Service Fanatics

by James Merlino

  Putting the entire blame on physicians is unfair. Some of the problem likely rests with patients’ interpretations of our interactions, influenced by memory or affected by high anxiety, drugs, or a medical condition. Patients’ reluctance to ask questions or desire not to challenge the physician can also affect how much they understand.

  I tested this theory anecdotally by sending a medical student into 20 patients’ rooms 10 minutes after my visit. The student asked the patient, “Do you remember the plan of care that Dr. Merlino discussed with you?” Fewer than half of the 20 patients could recite the plan for the day. Patients have difficulty with simple things such as remembering physicians’ names. David L. Longworth is chair of the Medicine Institute and associate chief of staff for professional staff affairs at Cleveland Clinic. As part of his responsibilities, he precepts medical residents in the ambulatory clinic. Longworth likes to demonstrate to the doctors in training that patients have difficulty remembering and that physicians must communicate clearly. When he walks into a patient room with a resident, Longworth introduces himself, providing his complete name. At the visit’s conclusion, he asks whether the patient remembers his name, and frequently fewer than half do.3

  For Doctors, the Patient Experience Is About Communication Skills

  When I assumed the CXO role in July 2009, our main campus physician communication domain was at the 14th percentile of all U.S. hospitals, among the poorest of all our HCAHPS measures. The performance of private practice physicians in our community hospitals was even worse. Improving physician communication was imperative, and there was nowhere to go but up.

  As I mentioned earlier, one of our first tactics was to show our scores to groups of physicians to familiarize them with the measurement process and data. We shared hospital-level performance and talked about the survey process, explaining the way questions were asked, how patients were allowed to respond, and the methodology that Medicare used to score surveys and distribute results. We talked to doctors in departmental, staff, and leadership meetings and held community dinners with private practice groups. We learned that our doctors were essentially clueless regarding their performance metrics. Few knew what the HCAHPS survey was, let alone that it contained questions evaluating how physicians communicated with hospital patients. Before I interviewed for the CXO position, I likewise had no idea physician communication skills were rated by patients.

  It was essential to get the data to the physicians. When Cosgrove led the Department of Thoracic and Cardiovascular Surgery before becoming CEO, he was tasked with consolidating Cleveland Clinic’s open-heart surgery programs across northeast Ohio. Getting a group of heart surgeons to work together and standardize their practices was certainly no easy task. One of the tactics he employed was data transparency. He released unblinded individual surgeon and program performance data throughout the department so that everyone could see everybody else’s data. He perceived that physicians, especially data-driven and inherently competitive heart surgeons, would use the individualized numbers as an improvement tool, and they did.4

  Cosgrove perceived that getting data to physicians would shorten their learning curve on the HCAHPS rating process and drive communication performance improvement, so we decided to release individual scores to every physician in our group practice. We chose to convert all of the HCAHPS scores to percentile rankings so that physicians would understand how they stood relative to peers across the country. In addition, the Medicare Hospital Value-Based Purchasing (HBVP) Program uses percentiles to judge performance, so this would provide consistency.

  I suspected that our release of data would be rather controversial and that our physicians would not necessarily like what was coming. Cosgrove wanted us to release unblinded scores immediately. “Post them in the lounges!” I recall him saying. However, I counseled that we proceed more gradually. Despite the educational meetings, most physicians still had little grasp of what HCAHPS was, let alone that they were individually scored on patient communication. Cosgrove humored me, and we released scores to individual doctors with their partners’ scores blinded, which coincided with a major campaign to educate physicians about the measurement process.

  Distributing individual data to doctors also taught us a great deal about the process we use to measure communication skills. We wanted to better understand what physicians thought about the HCAHPS survey process, given our reliance on it for our primary patient experience data. As we anticipated, after scores went out, doctors started paying much closer attention to the survey process and data. Physicians are highly trained in data interpretation and very skilled at using data to drive the way they practice. Complaints and criticism of the HCAHPS process started to pour in via phone and e-mail. The most common were:

  1. I can’t get the data.

  2. There is no comparative data.

  3. The standard is too high.

  4. The sample size is small.

  5. I don’t have the support I need.

  6. Other people impact the score.

  7. My scores are low because I work at high volume.

  8. No one will help me get better.

  Part of me wanted to say, “Look, the data is the data, and we can’t change the process. It’s the hand we’re dealt by the government. There’s really nothing we can do, so don’t shoot the messenger.” But we prepared responses for each of the top complaints because it was very important to address the physicians’ issues. We needed buy-in, and you don’t get that by ignoring people’s concerns.

