It’s all about the data. The wrong data. Old data. No data.

Every day, unit-based teams aim to improve their care of patients, but for one reason or another, at times they lack the measures they need to track their progress. Then it’s up to teams to create their own measures—and get the right data.

What they develop depends on the circumstances.

For example: A patient satisfaction survey was creating a head-scratcher for a facility-wide UBT at the Sugarhill Buford Medical Office Building in Atlanta. Overall dissatisfaction with the medical center was loud—but not clear.

“It’s been as low as 43 (43^rd percentile), and it’s jumped to 65,” says Jan Ritter, medical office administrator and management co-lead. Yet patients reported positive encounters, in the 90^th percentile and better, with their nurses, doctors and receptionists.

The team wants to find out what’s causing the split assessment, Ritter says, investigating likely culprits such as wait times: “Sometimes new members come in with the idea that it’s one-stop shopping, like Jiffy Lube—in and out in 30 minutes.”

The existing survey, however, doesn’t provide feedback quickly enough to be able to run a test of change according to the plan, do, study, act (PDSA) steps of the Rapid Improvement Model (RIM). Members aren’t asked about their visits until a few days or weeks after their appointments. The UBT doesn’t see survey results for at least two months.

The team’s quest for usable metrics has become a test of change in and of itself. Its first step was to create a basic survey asking, “Were your expectations met on this visit?”

The brightly colored survey cards included space for comments and were handed out to patients during the after-visit summary. Collection boxes were placed by the exits. The response was positive (99 percent answered yes)—and, the team realized, too simplistic to be useful. Now the team is working on questions designed to distinguish individual encounters from the overall visit.

“We need to pinpoint where and why we are having problems,” Ritter says.

Starting at the beginning

In Honolulu, the Labor and Delivery unit nurses at Moanalua Medical Center faced a different issue. Team members knew they wanted newborns to spend more time just after birth bonding with their mothers in a practice known as skin-to-skin contact, which has many benefits. But they had only a general sense of what was currently happening. Before proceeding, the team needed to develop a baseline measurement.

So this summer, the RNs painstakingly took notes on every patient, recording how much contact babies and mothers had: more than an hour, less than 30 minutes, or between 31 and 59 minutes. With that data in hand, the team created a goal that was SMART (strategic, measurable, attainable, realistic/relevant, time-bound), aiming for at least one hour of skin-to-skin bonding time.

Most of the staff thought 30 to 59 minutes was sufficient, says labor co-lead Kris Oishi, RN, a member of Hawaii Nurses Association, OPEIU Local 50. “Once we educated everyone on (the need for) a minimum of one hour skin to skin,” she says, “we immediately saw an increase in our percentages.” (For more on the team’s work, see "Nurses help newborns get closer to moms.")

 Eric Tom, LMP program manager and management co-lead for the Hawaii region, notes that project-specific metrics will vary widely. A Hawaii OB/GYN team measured patient time from admission to epidural injection for a pain management project. The Ambulatory Surgery Recovery UBT gauged its recycling efforts with total weight of recyclables collected.

But in all cases, Tom says, it’s critical to start a project by identifying a comprehensive set of measures and planning how the data will be collected. This ensures baseline data is available before the team sets its SMART goal and streamlines the process once the project gets under way. 

Methodology: Process vs. outcome

Teams also need to consider whether they want to measure outcomes (the final result, such as how many people with diabetes are getting retinal screenings) or process (what they have to do to get to that result).

“Regional goals are often outcome-based or (tied to) performance metrics such as HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) scores or workplace safety,” Tom says. But the team-level actions that affect those scores are things such as hourly rounding on patients and having safety conversations—processes the team can track on an hourly, daily or weekly basis, providing real-time feedback that can be acted on quickly.

In Southern California, a Riverside Medical Center perioperative team initially used an outcome metric—a patient satisfaction survey—that didn’t work well for it for two reasons. It wasn’t reflecting the team’s role in patient service, for one thing: The survey is aimed at patients who have had a stay of at least 24 hours, which omits the majority of the perioperative unit’s patients, who have outpatient surgery.

“It’s a good survey but was totally the wrong survey for them,” says Priscilla Kania, Riverside’s senior LMP consultant. Moreover, Kania says, in performance improvement work, teams often need a process measure—not an outcome measure—to determine how the system is working.

“You have to come up with your own way of measuring the patients’ experience with your process,” she says. “The team didn’t need survey results two months from now. It needed feedback today.”

The team’s solution was to begin measuring patient perceptions of specific team behaviors, says labor co-lead Tracie Girard, a UNAC/UHCP nurse. She and Josephine Murphy, a clinical nursing director, took the lead in developing a questionnaire that asks patients what the unit can do to improve the care experience for members.

“All but one of the questions were designed with a ‘yes’ or ‘no’ response,” says Girard. “We are off to a good start: We reviewed the first set of data in our last UBT meeting. Together, the team identified a SMART goal in UBT Tracker.”

Keeping it simple

Finally—the data should be easy to collect.

In Atlanta, the pediatrics team at Cumberland Medical Center reduced waste simply by making pencil marks on note cards taped to a cabinet. The UBT suspected its disorganized system for storing patient-care supplies was both creating duplicate materials and causing shortages.

It had to figure out: How do you measure wasted time and effort?

The team decided to place note cards on each of the supply cabinets. On each visit, the staff member noted what was retrieved. The team members tracked multiple trips to multiple places for the same item.

“In the next meeting, they had their tallies,” says Natalie Ines White, a Georgia performance improvement adviser for UBTs. “They invited the facilities person to talk about closets and cabinets that could be used to store all inventory in one place. They discussed par levels—the maximum or minimum amount of bandages needed at a given time—and talked about inventory they could give to other departments.”

In a month’s time, the team documented 154 individual trips to five different supply closets. Since then, they have eliminated one closet that wasn’t even in the unit and are working on consolidating everything into three closets: one for medication, one for general supplies and one for bulky orthopedic supplies.

“All of this came from thinking outside the box,” says White.