Quality improvement (QI) in medical practices is a method for continuously finding better ways to provide better patient care and service. At its core, QI is a team process. Why? Under the right circumstances, teams harness the knowledge, skills, experience and perspectives of different individuals to make lasting improvements. Teams are most appropriate when your process or system under study is complex, when no one person in your practice knows all the dimensions of an issue, when your process involves more than one discipline or work area, when your solution requires creativity, or when you need your employees' commitment and buy-in — in other words, virtually all QI projects. Whether your practice is seeking to improve patient waiting times, telephone service, diabetes care or anything else, a team effort will help you make significant, lasting improvements.

In the first article in this series (“Quality Improvement: First Steps,” March 1999), the authors walked through the process of identifying the key problems in your practice and thinking about how you might address them. It's now time to assemble the interdisciplinary team that will design and implement your QI intervention.

Getting started

Before you assemble your QI team, you must create an infrastructure within your practice that will support the team's work. Without such infrastructure, teams run the risk of being set adrift with no clear course, no resources and no destination. Here are some ideas for creating a QI infrastructure:

1. Be a QI champion and actively support your teams. Without strong endorsement, support and resources from practice leaders, teams will flounder.

2. In larger practices, pull together an interdisciplinary oversight group, such as a quality council, that keeps the QI momentum going and provides guidance and resources to the teams. For smaller practices, a physician and office manager can provide this function.

3. Cultivate a spirit of QI within your practice that encourages everyone to improve the quality of services and programs continuously. You can even develop a set of improvement principles to help guide your practice.

4. Identify internal experts or external consultants who have experience and training in QI to help you get your teams started.

5. Develop staff members' skills in data collection and analysis.

6. Develop staff members' skills in information retrieval, such as conducting literature searches and accessing databases.

Chartering a team. Before a team can be assembled and begin its work, you will need to establish what the team is to accomplish and why.¹ (In large practices, this may fall within the scope of the quality council; in smaller practices, a physician or office manager can lead the task.) Try to establish clear parameters and goals for the QI project, and define the team's mission. The final product of this planning will be a team charter that offers the following information:

• A brief description of the process under study,

• Why the process needs improvement,

• How the team is to demonstrate that the process has improved,

• Who is affected by the process,

• A timeline for team meetings,

• Resources available to the team,

• How the team should communicate its progress with the practice's leadership or quality council.

Recruiting team members. To work efficiently, QI teams need diversity: people with different skills, experience, knowledge and viewpoints. In addition, staff members who are directly involved with the process under study should be represented on the team. For example, a diabetes QI team might include a physician, nurse and lab technologist. Consultants to the team might include database or medical records personnel who could help the team with data collection and implementation of improvements.

As you assemble your team, try to keep it small (no more than six members). Smaller teams are easier to manage than larger teams. Also, you'll want to recruit enthusiastic members, those who are very invested or interested in the process under study. Preferably, allow people to volunteer for the team rather than appointing them.

You will then need to select a facilitator and a team leader. Teamwork has two components: task (which includes the content of the issues being addressed, holding team meetings, etc.) and process (such as how the team members interact with one another while performing their work).² The team leader generally focuses on task, while the facilitator helps with process. The team leader keeps everyone on track by ensuring that the team is making progress toward the charter objectives. The facilitator keeps the team healthy by monitoring participation and group interaction and by intervening as necessary.

Running effective meetings. The key to effective team meetings is planning. Good meeting planning and preparation really do make a difference! We have found that our team meetings are much more productive if the team leader and team facilitator meet briefly before each team meeting. They set objectives and the agenda for the team meeting; identify methods or tools needed to facilitate group idea generation, problem solving or decision making; assign responsibility for each part of the agenda; and allot time for each agenda topic.

A good portion of the first team meeting should be devoted to ensuring that members understand their charge and the standards for working together. The first meeting's agenda should follow this format:

1. Introduce team members, clarify each member's role on the team (why they are there and what they think they can contribute to the process) and explain the facilitator's and team leader's roles. If members do not understand why they are there and what roles they can play on the team, they will be reluctant to participate.

