it's worth it to do such a thing takes a little bit of enlightened thinking on the part of hospital administration. The service won't pay for itself simply by the physician billing for their services. So there does have to be a more comprehensive financial analysis in terms of professional revenue as well as the facility fees that can be generated for the services, and in addition to that any cost savings from length-of-stay reductions from more timely procedures. Lastly, any reduction in complication rates obviously translates into cost savings for the institution and better patient safety as well.

RW: What data do you have that demonstrate that in fact this improves patient safety?

BR: We have been collecting data since day one. That's an important point to make to anybody who's interested in starting such a service, that collecting and maintaining data are of paramount importance in order to demonstrate value. We have analyzed all data for the last 15 years, and our major and minor complication rate for all of the procedures collectively is less than 1%—we do not have a comparison group here at Cedars. It's not as if the Procedure Center started providing the service to half the hospital and left the other half to the masses to do the procedures so that we could demonstrate our effectiveness. But when you look at the scant data in the literature in terms of complication rates for the various medical procedures, whether it is central lines, thoracenteses, paracenteses, etc., the rates range anywhere between 2% and 5%. Comparing that to our data, we've demonstrated a lower complication rate.

RW: Is the feeling of the organization that the evidence of safety benefit is so self-evident and impressive from your data that that's good enough? Or for a new program starting up, should they be trying to figure out ways of collecting more comparative data to make the case for patient safety?

BR: To the extent that apples to apples comparisons can be performed, certainly any new service should try to do that. If an institution has a baseline complication rate for procedures and then a group wants to start a procedure service, they can and should certainly track their performance and compare it to the baseline. From our standpoint at Cedars, I think the jury is in—that the institution supports our Procedure Center and feels that there are a lot of benefits to its existence. At our institution at this point, the commitment goes beyond complication rates in the strict sense.

RW: How do you organize the center in terms of outpatient vs. inpatient procedures? It sounds like you also go to the ICU and various places around the building.

BR: Now this is also something that has evolved over time. There are a number of different components to this. We have a Procedure Center, which I refer to as the neural hub of all the procedural activities. At the front end there's a scheduler, someone who takes the calls, pulls new orders from the computer system, and puts them onto a schedule so we can stay somewhat organized. Somewhere between 60% and 75% of our work is on the inpatient side, which was the impetus behind the service. The outpatients have really just evolved as a bonus. And as part of that, as an aside, we do have an infusion center where we transfuse blood and give various immunomodulating drugs for rheumatologic diseases and IBD [inflammatory bowel disease], etc. Obviously, the lines that we've put in for those conditions then get medications infused through them, so we're kind of a natural extension of that center. But in terms of the inpatient services, we have patients come to our procedure center. We have a fluoroscopy suite with a C arm set up for more complicated procedures or for tunneled catheters or for PAS-ports [peripherally accessed subcutaneous ports]. We also have mobile carts and portable ultrasounds that we can take to the ICUs, isolation rooms, and patients who are not mobile and do procedures at the bedside. So we provide a whole spectrum of procedural services based on where the patients are located, how hard it is to move them, and what kind of procedure we're going to do.

RW: Who are the physicians doing the procedures?

BR: The proceduralists, as we call them, are generally physicians trained in internal medicine, plus or minus critical care, who enjoy doing procedures and enjoy working with their hands, separate and apart from all of the cognitive work they did as internists or intensivists. We also have a team of nurses dedicated to the Procedure Center. So a nurse–physician team will be doing procedures all day together, and the nurse basically functions as a circulating nurse. On the front end, the nurse helps to obtain consent and do any other necessary paperwork, helps set up the sterile field, and helps perform the Universal Protocol or time out. The nurse also helps to clean up afterward, helps collect the data, and helps to move the cart and the portable ultrasound around. In addition to that, the Procedure Center has nurses who regularly go out and help to maintain the lines. It's important for anyone considering starting up a procedure service to keep in mind that a procedure center or a procedure service is not just physicians running around doing procedures. After a central line is placed, whether it is a PICC, a hemodialysis catheter, or a central line, those lines require care and feeding. That means making sure that the dressings stay clean, dry, and occlusive to prevent a catheter-related bloodstream infection (CR-BSI) and making sure that the lines are flushed properly so they don't occlude. The nurses on the floor must be educated how to manage those lines so that there's not too much IV tubing (weight) hanging off of them that can result in a line dislodge. There is a lot that goes into making sure that once the line is placed, the line is well cared for so that it can last as long as the patient needs it. So we have rounding nurses who go out and do that kind of patient and nursing education, line maintenance, and declotting of the lines.

RW: Who removes the lines?

BR: Generally the Procedure Center—if it's a central line or it's a tunnel dialysis catheter, the physician will remove the line. If it's a PICC line, one of our nurses will remove the line. We discourage floor nurses or any physicians who aren't involved in the placement of the line from removing it. Not necessarily from a patient safety standpoint, but we have found that the decision of whether a line "should" be removed is often a matter of clinical judgment. Sometimes people are too quick to remove central lines if a line is in place and there is a concern about a low-grade fever, or one out of four blood cultures is positive at 48 hours. A lot of the patients that we manage in the Procedure Center have had 5, 10, 12, 15 central lines already. It may take a couple of hours to get another PICC in a patient, and we don't like finding out on a Monday morning that, well, the patient had a low-grade fever so the covering physician pulled it out. Because that may mean that that patient can't get another central line. Over time, we have a growing list of those very difficult patients. So we like to be involved in the decision whether or not the line is actually removed.

