Editor's note: Dr. Nasca is Chief Executive Officer of the Accreditation Council for Graduate Medical Education, the major accreditor of residency and fellowship training programs, and CEO of ACGME International, a position he assumed in 2007 after serving as Dean at Jefferson Medical College. He has been involved in medical education since 1981 and has authored more than 100 peer-reviewed articles, chapters, and other publications. We spoke with him about the ACGME's Clinical Learning Environment Review (CLER) program and its impact on medical education.

Dr. Robert M. Wachter: What led you to believe that a program like CLER was needed?

Dr. Thomas J. Nasca: We emerged from our discussions about resident duty hours understanding that the public was asking us to assure them of the safety and quality of patient care rendered in the teaching environment. Duty hours were an index issue in that debate, but the real issue was safety and quality. We gathered a group of individuals to try to understand how the ACGME could examine the issue, and what would be an appropriate way to do that without duplicating the effort of other agencies—many of which are centered around safety and quality of care rendered in all institutions and settings. We eventually centered around six key areas that we feel describe a quality learning environment, where the safety and quality of care rendered to patients today is assured, as well as a learning environment where the safety and quality of care rendered by our graduates in their future practice is assured. The third dimension is our ongoing attention and real concern about professionalism and professional values and their inculcation in the next generation of physicians, and for all who are involved in the care of patients in every clinical environment.

RW: One might have guessed that if you were going to look at safety and quality and try to improve it in the training environment, you would have somehow bolted it on to the existing accreditation process. It's a huge infrastructure, and you made an interesting design decision to have it become a separate program. Why did you do that?

TN: A number of luminaries in the field advised us that many groups had fairly significant leverage in the clinical care environment and were bringing that leverage to bear to work on safety and quality. We decided to take the tack of aeronautical reporting of near misses and errors with a no-harm model. This was the model in the VA under Jim Bagian's leadership, and it resulted in significant engagement of individuals throughout the clinical care environment in safety and quality. Rather than adding a series of rules around safety and quality to our program or institutional requirements, we chose to provide six major areas of focus and then reflect back to institutions based on findings from an external peer review process. Our goal was to provide formative feedback for them to improve.

Then we followed it with the development of a trajectory of a highly reliable and safe clinical care environment called The Pathways to Clinical Excellence, and that gives people a general roadmap. Another principle is that each clinical learning environment is fundamentally different. Each teaching institution serves a unique patient population. They function in a unique clinical context, under different reimbursement systems, and under different regulatory parameters. Therefore, we chose to provide a peer review reflection back to the training environment so that they could continue to improve with a roadmap that we provided to them in the six areas.

RW: I imagine there's some tension in creating something that's not quite as prescriptive and whose goal is to promote improvement; whereas, as an accrediting organization, you have a responsibility to make sure that people and programs meet certain standards. How do you balance that with CLER, particularly if you find a program that is not providing a safe and quality environment for trainees and for patients?

TN: There are two dimensions. The first is that going beyond minimum standards is the entire philosophy behind the Next Accreditation System, of which CLER is one of the pillars. Our goal is to hold programs and institutions to minimum standards, but to stimulate them to go far beyond minimum standards to endeavor to achieve excellence at a program or institutional level. That's the basic framework. The tension around accepting that framework and understanding that is that we develop a level of trust within the community—trust that that is indeed our motivation. We've been working very hard over the last 4 years to do that, and I think we are making progress. You'd have to speak to others in the field to determine whether we've been successful.

RW: That certainly is my sense as well. When it comes out of an accrediting organization, people always believe that there's going to be a stick there somewhere, and a pass/fail, but the feedback I've gotten so far is quite positive about the visits. Is that what you've been hearing?

TN: Yes, we've had very few organizations that have not enthusiastically embraced the findings as an opportunity to continue to improve. It has been the rare organization that has disagreed with our findings, less than 1%.

RW: Do you see a trajectory in which, over the course of years or a decade, you do identify certain best practices and maybe certain minimum criteria that everybody should meet, and then a part of this does become more prescriptive and embedded in your accreditation standards?

TN: I do see that possibility emerging. What we have said is that after each round of accreditation, after we visit each teaching institution in the country (it takes between 2 and 2.5 years to go through these extensive visits), we will publish findings. From that, the community will understand where we are. We'll have in essence a report card of this first round where we start with a goal of seeing improvement, being able to demonstrate to the American public improvement on round two, and then subsequently on round three some findings finding their way into frameworks for accreditation. But the CLER visits will remain completely separate from the accreditation visit programs.

Finally, this may be a dream but it is my hope that within 10 or 12 years, we'll find that there is no more need for CLER, or it may evolve into solving other challenges that we face in the learning environment. It's my hope that safety and quality will not be the issue at that juncture. We all have to be realists and examine the data as we go, but that would be a vision for success.

