Survey a launching pad for wide-ranging improvement

Despite buy-in, opportunities to improve identified

The performance improvement team at the M.D. Anderson Cancer Center in Houston has used an employee opinion survey as a guidepost to opportunities for improvement.

The ultimate result: an ongoing program called "Create Solutions" that has generated more than 150 PI projects to date. The survey, conducted about three years ago, was a first for the facility and generated a 60% response.

"We learned that everyone was really behind our mission and vision, but that there were still some opportunities for improvement," explains Tina Smith, MBA, manager of new programs for M.D. Anderson's Institute for Healthcare Excellence.

Each executive vice president in the organization was tasked with taking the survey and looking at those opportunities, she explains. "We report to the COO, and what we suggested to them was that under the COO umbrella, we take [the survey results] to the frontline employees and ask them what their suggestions were for improvement," Smith says.

Creating the teams

To accomplish their goals, the frontline employees first had to be taught how to facilitate outcomes, she says.

"We formed teams of employees, chose team leaders, and taught them PDCA [plan-do-check-act]," Smith recalls. Training also included teaching team leaders and managers how to coach teams.

The charge to the teams was to work with leadership to focus on one question off the survey that affected their area the most.

"Then we would try to do an analysis around what caused that score to be where it was and pick one of those areas to run a quick-win team on," Smith explains. (Topics chosen varied from nursing orientation to communication, from patient flow to room utilization to building trust.)

Two initial rounds of these team projects were run - which generated "a lot of learning and growth," says Duke Rohe, systems improvement specialist. "From that grew a more formal program from January to September this year, which we ran at a higher level."

It began with communication with top-level executives in January, followed by kickoff meetings with midlevel managers in February.

"In March, we held a conference-style session for team leaders and managers, and there were about 175 participants," Rohe adds.

Having gone through two pilot programs, it had become clear that there was a wide variety of skill levels among the teams, so the learning sessions were separated into three class levels:

  • Level 1 is for those with the least amount of experience and knowledge.
  • Level 2 is for those with greater knowledge and experience.
  • Level 3 is for the most advanced.

The four-hour session was kicked off by presentations from the COO and chief nursing officer, with a concluding keynote address by a physician who shared his thoughts on why he became a physician and what it meant for him to work at M.D. Anderson.

Once the teams were off and running, they had a growing body of knowledge upon which to draw; the round 1 and round 2 outcomes had been entered into a database, which they could access via the organization's Intranet to model, to brainstorm, and to benchmark.

"The database was originally set up two years ago and has evolved over time," Rohe says. "Its whole purpose is to be a proven practices database. You can, for example, do a key word search," he explains.

"You could also perform a search of any question from the survey," Smith adds. "You could see who else had worked on it and what they had done."

Rohe's goal is to have every improvement-related project entered into the database; at present, he says, there are about 175.

"There is also a virtual side," he adds. "If a team feels they want some hand-holding, I can sit there [at my computer] and ask, for example, if they have considered certain options. They can e-mail me a question, or I can view it myself and fire them off a note; it's also a great use of my time."

Hand in hand with this phase was a webcast, hosted by Rohe and Smith, that addressed key steps in Create Solutions:

  • What is the project purpose?
  • What are the main causes of the problem?
  • What was the implementation plan?
  • What was the baseline level of performance?
  • What was the solution?
  • How will you measure and maintain improvement?

"This included a team walk, where we'd go out and feature a team that did a particular step and what their experience was," Rohe relates.

Other sessions covered specific tools, such as cause-and-effect diagrams. "As we were coming up to our JCAHO survey, we sprinkled some of that in, too," he adds.

While the initiative still is in its infancy, Rohe says that so far, "It's been great." As of the summer, he notes, 105 projects had proven results - with the data to back them up.

While Create Solutions originally was designed for the employee survey response team, "We knew it would grow, and we would open it up to all PI projects," Smith notes.

"It could be services, safety - anything," Rohe concludes.

Need More Information?

For more information, contact:

Duke Rohe, Systems Improvement Specialist, M.D. Anderson Cancer Center, Houston. Phone: (713) 745-4433. E-mail:

Tina Smith, MBA, Manager, New Programs, Institute for Healthcare Excellence, M.D. Anderson Cancer Center, Houston. Phone: (713) 745-4435.