Published in the June 2005 issue of Today’s Hospitalist
When hospital and practice administrators want to improve the performance of their physicians, many immediately think about changing compensation systems. Align physician pay with performance, they figure, performance will improve and everyone will be happy.
But nothing could be further from the truth, according to Russell Holman, MD, a veteran hospitalist and national medical director for Cogent Healthcare Inc. In his opinion, tinkering with doctors’ pay is not even remotely the most effective way to increase physician performance.
At the recent annual meeting of the Society of Hospital Medicine in Chicago, Dr. Holman urged a standing-room-only crowd of hospitalists to change compensation only after addressing some more immediate “and more effective “performance measurement strategies.
From making sure that you hire the right candidate to clearly setting your program’s expectations and giving useful feedback, he outlined several strategies to make sure everyone is on board with your program’s goals. And just as importantly, he gave the group pointers on how to avoid common pitfalls when it comes to measuring and improving physician performance.
Here are some of the pearls that Dr. Holman has picked up during the years he has spent building and leading hospitalist programs.
Pearl: A really bad hire will leave, but a mediocre hire will stay forever.
Improving physician performance begins with bringing the right doctors on board from the start. While that may sound like obvious advice, Dr. Holman said it’s a point that too many practices overlook. That’s because finding a good fit requires more planning time than many physicians are willing to invest.
“Determining the right physician for you can only come from your definition of your group and its values, goals, strategies and vision,” he explained. “Recruitment is all about preventive medicine. It’s all about stacking the deck in your favor in terms of successful group performance.”
As an example, Dr. Holman described a hospital he recently visited that reached some interesting conclusions about their version of a “perfect” physician. “They said they would take a good clinician with a great attitude and great communication skills over a great clinician with marginal communication skills and a marginal ability to interact with others,” he recalled. The staff at that facility knew themselves well enough to recognize that great clinicians who don’t function well as part of a team would never really fit in.
While the superstar physician may be a bad fit for a team-oriented environment, Dr. Holman said those individuals usually see the writing on the wall and find another job. “Really bad hires will leave,” he explained. “They’ll realize that this is not the place for them, that they don’t fit into a hospital that doesn’t value strong individual performances, but instead values team-based care.”
What about physicians who have problems fitting in not because they have superb skills in one area and are deficient in others, but because they are mediocre in all areas? Dr. Holman said that from an administrator’s perspective, these are the physicians to really worry about.
“Mediocre hires,” he explained, “people who don’t fill out their charts, who show up to work late, who don’t really want to take that 4 p.m. admission, will stay forever. Why? Because their behavior is not dramatically different from expected norms, they grow comfortable in their routines. These physicians are difficult to manage, and they often go unmanaged altogether.”
That’s why Dr. Holman always urges hospitals and hospitalist groups to carefully watch who they hire.
Pearl: Do you want to “make” or “buy” the ideal physician?
When it comes to hiring physicians “particularly if you’re looking for a certain set of skills “you’ll eventually face a fairly basic decision: Do you want to “make” or “buy” the physician you’re looking for? While the terminology is often used by people working in human resources departments, Dr. Holman said it pertains to medicine.
“If we’re really looking for someone with highly effective quality improvement skills,” he explained, “someone who knows quality improvement inside out and has a proven track record, do we want to ‘buy’ that individual from another practice? Do we want to offer a premium salary and resources to attract someone to the group?”
“Or do we want to make somebody,” he continued, “take someone who’s already in the group and give them the resources and training to develop the skills we need? It’s a more labor-intensive approach.”
Dr. Holman said he suspects that physicians tend to avoid the notion of “buying” talent.
“We really don’t do nearly enough buying,” he explained. “We try awfully hard to make people into things instead of looking elsewhere and paying a bit of a premium so people can hit the ground running with the skills we need.”
Pearl: Defining your group’s expectations can help improve performance by itself.
According to Dr. Holman, setting physician expectations is a key part of improving clinical performance. If physicians don’t know what’s expected of them, after all, how will they ever be able to improve their performance in the eyes of their group?
If your group doesn’t have a formal list detailing what is expected from each and every one of its physicians “basic concepts like showing up for work on time, helping out during crunch times, etc. “take some time to come up with a list of 10 or 12 things you as physicians expect of each other. Dr. Holman said compiling this type of list is a great teambuilding exercise, and that its value will go beyond developing a list of measures.
“If the team participates in the process of developing many of the expectations,” he explained, “they in essence become the shepherds of many of those expectations. They develop no tolerance for behaviors and attitudes that don’t meet those expectations.”
Pearl: The job you take today will not be the same job you have next year.
When it comes to managing physician expectations, Dr. Holman said it’s essential to talk about the inevitability of change.
“One of my favorite sayings,” he explained, “and I say this to everyone who has ever worked for me, is that the job you take today is likely to be very different from the job you have next year.”
That kind of honesty helps set physicians’ expectations, a process that Dr. Holman said must not end once you’ve successfully recruited a physician to your group.