RW: When you think about who is accountable for creating the environment, is it the training program directors? Is it the institutional leaders of the training programs? Or is it the folks who organize and run the delivery system?

TN: I believe it's all of the above. The CLER program is designed to nurture and bring together the clinical leadership at the level of the chief executive officer, the chief medical officer, the chief nursing officer, the chief of quality and safety in the institution, and the leadership of graduate medical education. The first and last meetings of the CLER team are with that group. We go so far as to not conduct a CLER visit if the chief executive officer cannot commit to both of those meetings. Our goal is to bring back into the fiber of the clinical delivery system the graduate medical education programs, which over the last couple of decades have been separated for many reasons, some of which ACGME created to enhance the quality of the medical education components of the educational programs. That artificial separation from the clinical delivery system of these key providers of clinical care is not optimal when it comes to creating an ideal safety and quality environment.

RW: Do you feel like these groups have already begun to come together because of the other pressures in the environment, or CLER is actually making that happen?

TN: Early on, it was distinctly unusual to find an organization where this relationship existed. We did see some exemplar institutions early on who volunteered for CLER where those relationships did exist. While they were usually collegial relationships among the individuals, the relationships at a functional level did not exist. One challenge we find when we start to ring the bell about opportunities for improvement is that people look to us for solutions. One solution that we've piloted and will probably expand over the next 2 years is an educational function called CLER Conversations, where we actually bring teams of medical staff, nursing staff, and educational leaders to Chicago to talk about planning for improved integration of graduate medical education programs with quality and safety programs within the institution. That program was piloted in the fall in Chicago, and it worked extremely well with the first groups. So we believe we're going to expand that.

We're engaged with a number of other organizations because CLER had stimulated the need to pay attention to the clinical learning environment across professions and across levels of preparation. A coalition has formed to work together voluntarily, called NCICLE, the National Collaborative for Improving the Clinical Learning Environment, to improve the clinical learning environment. And that voluntary group includes more than 30 organizations now, including the Institute for Healthcare Improvement, the Association of American Medical Colleges, the American Medical Association, a number of nursing groups, educators and accreditors in pharmacy, and several patient safety groups that traditionally ACGME would not necessarily have the opportunity to work with. So we're very excited about this development because we will try to provide solutions that people can try out in their institutions to find what works best.

RW: One of the tensions marbled through this question is the autonomy-versus-oversight balance. One could argue that in the safest environment, you would be taking away a tremendous amount of autonomy from the trainees, and you might actually make it safer today, but over a 10- or 20-year time horizon you make it less safe because trainees come out and they're not ready to be autonomous. How do you think about that tension?

TN: Well, we're currently cataloging both circumstances of undersupervision and oversupervision. Our tripartite goal has two specific goals that address the question you asked. The first is the safety and quality of care rendered to patients in a teaching environment today. And that demands the appropriate level of supervision to be provided for the level of expertise of the trainee. That is often misinterpreted to mean excessive supervision of every trainee no matter what the level of training. And that's where we run into the oversupervision circumstance. We are very concerned that trainees at the senior levels need to have appropriate backup with supervising faculty available, were the resident to encounter a challenge in delivering care to an individual patient. But we need to recognize that a trainee in their last few months of residency or fellowship training has the skill set necessary to practice independently, and indeed on July 1 will be practicing with no supervision. So the dual responsibility of the training environment is to create that graded authority and responsibility and have appropriate supervision systems that not only protect the patient today, but also provide the trainee with the latitude and the space to demonstrate that they are indeed ready to enter the unsupervised practice of medicine right after graduation.

RW: I imagine programs are surprised when they get feedback about oversupervising. When most people think about quality and safety, their instinctive response is to put in more layers of oversight.

TN: Yes, and it's really not more layers of oversight. It's appropriate oversight available to the trainee based on their level of demonstrated competence. The logical question then is: what are you doing around assessment of competence? The Milestone Program is exactly that—to provide faculty with more clear understanding of the characteristics that the trainee must demonstrate in order to be entrusted to move to the next level of supervision.

RW: We're now more than a decade into duty hours and about 5 years since the last changes. What is your sense of how it's going? And what's your reaction to the literature, which has not demonstrated improvements in quality and safety, at least in terms of the way people are trying to measure it?

