There are several specific improvement frameworks that can be used by a team to address a quality-of-care problem and perform a quality improvement project. The framework chosen depends on the type of problem that is being targeted and the training of the individuals on the improvement team. Three of the most commonly used improvement frameworks include the following: 1) Six Sigma; 2) Lean; and 3) Model for Improvement.
AGA Institute
Figure 1. Common diagnostic tools used for root cause analyses. (A) Fishbone diagram and (B) Pareto chart. HD, high-definition; prep, preparation.
Six Sigma
Six Sigma is focused on improvement by reducing variability.8 It is a highly analytic framework relying on statistical analysis and mathematical modeling. It is best suited for projects in which the root cause and contributors to the target problem remain unclear and the aim of the intervention is to reduce variation.
Lean
Lean emphasizes improvement through elimination of waste and classifies all parts of any process as value added and nonvalue added.9 It is estimated that 95% of activities in any health care process are nonvalue added and the objective of Lean is to identify opportunities to simplify and create efficiencies. It is best suited for target problems that directly can be observed and mapped out, for example, process of care, flow, and efficiency of an endoscopy unit.
Model for Improvement
The Model for Improvement has been popularized by the Institute for Healthcare Improvement.10,11 It is well suited for health care teams, and its advantages are its adaptability to many improvement targets and lack of extensive training, consultant support, or statistical training as required by the previous frameworks mentioned earlier. As a result, it is the most commonly used improvement framework.
Using the Model for Improvement
The Model for Improvement is organized around three main questions: 1) What are we trying to accomplish? 2) How will we know that a change is an improvement? and 3) What changes can result in improvement?
Question 1: What are we trying to accomplish?
The first stage using the Model for Improvement is developing a clear project aim. A good aim statement should be specific in defining what measures one is hoping to improve and setting a concrete deadline by which to achieve it.10,11 It should answer the questions of what the team is trying to improve, by how much, and by what date. It is more effective for the target to be an ambitious, stretch goal to ensure the effort is worth the resources and time that will be invested by the team. Not only does a good aim statement serve as the foundation for the project, but it can redirect the team if the improvement effort is getting off track. In the earlier example of improving ADR, an aim statement could be “to increase the ADR of all endoscopists who perform colonoscopy at your hospital to 25% over a 12-month period.”
Question 2: How will we know that a change is an improvement?
This step involves defining measures that will allow you to understand if changes implemented are impacting the system within which your target problem resides and if this represents an improvement. This usually involves continuous, real-time measurement. Outcome measures are clinically relevant outcomes and are the ultimate goal of what the project team is trying to accomplish. In the example of ADR, this could be the proportion of endoscopists at your institution with an ADR greater than 25%. Process measures are relevant to the system within which you are working and your target problem resides. Typically, the intervention that you implement will have impact that is measurable much earlier by process outcomes than outcomes measures, which are usually a downstream effect. As such, an improvement project still may be a success if it shows improvements in process measures only. For example, the proportion of endoscopists measuring withdrawal time would be a process measure in an intervention aimed at improving ADR. In time, improvement in process measures may translate to improvements in the outcome measure. Balancing measures are indicators of unintended consequences of the project. Not all that comes from an improvement effort is necessarily positive. If improvements in certain process measures come at the cost of harms shown by the balancing measures, such as deterioration in staff satisfaction or increase in time per procedure, the improvement project may not be worth continuing.