2. Develop team guidelines (also known as standards of behavior or ground rules). A team's guidelines should outline the following:

   The team's decision-making model (e.g., consensus),

   Behaviors encouraged by the team (e.g., respect and good listening skills),

   Team etiquette (e.g., starting and finishing meetings on time).³

These guidelines can move the individuals in a group from thinking in terms of “me” to “we.” Guidelines also help keep the team on track, get everyone involved in the team's work and strengthen the team's ability to discuss issues openly and develop effective solutions.

1. Review the team's charter for clarity.

2. You may find, as the team progresses with its work and gains more information, that the charter needs to be revised.

3. Ensure that everyone understands the problem-solving model being used to keep the group on track with its work. A common problem-solving approach used in QI is the FOCUS-PDSA model, which we discuss below.

4. Begin planning baseline data collection. Your QI efforts must be based on data that show how your practice is presently performing in the area under study. For example, if you were studying diabetes, you could begin by collecting data on the number of HbA_1c tests performed in your practice and the range of results for patients with type-2 (non-insulin-dependent) diabetes.

5. Develop an action plan for the team's next meeting. In other words, determine who will do what and by when. Plan for a focused literature review (if one has not yet been done), which will show you what others have been able to accomplish in the area you are studying.

6. Propose an agenda for the next meeting.

7. Evaluate the meeting. A simple discussion at the end of the meeting of what went well and what you would improve provides immediate feedback and brings closure.

In subsequent meetings, the team will work through the steps of the problem-solving process you've selected.

A QI team at work

QI teams benefit from having a common framework for thinking about the improvement process. One widely used approach is the FOCUS-PDSA model. Very briefly, it's defined as follows:

• Find a process to improve,

• Organize the team and its resources,

• Clarify current knowledge about the process (analyze baseline data),

• Understand sources of variation and clarify steps in the process,

• Select an improvement or intervention.

The second part of the model is the Plan-Do-Study-Act (PDSA) cycle:

• Plan how you will implement the intervention,

• Do it (carry out the change, preferably on a small scale),

• Study the process to see whether your intervention has made an improvement,

• Act on what you have learned, which may mean either implementing the change on a larger scale, starting over or tackling a new area of improvement.

To illustrate the FOCUS-PDSA model, let's examine our group's diabetes QI project.

Find a process to improve. The first article in this series dealt with this step. Remember that the emphasis in QI is on improving processes, not people or single steps. Remember also that the process you choose to address must be one you can measure to judge whether interventions are working. For example, we decided that we needed to improve our care of patients with diabetes, specifically by improving the process of ordering HbA1c tests. Based on our current record-keeping system, we knew we could monitor the number of HbA_1c tests performed.

Organize. This step means bringing together the people and resources that will be most effective in making the needed change. For our diabetes project, we brought together physicians, a lab technologist, a nurse and a staff person from our business office to form an interdisciplinary team. One of the physicians assumed the role of team leader (although a staff member may also be an appropriate team leader), and a staff member who had been trained as a facilitator fulfilled that role.

Clarify current knowledge. This is where your team will use its baseline data and literature search results. The baseline data gives you evidence of how well your practice is performing in the area of study. The literature review gives you insights into how others have addressed similar problems.

At this stage particularly, but throughout the QI project as well, good data collection is essential. The data will provide your team with ongoing feedback, which if positive reinforces the intervention and energizes the staff to keep it up, or if negative motivates the staff to try harder or look to see why they are not meeting with success.

Collecting data isn't always easy, however. Many practices do not have computer systems that can tap the sort of data they want to follow. In those cases, you need to develop manual systems that can store the data in a readily usable format. Even if you do have computerized systems, it may take considerable creativity to get your data into usable form.

Your QI team's data collection plan needs to be very specific. Here are some questions your plan should answer:

• What population are you studying (all patients, only active patients, a random sample)?

• What defines the population you're studying?

• What time frame are you examining?

• Are you interested in values (such as HbA_1c values) or data points (such as the number of HbA_1c tests performed)?

• How will you display and communicate the data?

