Evidenced-Based Management and a Culture of Safety
William C. Rupp, M.D.
My job is to ask “Is this doable?” You have heard 18 recommendations. Ten of those belong directly to health-care organizations, and I would like to suggest that they are all doable, and I say that as somebody working in and helping lead a health-care organization.
No one recommendation is any more important than the others, but certainly there is going to be more press about such things as staffing levels, hours worked, and design of work. Yet issues about evidence-based management and culture are crucial if we are going to make a safer environment.
Management and culture are less tangible than number of hours worked, but they are a very important part of the safety. So, I want to talk about three very specific examples of how we can introduce and trust evidence-based management and how we can create a culture of safety.
I won’t give the whole outline but just some examples of the kinds of things that can be done in health-care organizations. This comes directly off of recommendation 5.2 that we employ nurse staffing practices that identify needed staffing for each patient care unit per unit, empowering nursing unit staff—and I emphasize nursing unit staff—to regulate work flow and set criteria for unit closure to new admissions and transfers as nursing workload and staffing necessitate.
Now, Ada Sue referred to some leadership practices. It is our job as leaders to balance that tension between efficiency, safety, and making money for the organization so that we stay in business. If this were easy, they wouldn’t need us.
Creating trust throughout the organization is a major challenge be-
cause that is how we get at that culture of safety, and I will give you a very specific example.
We have to actively manage the process of change, and there is a whole literature about managing change: such information as don’t do an across-the-board change in your whole organization on day one. It is the role of pilots and gradual change in organization. It is involving workers in the decision making on issues around work design and work flow. As a leader, it means that I admit I don’t have the faintest idea in most cases what happens at the sharp end of health care, at that individual point of care delivery, and going to those people and asking them the best way to do things and then finally establishing a learning organization.
Now, let me show you a specific example that comes out of my previous site. We called it a capping trust policy. Call it anything you want. It is the basic idea that the nurses on a unit determine on an hour-by-hour basis whether or not they can accept new patients either as new patients or transfers. Notice I didn’t say the nursing supervisor; I didn’t say the unit coordinator. I said the nurses on the unit, and it is done basically hourly, on an updating basis depending on what new patients have come in.
These hourly upgrades are made available to every unit just by using the Internet, and we basically rate each unit on a red, green, or yellow depending on how busy they are.
It is asking the nurses can you handle another patient, and if they say, “Yes,” the unit is green, and if they say, “No,” that unit is red, it means “We can’t handle another patient.” They don’t get another patient no matter what. If there are five empty beds on that unit they don’t get another patient. This means sometimes working with other hospitals to send patients elsewhere or transferring or working around, but it is respecting and treating as professionals the nurses who work on that unit.
Now, I have had hospital administrators say to me, obviously, “Well, you can’t do that; they will cheat and stop working.” My response is very simple, “If you really believe they would cheat, that means you would do that in that position.” I simply don’t believe it.
So, this is one specific example. We call this a capping trust policy. Now, it works into the whole culture of retention. This is the vacancy and registered nurse (RN) turnover rate at Luther Hospital in Eau Claire, Wisconsin, going back from about September 2002 through September 2003.
At the very end you will see the vacancy rate rise. That has to do with a decision to hire about 20 more RNs. So, it factors in, and there are some vacant positions, but notice that turnover rate. It is under 1 percent most of the time, and the vacancy rate is about 1 percent. That is because the nurses are happy working there.
Now, it isn’t a single thing. It isn’t just the capping trust policy. It is all
the things we talk about in this report. There are market-based comparisons to make sure that we are in the market financially. We have a nurse recruiter. There is a recruitment pipeline management. We have an accelerated specialty orientation and education program. All of the things that are mentioned come together to make that culture, but my point is that it is a culture of trust, and it involves recognizing those nurses and listening to those nurses on the floor.
Actually, at Luther the only one who could overrule those nurses was me, and I never did that. I often had a physician in my office saying, “But I was up there and I saw an empty bed.” The answer is, “Yes, would you like to be admitted if they just said that they can’t do it safely?”
