Using frontline staff to improve hand-offs

Let nurses identify problems in processes

When Intermountain Healthcare's LDS Hospital joined with the Joint Commission Center for Transforming Healthcare and nine other hospitals to work on hand-offs, the health system's associate chief medical officer says the first step was identifying which hand-offs the hospital wanted to work on.

"There are lots of different hand-offs that take place in hospitals every day," says Douglas Smith, MD, from the ED to the floor, floor to the ICU, ICU back to the floor, OR to the floor. There are transfers from surgery services to medicine services, from the hospital to the community doctor or extended care facility, as well as from doctor to doctor and nurse to nurse, Smith says.

The first step the participants in the project took was defining hand-offs and elements to measure their effectiveness, Smith says. Then each hospital chose which particular settings it was going to address. LDS chose two specific hand-off situations: from the ED to the med/surg floor and from the OR to the PACU floor. As the project progressed, he says, the team began focusing further on nurse-to-nurse hand-offs.

For these hand-offs, the team looked first at the processes in place and then surveyed both senders and receivers of hand-offs to gauge their satisfaction — a component the team selected as a measure to stratify successful hand-offs — with the current practices.

"On our initial surveys, both senders and receivers were pretty happy with the hand-offs," Smith says. But once the team started digging deeper, it became apparent that each nurse had his or her own style for handling hand-offs; some used "cheat sheets," where they would scribble a few notes. "But a lot of times," Smith says, "it was an index card or a sheet of paper that would get stuck in a pocket." Looking deeper, the team used focus groups and meetings with senders and receivers to answer these questions:

  • In an ideal hand-off, what information would you want? What are the data elements? What are not required data elements?
  • What does the process look like? Think about timing and interruptions and the ability to ask questions.

"We had [nurses] define the elements, and I think this was the powerful aspect of the project. It wasn't The Joint Commission telling this group of nurses what should be in a hand-off. It wasn't the chief medical officers telling them what should be in a hand-off... We really let the voice of the customer speak on this, and because it was their process, not my process, not The Joint Commission's process, it was really adopted very nicely by our nurses," Smith says.

Once the team had identified which elements frontline workers thought should be in a hand-off, it surveyed nurse satisfaction monthly. While the satisfaction rate initially had scored between an eight or nine on a 10-point scale, it dropped to about a six.

Taking the information gleaned from the nurses, the team created a checklist for hand-offs. Basic information was included, Smith says, such as name of patient, medications, allergies, diagnosis, recent labs, pain scores, and pending X-rays. Use was voluntary at the beginning, but staff were told, "In four months time, we'd like everyone to be using the checklist."

"Toward the end of that four months, we were up in the 95% usage of the tool, even though initially there was some resistance... But the more people used it, the more people liked it," Smith says. Staff began asking for more checklists. Over that month, satisfaction rose again to the eight to nine range and defective hand-offs dropped, he says.

"We think that it has led to safer hand-offs. Can I prove that hand-offs are safer? That's hard to do," Smith says. But long-term, he says, once the now-hospitalwide process moves to the entire system, looking at which communication errors led to sentinel events will be a goal.

Initially the tool was tweaked, but "then we got to a point where we said, 'OK, this tool has undergone a lot of revisions, this is what we are going to stick with for the remainder of our project,'" Smith says. Suggestions for improvement keep coming in, and at the end of the pilot phase he plans to introduce some of the changes before rolling out to other hospitals in the system. In preparing for that, Smith says he's also working with information technology to see which data elements can be automatically populated.

"It's not really time-efficient for one nurse to be reading off the electrolytes, hematocrit, and blood gas values and another nurse to be writing it down when it's all there in the computer," Smith says. If those values are populated automatically, "then that lets the sender and receiver focus on the hand-off elements that really need verbal communication. What's the plan for this patient? What are you worried about? What sorts of things do you need to look out for in the next couple of hours to keep this patient safe?"

What causes defective hand-offs?

Smith says one of the biggest reasons for defective hand-offs is the sender not being able to connect with the receiver. Another barrier, one he thinks all sites deal with, is interruptions. He suggests practitioners go to a special place away from the nursing station to do a hand-off.

"Instead of being at the nurses station, we want them to be back in the work room or some place where they are not going to be interrupted," he says, pointing to a "really simple solution" used at Stanford. Little bicycle lamps that flash were placed on medication charts. "The understanding was that if the light was flashing, don't talk to the person or the nurse who is there passing medicines." He says hospitals are looking at "low-tech, simple, creative solutions to getting the word across that this person's involvement is something important. Don't interrupt them."