  The first two complaints essentially became moot, as we now were distributing data. We chose the 90th percentile as the standard because it’s Medicare’s benchmark for comparison against all U.S. hospitals and physicians. The 90th percentile also represents an “A” grade, and this is where we should be as an organization.

  The sample size being too small is absolutely a valid criticism. Even the number of surveys Medicare requires from a hospital is not statistically valid. We advise physicians and their leaders to look at trends in the numbers and avoid seeing them as a snapshot. If a physician is in the 10th percentile in a single quarter, the communication scores are possibly invalid. But if the physician is in the 10th percentile over three or four quarters, it’s probable that the physician needs to improve his or her communication skills.

  The comment that “others impact my score” is an important one. The HCAHPS communication domain is linked to the discharging physician. If there is only one doctor who takes care of the patient while hospitalized, then it’s easy to assign accountability. But hospital patients typically are cared for by many physicians. On my service, even if I see my patients daily until discharge, there are still interns, residents, and fellows participating in care. A patient with complicated medical issues may have many staff physicians providing care. So others do impact a physician’s score. But achieving a high score on this metric requires teamwork, not just individual performance. I tell physicians the scoring process is determined by the Centers for Medicare & Medicaid Services (CMS), something we can’t change. But determining who helps us take care of patients is our choice, so we have an obligation to monitor our consulting physicians and house staff. If a consultant or resident is not communicating well with patients, that negative interaction will likely be reflected in the responsible physician’s HCAHPS scores, so perhaps we should choose different consultants.

  To prove this point when educating physicians about the process, I unblinded one-quarter of my individual HCAHPS data from 31 inpatients. Seven belonged to my partners, but I either had admitted them while on call or had briefly covered for a partner traveling out of town. For the total 31 patients, I ranked in the 50th percentile for communication performance. If I removed the 7 patients that were not mine, my score jumped to the 99th percentile. I did this not to prove I’m a great communicator but to illuminate reality on the units. Taking care of in-hospital patients is a team sport, and we have a responsibility to work together and police the team.
/>
  Many physicians complained they had low communication scores because they took care of a high volume of patients. They legitimately argued that there is a trade-off between productivity and good communication. Our main campus hospital is full of focused, high-volume specialists. Scholarship confirms that high-volume proceduralists tend to have better-quality outcomes.

  We probed this argument by closely examining the performance data of a relatively homogeneous practice group. Cleveland Clinic has one of the largest and highest-volume cardiac surgery programs in the world; we are the highest-volume U.S. provider, with the next competitor at just half of our volume. Each of our cardiac surgeons is very productive in relative-value-unit (RVU)5 performance. Essentially, they are a group of about 10 surgeons who are top in their field, see a comparable profile of patients, have extremely high quality standards, and have similar office and clinical support. Most of these surgeons had excellent communication scores, but a couple did not. While this was a small analysis, it clarified that highly productive physicians can also have excellent communication skills.

  The cardiac surgeons and their leadership were early adopters of patient experience initiatives. Bruce Lytle, chair of the Sydell and Arnold Miller Family Heart & Vascular Institute, frequently brings a patient’s family into his office after an operation and talks until the family is comfortable and has no more questions. He invited me to speak at a Cardiovascular Surgery Department meeting, and I presented a slide showing blinded physician communication scores. All except two were at or near the 90th percentile. At the end of my talk, Lytle remarked, “Everyone here knows who has the low scores and why. That ends today!” The two surgeons did not regularly round on their patients, and Lytle rightly believed this was reflected in their scores. Subsequently, the two started rounding routinely, and their communication scores went up. This is a great example of physician leadership at the local level and demonstrates how a simple tactic—rounding on patients—can have meaningful impact on patient perceptions of physician communication skills. It also reinforced that institute and department chairs could drive significant improvement in the communication scores by taking responsibility.

  The final complaint expressed by physicians about HCAHPS scoring data, the lack of available improvement assistance, was especially important feedback. Since we started discussing the scores, physicians had asked for tactics they could use to improve. While we gathered some tactics previously analyzed in physician communication scholarship, most improvement recommendations came from asking our own high-scoring physicians for their success secrets. Some of these suggestions are common sense and used by many physicians. But consistently using all of them during patient interactions is critical. They include:6

