Change concepts are often useful for generating ideas for specific process changes. Sometimes, however, process changes can be identified first. Other sources of ideas for changes include brainstorming with the team, asking front-line staff and patients how they perceive the problem, talking with other departments that may have similar process issues, etc.

Try to identify why previous efforts have failed: separate out whether the idea was flawed or the attempt to implement it ran into barriers.

Emphasize that teams often learn more from failed tests than from successes.

Remember that you will be testing small changes first, so you will be able to identify and adjust to problems as they come up.

Remember that you will be testing ideas for change, so you can learn quickly about which approaches seem to work.

Plan to test several ideas at once.

Use multivoting or other group process tools to make a quick decision about where to start; revisit your decision after you have completed initial testing.

Share plans with your senior leader and get initial feedback; be specific about the skills and/or resources the team needs.

Test on a small scale to see if plans work before making a formal request for resources.

Use small-scale data collection to minimize additional resources needed.

Consider redefining the scope of the project if resources are an impediment; set initial aims for one department or one unit.

Ask senior management to provide additional resources for these individuals.

Make the work a normal part of their day.

Try to design changes that make everyday work easier rather than more difficult (e.g., consolidate different reporting forms).

Identify opportunities to eliminate wasted efforts and unnecessary work.

Make sure front-line staff is involved in designing changes.

Large PDSA cycles are often difficult to complete, absorbing time and energy. Cycles should be short but significant; test a big idea on a small scale and in a short time frame (for example, on the next ten patients In the ICU; on Dr. Smith's patients next Tuesday) so that you can identify ways to improve or change the idea.

In reflecting on what was learned from the test ("Study"), make sure it helps to achieve your aim. Plot outcome measures related to the aim over time.

The "Study" phase of one cycle should be connected to the "Plan" phase of the next one. Schedule specific times for reflecting on what was learned in carrying out cycles.

Documenting all four steps of a PDSA cycle — Plan, Do, Study, Act — has three clear benefits: It helps teams get into the habit of doing all the steps. It is often the only way to follow the thread of the improvement story. It also really helps in "bragging" or creating a coherent demonstration of progress for the future, for senior leaders, or for funders.

Teams often forget to predict. As Deming says, "Learning depends on having a theory." A prediction is the team’s theory about what will happen if this change is tested. Without a prediction, it is hard for a team to assess whether learning occurred or to compare what they expected to what happened.

Use the aim as a reference for defining the outcome measure. Aims should specify clear goals that reflect the change in the system. One approach is to define what you would like the results to show at the end of the project, and derive appropriate measures.

Outcome measures tell whether the changes being made are leading to improvement — that is, helping to achieve the aim (e.g., waiting time in emergency department).

Process measures tell whether a specific process change is having the intended effect (e.g., percent of telemetry patients not meeting criteria for telemetry utilization).

Make sure that measures match aims. Select a few measures that you will track throughout the project. Evaluate whether you need additional measures to help guide changes.

Collect only enough data to support the "Study" phase of PDSA cycles.

Just enough — that is, just enough data to see if an improvement occurred. What matters is testing under a wide variety of conditions with multiple cycles.

Use sampling instead of waiting for information services to crunch numbers. You can compare the results from sampling with those for all patients at a later time.

Use the resources available to you. Paper and pencil are good enough. Updating the organization’s computer system is not a feasible endeavor for a short-term project; a great deal of change and improvement can occur using available resources and just enough data.

Check on the clarity of and commitment to the aim.

Present data simply (use graphs rather than tables).

Do not collect too much data.

Check whether the data help with PDSA cycles; if not, you are probably collecting the wrong data.

Make sure everyone has bought into the aims of the project and that the data reflect those aims.

Show results of data collection to staff on a daily basis.

Make sure the staff understands that data will be used for learning and improvement, not for judgment.

Check whether you are collecting too much data, unnecessary data, or both.

Make it easy for staff to collect data by integrating it into their daily routine.

Emphasize using data in a test, with the focus on how data can guide the next PDSA cycle.

Force an answer to the question, "What action could these data lead to?"

Mandate a minimum data set and stick with it for a month.

Suggest trying a PDSA cycle using existing data.

Differentiate between the level of data sophistication needed for research and that needed for improvement. Randomized clinical trials are needed to establish standards of practice, but not to test best methods for putting standards into practice.

Use sampling to test changes on a small scale; once you agree on improvements, you can collect additional data to verify results. Sampling is based on scientific principles and can satisfy many concerns about validity.

Revisit the basics: aim, measures, and changes.

Structure meetings around the Plan-Do-Study-Act (PDSA) cycles being tested.

Visit other teams to see how their meetings work.

Take advantage of the organization’s resources for facilitators or training in team building.

Consider changing the team membership if problems persist.

Do a PDSA cycle that focuses on improving team functioning.

Check the composition of the team; it should include systems, day-to-day, and technical leadership. A missing component can hamper implementation.

Clarify the specific responsibilities of individuals and the time frames for completion.

Investigate if the staff just does not have time to complete responsibilities; discuss with the team and with the senior leader.

Establish team rules regarding agenda-setting, timekeeping, roles, and responsibilities.

Make sure each meeting has a clear focus and objectives.

Use PDSA cycles to test ways to shorten meetings.

