Leigh Hamby, MD, CMO of Atlanta-based Piedmont Health, thought his system was effectively executing quality initiatives when a surprising Leapfrog rating shook the status quo. Suddenly, he realized his team needed to reexamine their understanding of quality.
Dr. Hamby and his team consolidated individual quality programs from hospitals across the Piedmont system to form the System Quality Office, which standardized all quality efforts. As part of Becker's "Steal this idea" series, Dr. Hamby discusses how he instituted and scaled the System Quality Office and how it has helped all 11 of Piedmont Health's hospitals.
Dr. Hamby invites you to steal this idea.
Editor's note: Responses have been lightly edited for length and clarity
Question: What led you to create the System Quality Office?
Dr. Leigh Hamby: In 2014, our largest flagship hospital received a 'D' in the Leapfrog ratings. It was really a wake-up call because, prior to that, all of the ongoing quality indicators we held ourselves to were fine. At the time, our structure for quality was pretty traditional, with each hospital determining its own quality priorities.
The Leapfrog rating really forced us to think about the function of our quality departments. The service quality teams deliver is improvement at the bedside, and we asked all the folks that work in the hospitals in quality roles to tell us what they did on a daily basis to achieve this function.
Our hypothesis was that you really need four things to improve care at the bedside: surveillance, analysis, design change and implement change. We found 80 percent of what our people were doing was surveillance and analysis. They were reviewing charts and defects. If somebody had an infection, they'd run over there, review the chart and say what went wrong. In terms of the time spent actually designing and implementing improvement at the bedside, we had very little activity — I would say 20 percent. So we had people that were good at counting things — and again, that's an important skill — but that's about all we could do.
Q: So what did you do to address this issue?
LH: We redefined all the job descriptions in our collective functions. At the time it was all hospital-based. We brought it into a single corporate office with 100 employees to say all of these resources are going to be corporate assets that we move around the system as needed, based on where the problems are.
We mapped out a standard work process. We realized we have three or four people looking at the same chart three or four times. Instead of eight people spending 10 percent of their time on surveillance at hospitals, let's have one person do it for the whole system, and let's automate it. Let's build a computer algorithm that says if the lab result is normal, you don't even need to look at it. We wrote pretty simple computer algorithms that highlight the reports we should probably start with so we could surveil the whole system with just one or two people.
Then we reloaded our labor balance. Instead of being 80 percent finding problems, 20 percent fixing problems, we changed it to 50-50. Fifty percent of the labor hours we had were surveillance and analysis, the other 50 percent were people who were going to drive improvement teams that would define standard work processes and actually help implement them at the bedside.
Then we said let's define the top three or four existing problems we already know about, and let's fix those so they stay fixed rather than review more charts looking for more problems.
Q: What is the design process like?
LH: We needed people who could actually design standardized processes. So we actually got process engineers, and now half of our staff is clinical and half are process engineers. Our two business units are led by a process engineer. One leads a surveillance and analysis group, and the other leads the improvement group.
We've actually gotten really good clinical feedback. Because these engineers don't have a lot of clinical experience, they can ask questions and involve the clinical staff to design these processes that help standardize care and lower incident rates. It helps to have a dialogue where a surgeon or physician explains the rationale behind their process. Those insights allow the engineers to have a good, unbiased conversation with clinicians.
Every order set or every standard work process is co-designed with physicians who have these engineers alongside them, as well as the other clinicians. The doctors like having the engineers help them pull the data, ask the questions and map the process so they can really be left to think about the actual delivery of care.
Q: How do you implement these standard processes?
LH: We have a very specific way of doing it, so that as we move along, we share it with the rest of the organization. We test processes for 60 days before we roll them out system-wide. We have a very systematic approach, and we are constantly looking for feedback.
Most organizations already know about all of the problems, but they don't work on a lot of them. So in our way of looking at things, we don't go actively looking for more problems. We focus on the ones that we know about and work on them. Our effectiveness measure is to reduce harm to patients.
When someone comes to me and goes, "Hey, there's a really important project, I need your help with X, Y, and Z," the first thing I'm going to ask is how many people are being harmed by that now, because I've got my team currently working on something that we know has either caused harm in the past or is sufficiently likely to cause harm in the near future because it's either poorly designed or there's no standard of work. So it gives us a barometer by which we measure what we spend our time on, and we prioritize those things that have demonstrated harm.
That's a really hard philosophical concept for folks to get their heads around — that you're going to prioritize the work and you're going to stick with that priority based on potential for harm or actual harm and stay focused on that. It means you're going to say no to a lot of stuff.
We have resources that are assigned to every facility. Those facilities can help deploy those resources to a project that they want to work on if that project measures up to our quality safety and service scorecard. We have a three-page scorecard with 90-plus metrics on it. If a problem meets our standards, we can focus on it.
Our primary focus is reducing hospital-acquired infections, and through this process we've reduced the five most common hospital acquired infections by 54 percent over the last two years. We had hospitals that have gone a year without an infection. Some of our bigger hospitals went three months in a row without specific types of infections. Zero is the goal. We're trying to get to zero harm by 2024.
A lot of places in the country do this, but we've achieved an 80 percent reduction on our serious safety events in the last three years. It's held for the last 18 months. We feel pretty good about that.
Q: How do clinicians interact with these standardized work orders?
LH: We hardwire these packages into the EMR. So if you're a nurse on the floor caring for the patient, the computer is expecting to see you do certain things for certain procedures and conditions. We built in the ability in real time for caregivers to see how they're performing according to the standard work process instead of waiting on a report to come out three months from now.
A nurse at one of our facilities, who's been around for 20 years, said, "You know what? I finally feel like for the first time in my career quality is here to help me." In the past, quality was usually the team that would show up on your floor and say, "Hey, guys, your hand hygiene is not meeting requirements. Hey guys, you've got more infections. Stop having infections."
When we implement a new program, our 100 system office employees take 24/7 hour shifts at the hospital and we stand by and answer any questions clinicians have about the new documentation, about the new policy, about the new devices. We call it elbow-to-elbow support.
After implementation, we go back and audit. We have folks go around and randomly complete observational audits to make sure we're still doing the work. That's part of this whole improvement cycle, so we feel like once it's fixed it stays fixed.
We don't wait for infections to occur. We look at the process measures that tell us that we're not doing everything we want to be doing. Let's go jump on it. It's a proactive intervention rather than waiting for problems.