  1. Introduce yourself. Tell the patient and family who is in charge of their care.

  2. Address the patient by “sir” or “ma’am” and use the patient’s name if you know it.

  3. Partner with nursing on rounds and to discuss plans of care.

  4. Ensure that the patient and family understand the care plan.

  5. Set and manage patient expectations.

  6. Answer patient questions.

  7. Engage others who may impact patient perceptions, such as extenders and residents.

  8. Respect patient privacy.

  9. Recognize that patients judge you by how you look and what you say.

  10. When possible, include the patient’s family in discussions.

  11. Ask patients and visitors how they are being treated and whether they need anything.

  12. Discuss pain management.

  Our team also produced a comprehensive communication guide, designed by our doctors for our doctors. We balanced information from different practice environments and specialties, with more than 50 private and group practice physicians contributing. An easy-to-use checklist with simple suggestions accompanied extensive material for more in-depth study. Providing information from busy clinicians and keeping it practical and useful gave the guide credibility with frontline practitioners.

  Communication Skills Must Be Developed

  But releasing HCAHPS data, educating physicians about measurement, and distributing a communication guide were not enough. Physicians are important engines of our organization, critical assets that require continual investment. We had an obligation to help them communicate better. We needed a new program to help improve their skills. And teaching established physicians to communicate better with patients would not be easy.

  Over the course of my first year as CXO, chief of staff Joseph Hahn and I had numerous discussions on how to approach this training. Substantial commitment and resources would be required for success. Doctors would have to buy in and perceive it as worthwhile, but there would be pushback and criticism. The program would directly impact the sacred doctor-patient relationship. I suspected it would be one of the hardest things we would do in the patient experience.

  Hahn and I finally established a few ground rules. The effort would be led by respected frontline physicians. It could not be physicians in leadership or those at career start or end needing something to do. The planning group also needed to encompass believers and those not yet fully convinced. A healthy dose of realism and skepticism would ensure a much more robust and successful program.

  It needed to be practical and interactive to effectively serve and appeal to our high-performing, world-renowned medical staff. The training had to be highly relevant to busy, frontline clinicians, improving not only how they communicated with patients but how they practiced medicine. The training had to be as much about improving physicians’ effectiveness, efficiency, and expertise as about enhancing patient perceptions of physician interactions and communication. While we had little idea what the final program would look like, we knew that lecturing to a bunch of experienced doctors in a classroom about improving their communication skills would not work.

  We took an important lesson from the nursing education world. Non-nurse experts can teach nurses about any topic except nursing practice, where they have zero credibility. Only nurses can teach nurses about practice standards. Why should it be any different with physicians? The program had to be led and taught by active staff with peer and clinical credibility. Even though there are programs in hospitals across the United States using different professions (nurses, social workers, and so on) to help physicians improve communication skills, we felt we should use only physicians.

  I don’t believe I could be successfully coached to improve my patient communication skills by a nonphysician peer. He or she might have different perspectives and expertise than I do, yet what I most value is someone who has actually shared the experience of caring for patients, being ultimately responsible for them, and owning the patient experience medically and emotionally, as I have tried to do as a staff physician. We are a physician-led organization, and our leaders must model relationship-centered skills, as all other staff members are watching our behavior. Our team also was sensitive about avoiding a power differential between physicians and other providers in the classes.

  We wanted to start the training with our staff physicians for these reasons, and we found our leader in Adrienne Boissy, a neurologist and neuroethicist serving in Cleveland Clinic’s Mellen Center for Multiple Sclerosis. As patient experience leader for the Neurological Institute, she had had several highly successful projects, including ones to improve staff and resident communication skills. She had a passion for patient engagement and communication and was a well-respected physician within Cleveland Clinic. In the summer of 2010, Hahn and I asked Boissy to assemble and lead a team of other medical professionals to research, benchmark, design, develop, and pilot a program for improving medical staff communication skills. She agreed, with one condition: that the training outcome would not be an improvement in HCAHPS scores. She argued that the HCAHPS questions did not capture what truly constituted the physician-patient relationship and that healthcare providers aren’t automatically inspired to change
behavior because of fallible numbers. Rather than an organizational imperative to move a score, Boissy believed the communication curriculum should recognize the extraordinary work done by healthcare providers and capitalize on their experience and insights. She advocated very successfully that this was as much about enhancing physician practice as it was about improving patient perceptions. We agreed.

  Communication Training Must Be Practical

  Our charge was very specific. The program could be based on theory but had to be practical and taught interactively so that doctors would practice new skills. The program needed to be applicable to every type of physician and patient encounter and relevant to busy clinicians. It had to be useful to physicians who brought effective skills to the table as well as physicians particularly challenged by communication, without being viewed as remediation. Credibility and utility were paramount, given that many physicians might not see need for the training.