These are the reported medical errors on one unit back in about 1998 when we started collecting some of these data by month, and then we began a process of going out. The head nurse and I actually went around to every single unit on every single shift and said, “Do you know how important it is that we start finding out about these errors?” This had all come about after To Err Is Human came out and we looked around and said, “Gee, is that us? That can’t be us.” So, we went out and started asking the nurses.
We explained the fact that we won’t punish you for these things, and in fact I will punish you if you don’t tell us. What happened was we initiated this policy called Fair and Just, and the reported errors skyrocketed. Was that new? Of course not. They had been there all along, just nobody told us.
So, we started doing some things about that, various interventions. They came down some, but still remained quite high because people were telling us about more things that were happening.
We did a culture survey, which was very interesting, Fair and Just; we called it non-punitive initially. I like this term Fair and Just better, but we had a non-punitive policy. I can tell you that. I was the CEO.
One of my staff said, “Maybe we ought to ask the staff out there if we have a non-punitive policy.” So, we surveyed nurses, physicians, pharmacists, and ward clerks; had several different scenarios that we took to them; and said, “In this scenario involving a near miss, a patient with minor harm, and a patient who died, what would happen to you?” Fascinating information. Only 5 percent said that they would even tell us about a near miss. It didn’t happen. It wasn’t an error. Why would I tell you? That means 95 percent of the time when things happened they didn’t even tell us. Would they be criticized for an error? Thirty-four percent said, “Yes”; 23 percent said that it would be used on evaluations; and 76 percent said that there would be disciplinary action if the patient died. Remember I was the CEO of this hospital, and I could have told you we had a non-punitive policy. Staff didn’t quite see it that way.
Going across, the ward clerks who were at the ultimate sharp end, 7 percent wouldn’t tell us about a near miss. Seventy-seven percent thought they would be criticized for an error. Thirty-eight percent thought they were used on evaluations. Seventy-six percent said, “We are out of here if we make a mistake that harms somebody,” and only 50 or 25 percent thought we had a non-punitive policy. It goes over all the way in physicians, of whom 85 percent said, “We will be punished if a patient dies.”
Coming back to the old philosophy, as a nurse or as a physician I am trained that if I work hard enough and study hard enough I won’t make a mistake. If a mistake happened, then obviously I didn’t work hard enough or study hard enough.
So, this baseline survey got us going, and we continue to survey like this to see it improve as people understand that we really are after this reporting.
Now, just a bit of background about getting that culture of safety going, and then I close. When To Err Is Human came out, we went back to our organization and started looking at incident reports, interviews with a number of pharmacists saying, “Gee, we just heard about all this stuff that is happening; is it happening here?” We interviewed a couple of nurses and—who was doing this? It was myself and two other physician leaders; then we sat down and did some chart reviews.
On one unit, for 6 weeks we reviewed 20 charts a week in detail looking for adverse drug events now not just errors but adverse events. We found 5 per 100 admissions. We were flabbergasted; 23 potential adverse events per 100 admissions, 14 pharmacy interventions, and in those 6 weeks there were seven major adverse drug events. Not one got reported up through the system. Not one.
As we took apart that 6 weeks of data, what we found is that 56 percent of the adverse drug events happened at the interface. The interface was from home to admission, admission to transfer, as Ada Sue was talking about, and then discharge, 56 percent of those.
We put in place a very aggressive process to begin reconciling, so roughly 213 adverse events and potential adverse events. We started an admission reconciliation process guaranteeing that we knew what those patients were on once they got into the hospital. It required nurses calling home and getting somebody to go home and look in the medicine cabinet. It is complex, but we got to the point where we were 99 percent certain we knew what a patient was on when they came in the hospital. That dropped the adverse event and potential rate significantly.
We then put in one for discharge because we discovered that patients went home on our medications and continued the things they had been on at home, so that discharge reconciliation and then a transfer reconciliation took out a significant number of those errors. This is all very public
information in the organization. Everybody understands this. They are very proud of this, and it makes it a much better place to work.
My message is that this is doable. There are a number of different ways it requires us as health-care leaders to think differently about our organizations and working with the multiple professional organizations, but those 10 recommendations are doable and there is data and management data to show that we can dramatically increase the safety of our organizations.
