Improving Hospital Safety, One Physician at a Time

The Center for Patient and Professional Advocacy at Vanderbilt has spent two decades building a professional accountability model. Its director discusses how co-worker observations can be leveraged to promote greater accountability among physicians.

What happens when a lower-ranking healthcare worker observes bad behavior in a physician or another advanced practice professional?

In many organizations, that worker will keep the information to himself.

But that's not necessarily the case at Nashville-based Vanderbilt University Medical Center, says William Cooper, MD, director of the Vanderbilt Center for Patient and Professional Advocacy. He recently co-authored a study regarding peer feedback and physician behavior modification.

Published this month in The Joint Commission Journal on Quality and Patient Safety, it found that almost three quarters of clinicians confronted about bad behavior by a peer did not to be confronted again.

Establishing a culture that values safety is mandatory for a program like this to work, Cooper told me in a recent interview. "You have to have commitment to it, and it has to permeate the entire culture." The transcript below has been lightly edited.

HLM: Tell me a bit about the background of this study.

Cooper: The Center for Patient and Professional Advocacy has spent the last 20 years building a professional accountability model, which was originally created for identifying and intervening with physicians at risk of medical malpractice.

We began to recognize over the years that there are also ways physicians and other healthcare professionals interact with team members that might have an impact on key performance and quality.

Our malpractice work has used patient complaints as a source of [identifying] high-risk physicians. We know that 3% to 5% of physicians account for 45% to 50% of patient complaints. And, those same physicians account for 50% to 60% of medical malpractice risk.

But we hadn't had the chance to explore whether staff complaints were also non-randomly distributed. What we found was that in three years' worth of staff complaints, 3% of our physicians accounted for almost half of our staff complaints.

Over the course of a two-to-three-year period, we were able to create a system for bringing concerns to healthcare professionals that were generating complaints by other staff, particularly those who generated more than their fair share.

HLM: How do you intervene when you receive a complaint about a clinician?

Cooper: The model for intervention is based on a peer-to-peer conversation. A nonjudgmental conversation with a peer can actually be quite effective in changing behavior. Based on academic detail and research that Vanderbilt did… [regarding] physician prescribing behavior… if you sent a physician to have a conversation with another physician, and [their peer] showed them… data [regarding their] peers' [behavior], they can self-regulate and change their prescribing behavior.

We've used [this model] for reducing patient complaints, and now with this most recent paper, we've been able to address staff complaints as well.

HLM: What is the conversation like?

Cooper: We do this in a non-public place, so there's no public embarrassment. It usually takes about 90 seconds to two minutes, and usually goes something like, "we received this report about an interaction you had with one of our team members. It's inconsistent with the values that we have here at Vanderbilt, and all we want to do is make you aware and tell you a bit about what happened."

Afterward, I'd ask for your views. I'd then end the conversation by saying, "I appreciate your time."

HLM: By what percent were patient and staff complaints were reduced once this model was adopted?

Cooper: What we find is that in our patient complaint model, 80% of physicians who received "peer messages," or a peer-to-peer intervention, will self-correct. In the early experience we have had, around 70 – 75% have had no subsequent staff complaints after receiving a physician conversation.

What's especially interesting about this is that these are sometimes people who have been doing these behaviors for years, and no one has ever told them [their behavior was unacceptable].

HLM: Does this program only apply to physicians, or to nurses and other employees as well?

Cooper: We do share these concerns with our advanced practice nurses as well.

HLM: What's a good example of disrespectful or unsafe behaviors that might trigger an intervention?

Cooper: Some might include something like a physician receives a call about a patient, then comes to the nursing unit where the call was received. He then says something like, "this patient is fine, I can't believe you wasted my time on this call. Don't ever call me again." Or maybe a physician has an outburst in a procedure room.

It's not all about cussing and spitting and throwing scalpels, though. Sometimes, these behaviors are more passive.

Not returning phone calls or not answering pages can create challenges with our goals to provide high-quality care to our patients. As leaders, we want to make sure to give the physician a chance to self-correct.

HLM: What would you say to concerns that this policy asks employees to tattle on each other?

Cooper: We definitely are mindful to the fact that in an ideal circumstance, [workers] could have… conversation [with offending clinicians] directly. But we recognize that within an organization, there are often barriers to that kind of conversation.

But we need to give people an opportunity to share their observations. This [set up] also provides a great way to [collect] data, because if a physician works in multiple areas of the organization, and we don't have a centralized way of seeing this data, we'll never know that this [behavioral] pattern is developing.

But to do that, team members have to trust that they'll be safe, and that no one is going to retaliate.

We analyze the data to make sure we don't have super-reporters, or groups of nurses ganging up on a physician, for example. But we've found that super-reporters are almost non-existent. Which is good, because that would be unprofessional, and that team member would have to get feedback about that unprofessional behavior.

HLM: Has leadership noticed the results?

Cooper: Recently, I was sharing some data with our CNO. She told me that she's had two nursing managers come to her recently and say, "I don't know what happened to Dr. Y, she used to be so mean, but she's been just amazing to work with recently." The CNO didn't know, but Dr. Y had recently received an intervention.

That's how you begin to get traction for this kind of culture change.

HLM: What advice do you have for leaders who wish to pursue a similar cultural change?

Cooper: One of the most critical lessons we learned in putting this together is that you have to pay a lot of attention to having the right people and processes. You have to have leaders that are committed. They have to align with the organization's values, align with the policies and procedures, and you need an intervention model to identify those [best suited to intervene] and train them to give non-judgmental feedback.