RW: If a hospitalist trained as an internist joins the staff at Cedars and decides to become a proceduralist, what happens to make that physician become competent in this? Does it become a full-time job or can it be mixed with other more traditional activities?

BR: There is no strict script that has to be followed in order to do procedural activities or to become a proceduralist. There certainly is some additional training needed on the front end. Traditionally, we have found that, certainly what my experience personally was, it takes somewhere between 3 to 6 months to learn and become facile with all of the skills required to be able to fly solo as a proceduralist, at least at Cedars. And then, even more importantly, is knowing how to respond when a procedure doesn't go well. In other words, the analogy that I give is that anybody can strap on a pair of roller blades and roll down the street or down a hill. That's easy. But you have to know how to stop, and that's what takes the practice. So in terms of procedures, you have to know what to do when something doesn't go right, how to adapt, and what other maneuvers or adjustments you can do so that the procedure ends up going smoothly. Generally by the time that the 3- to 6-month time frame is up, assuming you've been doing this more or less full time, you should be pretty facile in the vast majority of procedures that we do.

In terms of how frequently a proceduralist or someone interested in procedural work does the procedures, I think there's a fine balance. To be efficient and effective as a proceduralist, the person does have to do enough procedures frequently enough to keep the skills honed and at the same time balancing that with not letting the person burn out. What we have found is on average, I would say about 75% of the time, a proceduralist here at Cedars is doing procedures. It seems to be a good balance. The remainder of that time is rounding in the ICU or rounding on the general internal medicine wards, either academic or nonteaching, and then mixing into that some research time, some vacation time, etc. This is just my opinion, but I think if you drop down below 50% of your time doing procedures, you're going to have to ramp yourself back up, and I think that you start to lose a little bit of your edge.

RW: Talk now about the integration of trainees, and not only practically what happens for a medical student or a resident, but philosophically what your service means to the "see one, do one, teach one" model?

BR: On the teaching front, one of the additional value-added services that we provide as a dedicated Procedure Center is that each year as part of their orientation, the incoming interns go through a half-day orientation on how to perform safe procedures with ultrasound guidance. We run hands-on training sessions during their orientation week that give them all of the basics. We don't expect them to remember or be able to do a procedure on their own after that. But we expose them to the concept of wide sterile barriers, get them familiar with the ultrasound, and get them to practice on our tissue simulators so that they at least get familiar with the concepts. Then we talk about some of the indications and contraindications, and some of the regulatory forces at work. So, on the front end, we do some training for all of the interns coming in through the Cedars-Sinai Medical Center Internal Medicine Residency program. We also create opportunities for medical students through UCLA as third years, or any rotating resident in the fourth year as a sub intern (a fourth-year student), and our interns and residents can do a 1-month rotation through our Procedure Center where they're paired up with one of the proceduralists doing whatever procedures are happening at that time. So there are opportunities on that end. There's a combination of some information didactics as well as mostly hands-on training.

In terms of what this means philosophically to the "see one, do one, teach one" model and whether we should have residents doing procedures at all, that has been a topic of great discussion recently. The American Board of Internal Medicine recently modified their requirements for residents from a strict number to a more qualitative understanding of indications and contraindications and generally how to do the procedures. My understanding of that is based on a recognition that, first, regardless of whether you do 2, 5, 10, or 15, you may or may not really know what you're doing; and second, you may never need the skill set again, so why subject patients to your learning if you're not going to end up applying that? I don't have any wisdom on this topic beyond what many people have already published, discussed, or written about. I think that there is something to be said for having a dedicated service that specializes in doing procedures to ensure that patients are getting high quality care—that we don't have trainees learning with sharp objects on real patients. Beyond that, I'm not sure whether or not residents should be expected to or should be required to do procedures. I think that they still are the most prevalent and available work force we have in many hospitals, and in the middle of the night they are the physicians taking care of patients. And if a procedure needs to be done, they need to be able to do it.

RW: At your institution, I'm assuming that there is not a procedural service at 2:00 in the morning?

BR: That's correct. We have at our institution, as with many academic medical centers, a residency training program with a fellowship track in critical care, so there are layers of support in place. But just like at many institutions at 2:00 in the morning in the ICU, if a patient is crashing and needs a line for pressors or blood products, it's going to be the intern and the resident, whether or not the fellow is available, to put in that central line to make sure the patient is getting appropriate care.

RW: In that circumstance, has the service made things potentially even slightly less safe because the trainees may be less practiced?

BR: That's an excellent question. There is no evidence that patients are getting less safe care as a result of the existence of our Procedure Center. Residents and interns still are getting opportunities to do procedures, especially in the ICUs. In fact, our Procedure Center generally expects that the intern, resident and/or fellow, and/or attending will have attempted a procedure or deemed it too dangerous for them to try before they call us. So we do try to encourage the learning to take place under appropriate guidance and supervision before we're called to "take over the case." At Cedars, the Procedure Center has plenty of work to do on the regular medicine/surgical floors from patients referred to us by either the teaching service or more commonly the private doctors, that we certainly don't need the business in the ICU. But there are patients who are wildly coagulopathic that the team just doesn't want to stick a needle in, or the vessels are just too tiny or thrombosed that they need assistance with. So there's certainly a role for us. With those cases, if the interns or residents are around, we try to get them involved in the case, either scrubbed or at least observing so that they still have the opportunity to learn. It's a very important issue, which I think touches on your original question of should residents be doing procedures or not? In an ideal world I think that any time a patient receives a procedure, it should be done by someone who has as much experience as possible. In the real world I think that we have to find ways to maximize patient safety while taking into account the realities of staffing and patient demands.