Our diabetes team conducted a literature review, including clinical guidelines for the care of patients with diabetes, then collected and reviewed data from our practice. We identified all patients who had had an ICD-9 code of 250 within the past year and all patients who were taking oral hypoglycemics. We then audited a random sample of those patients' charts to determine our rate of ordering HbA_1c tests and the average HbA_1c value. Based on the results we decided that, to improve diabetes management, we would initially try to increase the number of patients who received at least two HbA_1c tests per year.

Understand sources of process variation. A key premise of QI is that you cannot improve a process until you fully understand that process. The next step, then, is developing a flowchart to map out the steps of the process you are trying to improve. Flowcharting involves setting the process boundaries, identifying and agreeing on the major steps from beginning to end (the hardest part!), sequencing the steps and showing the flow of action. (For a sample, see “Flowcharting the process.”)

Flowcharting can clarify a team's current understanding of the process and help the group identify problem areas. It's also an excellent way to bring the team together because it fosters discussion, full participation, good listening, acceptance of differing perspectives and learning. In addition, flowcharting helps teams identify key points in the process where data might be collected to shed light on what is actually happening.

Our diabetes team flowcharted the process by which patients received care and HbA_1c tests were ordered. In doing so, we recognized several missed opportunities when patients sought treatment for acute problems unrelated to diabetes and data related to diabetes (HbA_1c values, eye and foot exam results, creatinine values, etc.) were not organized in one obvious place in the medical record.

Select an improvement. With the process clearly defined and your data and literature assembled, your team should be able to choose an intervention. In our case, we felt that to improve overall care of patients with diabetes, we needed to develop a computerized reminder flowsheet (a printout of which would appear on the first page of the medical record at every encounter with patients who have diabetes) that would organize patient data relevant to diabetes management and provide guidelines on the frequency of tests and procedures. Specifically, we hoped the flowsheet would help us increase the number of patients with at least two HbA_1c tests in one year.

Plan. This stage is planning how to implement the intervention you have identified. It's a more challenging step than you might first think. To implement our diabetes reminder flowsheets, for example, we discovered that we needed to educate our office staff, nursing staff, lab staff and physicians about proper diabetes care and how to use the new flowsheets. We had to determine who would print out the flowsheets, who would ensure that the flowsheets were on the first page of the medical records, who would enter the lab values into the computer and who would transfer the ophthalmologists' notes about dilated retinal exams into the flowsheets.

Our diabetes team systematically reviewed the flowsheet system to identify who should be responsible for what activities and then planned and completed the appropriate staff education through our practice's QI bulletin board, e-mail, and group and individual training sessions.

Do. At this point, your team finally is ready to execute its plan. Depending on the project, your practice may choose to implement its intervention on a small scale before going practice-wide. In our case, implementation began when we started using the diabetes reminder flowsheets. Two physicians monitored the appearance of the flowsheets in their patients' medical records to ensure that the intervention was occurring consistently.

Study. Once you have enacted your change and have allowed sufficient time for the change to take effect, review and evaluate the results. To monitor whether our new process was improving our ordering of HbA_1c tests, we used a run chart to track the monthly percentage of active patients who had received at least two HbA_1c tests in the previous year. (See “Using a run chart.”) The chart illustrated that our baseline frequency of about 18 percent steadily increased to the current level of approximately 42 percent.

Act. The next phase of the QI cycle is to reflect and act on what the team has learned. In some cases, this may mean going back to the planning stage and coming up with another solution. In our case, the diabetes team has reviewed its data and is pleased with the improvement. We have given each physician individualized and practice-wide data on HbA_1c performance, and have even celebrated our success with a catered lunch!

For the future, our team hopes that 80 percent of our patients with type-2 diabetes will have all tests and procedures completed (and documented on our flowsheet) and that 80 percent will have their most recent HbA_1c values at less than 7.5 percent. To meet these goals, our diabetes team will explore additional interventions (such as automatic HbA_1c reminders for patients and providers and educational support-group sessions for patients held at the site where blood draws are done).

Not just for diabetes

The FOCUS-PDSA process-improvement model improved care for patients with type-2 diabetes in our practice. The model works equally as well for other areas of improvement and requires only the formation of an interdisciplinary team and use of a few simple data collection and reporting tools. QI has helped us provide better patient care, document our success and energize our practice to do even better. It can do the same for you.