Use brief huddles between meetings to keep in touch and report on progress.

Scale down the project based on time constraints.

Investigate whether the issue is lack of time or lack of a focus.

If time is the issue, discuss resource availability with the senior leader; have other staff cover during meetings; have meetings before or after a shift.

If lack of focus is the issue, revisit aims, and plan for testing changes.

Schedule the next meeting at the beginning of each meeting.

Use brief team huddles between meetings.

Bring staff from other departments into the project; have staff from one department visit the other.

Make patient care goals a priority.

Seek the assistance of the senior leader.

Work with those who will work with you.

Communicate goals and progress throughout the project.

Building relationships is the key to winning over others; involve others in small tests of change to slowly help redefine roles and relationships.

Identify one physician champion and work with that person. Use his or her involvement and improvement results to convince others.

Identify a nurse champion and include nurses who work with the physicians in your projects. The nurses can then work to help you influence those resistant physicians. Building nurses into the process can help with involving physicians.

Perform small tests of change based on good knowledge and experience. If it does not look or feel right, do not proceed with the test. Allow the implementer(s) of the actual change to stop the test if they deem it unwise to proceed (e.g., nursing or respiratory therapy personnel should feel free to not institute a protocol with which they feel uncomfortable). These situations should be viewed as additional learning opportunities.

Costs add up. Calculate the potential cost savings throughout the organization over the course of a year.

Propose that the team be a "skunkworks" to test the proposed improvements, and identify specific resources needed.

Write articles for the organization newsletter. Use storyboards or posters to display progress.

Align the project with organizational goals.

Stress staff involvement in changes, in contrast to top-down decisions.

Focus on patient care as the ultimate goal.

Emphasize connections among cost, productivity, patient care, and patient and staff satisfaction.

Build tension for change.

Start small.

Create opportunities for participation in change.

Ask resisters to suggest alternatives.

Identify and publicize prior successful changes.

Build momentum ("the theory of small wins").

Communicate intentions and progress.

The Model for Improvement is not meant to replace change models that organizations may be using, but rather to accelerate improvement. The model creates tension to test ideas on a small scale, rather than waiting until a solution is fully developed before action is taken.

Align incentives: use data to show potential benefits of your work.

Be realistic: offer senior leaders suggestions about highly leveraged uses of short amounts of their time.

Focus on the concerns of the senior leader and help the leader understand how the project fits into institutional aims and vision.

Request a brief team presentation to be put on the agenda for the senior leadership group or the quality council.

Identify an advocate who has the ear of the senior leader.

Identify specific things you need from the senior leader and let the leader know what they are.

Invite the senior leader to a team meeting.

Keep the senior leader informed of your work; use e-mail, newsletter, briefings, etc.

The most important first step to spreading improvement is to have a model of the new process or system so that everyone can see how it works and how it's better than the old way. The team can help with spread by getting great results and by making their work visible and understandable to others. It is the responsibility of the senior leadership in the organization to develop a plan for involving other sites in the organization in adapting the new process or system.

Establish "stretch aims" or "stretch goals." Show by your constancy and your commitment that you intend to do what is necessary in the organization to achieve the new level of performance. State and restate the aims.

Monitor the process frequently. If you really want improvement to move fast, review progress daily, or hourly if necessary.

Stay in touch with the improvement effort. Support the team leader. Phone the leader. Pay specific, personal attention.

Visit the team. Drop in on meetings, even if only for a couple of minutes.

Visit the site of the change. Walk through the Intensive Care Unit, or the clinic, or the Operating Room. Show that you're aware of the improvement work being done. Look at the charts that are monitoring progress. Ask how you as an executive can be more helpful.

Move the improvement effort up on the agenda. Make it clear that improvement has high priority in the organization. One senior leader who had been formally reviewing the progress of an improvement effort as #9 or #10 on the agenda moved it to #2. That simple act was a key factor in giving the project momentum.

Make sure that the team has the resources it needs. For example, the team may need designated staff time or access to key data that others have collected.

Celebrate successes. Public celebration of improvement efforts or personal expressions of appreciation are just as important as any modification of pay structure.

Keep the bigger system in view. The many local improvements are connected in a larger way to the system as a whole; it is the job of senior leaders to keep sight of, and help others to see, this larger context.

Your aim statement should include what you are trying to accomplish, for whom (patient population or location), the magnitude of change, and the timeframe. Example: Reduce Adverse Drugs Events to less than .10 per ICU day for medical intensive care patients by March.

Enlist the senior leader’s help. The leader has the authority to take the status quo off the table and encourage the staff to move beyond "safe" goals.

As long as the aims are not conflicting, agree that the team will have a dual focus; then work toward unifying the aims as the project develops. If the aims conflict, ask your senior leader for help.

Resist the temptation to weaken goals. Identify barriers to progress and seek solutions instead of changing the goal.

Starting with parts of a system is okay, but be ready to expand the scope of the project once initial aims are achieved.

Constantly focus on aims by repeating or reviewing aims at the beginning of each meeting.

Set numerical targets, and outline an approach and timeline for achieving them. If you can’t see clearly how to plan changes or how to measure progress towards the aim, try redrafting it to make it more actionable or seek guidance from your senior leader.