“What is the one constant in our work environment?” he asked. “Change. So if you’re setting expectations, they’re likely to change a week or six months or a year from now.”
He explained that his approach is particularly useful when physicians are reluctant to take on new duties.
“Whenever I get complaints from a physician nine, 12 or 18 months later saying this isn’t what I signed up for,” Dr. Holman said, “I come back to that point. I remind the physician that we talked about this, and that change is one of the things you’ll experience anywhere. I am very honest about that.”
Pearl: All measures are flawed, but some are useful.
Try to come up with a list of performance measures, and you’ll inevitably hear complaints like, “My patients are different” or “The data are wrong.” According to Dr. Holman, the best way to handle this kind of criticism is to accept it, clearly explain the relative value and purpose of the chosen measures, and move forward.
“All measures and systems are flawed,” he said, “but some are useful. It helps if you admit that up front and take the words out of physicians’ mouths. Admit that the data are wrong, but that we’re using the same data for everyone, so let’s just move on.”
That doesn’t mean that your group should work with performance measures that are sub par. Dr. Holman explained that measures must be directly related to the work at hand, and they must also give individual physicians a chance to influence or directly control the outcome being measured.
“If you say you expect 100 percent patient satisfaction, 100 percent coding compliance and 100 percent ACE inhibitor compliance with certain patients,” he explained, “you’re probably not being realistic. Trust in you as the person setting expectations will suffer.”
And if physicians are going to value the data they receive, Dr. Holman said, that information must be timely. “If I’m seeing readmission rates or patient satisfaction rates from the last quarter, including two weeks ago,” he explained, “that’s very timely.”
Pearl: Avoid questions that force respondents to make judgments.
While surveying staff about each other can provide useful information to manage performance, Dr. Holman said the process all too often asks for subjective, not objective, information.
“One of the major pitfalls of surveys of primary care physicians and nurses is that they ask questions like, ‘Do you think this person is a good team player?’ ” Dr. Holman said. “They pose questions in terms of character judgment and ask you to judge another person.”
A better approach, he explained, is to focus on objective measures. For example, ask whether a physician comes to meetings on time always, sometimes, occasionally or never.
“You’re quantifying a specific observable behavior,” Dr. Holman said. “In this case, it’s showing up at meetings on time. Here’s another example: Instead of asking doctors in the group if they think that a certain physician completes their medical records, try to quantify that by working with the medical records department. Generate a fair, objective and meaningful report of completion rates that can be shared with each member of the group.”
Tracking data is one thing, but how do you collect information about personal interactions without getting personal? While these questions can be tricky, Dr. Holman said there are some basic strategies to make sure the evaluation process remains fair and objective.
“I’ve seen one group that looks at admissions between 3 p.m. and 6 p.m. and the likelihood of each physician accepting a new admission during those hours,” he explained. “In one seven-physician group, six were clustered on one end of the spectrum and one was at the other end”
Dr. Holman said that everybody in the group knew that one physician rarely took late afternoon admissions, but the group wanted to quantify it. When analysts examined that physician’s productivity, he was the least productive individual in the group.
“It wasn’t about making a value judgment about what kind of team player the person was,” he said. “As it turned out, the physician had horrible organizational skills.” Members of the group shared organizational skills with the physician and are gradually helping improve his performance.
Pearl: Give specific feedback
Dr. Holman said that when giving feedback, try to be specific, relevant and timely. “Going up to somebody and slapping him on the back and saying, ‘You’re doing a great job’ is nice to hear,” he explained, “but it doesn’t mean anything.” As a result, he suggests the following approach, which provides infinitely more detail:
“I was on the ward and I saw you sitting with that patient and holding her hand and talking with her family and listening and not interrupting. I saw you diffuse what could have been a volatile situation. I like the way you handled that in terms of your nonverbal communication and your verbal communication. Taking time out of your day was exemplary. If you’re doing that all the time, we need to learn from you. Keep it up.”
Compared to the five-second pat on the back, the second approach took a mere 20 seconds “and will make a world of difference to that physician.
“If you take time to notice and observe someone else’s behavior and then feed it back to them, do you think that physician going to engage in that behavior again?” Dr. Holman asked. “Absolutely.”
Pearl: Focus on the behavior, not the person.
Finally, Dr. Holman said, focus on objective measures and don’t give in to the impulse to judge physicians when critiquing behavior and performance.
“Focus on the behavior,” he explained, “not the person.” The discussion should not focus on whether someone is a bad person, he added, but on the fact that the behavior was wrong.
“For example,” Dr Holland added, “swearing at the nurses’ station or calling a patient a liar is behavior that is not acceptable under any circumstances, but it does not give us the license to provide physician feedback in the form of character judgments. Although it requires a high level of self-discipline and skill, you will be far more effective if you focus on behavior and not the person.”
Edward Doyle is Editor of Today’s Hospitalist.