TN: My personal view of this now 30-year-old issue beginning back in the 1980s with the unfortunate circumstances surrounding Libby Zion is that we have seen significant benefit on the basis of some standards. I believe that the workload of 80 hours per week averaged over 4 weeks is a very reasonable educational standard, with appropriate variability based on acute needs of individual patients. I believe that night call should be no more frequent than every third night. I believe it's reasonable to average 1 day off in 7 for rejuvenation and rest, as well as for residents to have the time to study, prepare, and accomplish routine activities of daily life and family life in this rigorous training circumstance. So I believe that we have had some improvement both in the level of alertness, the ability to internalize the educational information that's provided during our clinical experiences, and the didactic experiences that they had. So those rules appear to be reasonable. There was some evidence of improvement in certain circumstances with implementation of those rules, especially in the critical care setting.

The question then becomes: do the physiologic studies translate into a complex clinical care and learning environment, where confounding variables often result in outcomes that were not intended or expected. We need to understand the net impact of our standards in the delivery system environment. The medical community in general and ACGME in particular was criticized by the IOM in their report because we did not conduct, nor cause to be conducted, research to further understand the impact of standards. We have taken that to heart, and two investigator-initiated trials are now underway, one in internal medicine and one in general surgery, to try to understand whether parsing the rules beyond those core three or four basic ones has resulted in improvement in patient care or patient safety, or perhaps has been harmful either to patient care, patient safety, or resident education.

ACGME funded a portion of both studies, but have been completely independent of the design and conduct of the studies. We want these large multicenter trials to not be tainted by any question about ACGME's engagement in them beyond the funding. What we have committed to and actually accomplished is waivers from certain standards in programs that are randomized into the experimental arms. And we've agreed to share, in a deidentified fashion, key elements of our information that we gather in order to understand educational and other impacts of the experimental design and the standard design, which are in compliance with our standards. We hope to have data from both of those studies within about 2 years; at which time, unless something intervenes in the interim, we would use that information in review and revision of our standards. [Editor's note: Since this interview was conducted, one of the studies, in surgery, was published in February 2016. It found no significant differences in patient outcomes, resident well-being, and education between programs adhering to the usual duty hours limits and programs that were given additional flexibility.]

RW: What have been the biggest surprises of what you've seen as you and your staff have gone out into the field and looked hard at the learning environment?

TN: The positive surprises: the first is the welcome with which we have been received at every institution. Every institution is very interested in trying to be better. It is gratifying to see at the administrative level, at the medical and nursing leadership level, the patient safety level, and the GME leadership level, as well as the residents, nurses, and other health providers of our institutions, that they all aspire to provide better and higher quality and safer care.

Second, we had been gratified by the interest of other members of the clinical care team and the administrative team in helping engage residents in safety and quality. They recognize that it's a challenge because the educational structure often sits in parallel to the administrative structure of the institution. But as the teams have gone around on the floors and engaged nursing and respiratory therapy and the dieticians, they all have recognized not only the importance of residents becoming expert in these dimensions. They have also recognized the importance of their input in achieving the clinical excellence in quality and safety goals of the institution at a local level.

Some things that have surprised us have been the degree of separation of the graduate education programs from the clinical care environment, even in environments that one would expect were traditionally, classically "academic" and would have been expected to have extensive integration already. The third is that only when every physician and every caregiver believes that it is part of their professional responsibility on a minute-to-minute basis to look for and seek out unsafe conditions and help resolve those will we get to the issues.

All too often we are seeing environments where safety and quality is someone else's responsibility. Residents have lectures about it. They may have quality improvement projects in the clinic. But they're really not integrated into the institutional effort in safety and quality. And that has two downsides. First, they're a huge workforce that is not engaged. Resident physicians and fellows are one out of every seven doctors in the United States. That's one-seventh of the workforce that is not engaged systematically in safety and quality.

The other key element is that we will not produce physicians in practice who actually are experienced in and understand how to work within the institutional systems of safety and quality unless we expose them to that during residency training. That means programs need to be configured so that residents have the opportunity to do that. Not just because they're self-motivated to do that, but because it's systematically built into the educational program. We do not frequently see systematic incorporation of resident engagement into safety and quality beyond what traditionally exists, for instance Morbidity and Mortality Review and the like—conference-based quality discussions around clinical care. For example, we don't see active engagement in root cause analysis by multidisciplinary teams. All of it's there, but it's spotty and it's largely based on the interest of the trainee to go beyond what is part of the curriculum. So on the GME side, those issues need to be dealt with. On the institutional side, the systems of quality and safety have not been built to incorporate resident input. So whether it's reporting unsafe conditions or systematic provision of the opportunity for residents to be engaged in safety and quality, all of those areas are ripe for development over the next cycle. That's where we hope to see the benefit.

[Editor's note: Since the time of the interview, the ACGME has announced the Pursuing Excellence in Clinical Learning Environments initiative, a competitive grant program designed to promote transformative improvement in the clinical learning environments of ACGME-accredited institutions.]