5.4 Improvement Science: Model for Improvement, PDSA, Lean, Six Sigma & Gap Analysis
Key Takeaways
The Model for Improvement asks three questions—what are we trying to accomplish, how will we know a change is an improvement, and what change will result in improvement—then tests changes in PDSA cycles.
PDSA tests start small, include a written prediction, and scale up through ramps as evidence builds.
Improvement projects track outcome, process and balancing measures; balancing measures reveal unintended problems elsewhere.
Lean removes waste such as waiting and motion from the patient's perspective, while Six Sigma's DMAIC sequence reduces variation and defects.
A gap analysis measures the distance between the current state and a defined target and plans how to close it, while a SWOT analysis shapes strategy.
5.4 Improvement Science: Model for Improvement, PDSA, Lean, Six Sigma & Gap Analysis
The CENP outline expects nurse executives to know patient safety and improvement sciences, to build them into the organization's practices, to use analytics such as gap analysis and SWOT to decide whether change is needed, and to use change tools such as PDSA and Lean methods. Safety science (sections 5.1–5.3) explains why harm happens; improvement science supplies the disciplined methods for testing and spreading better processes.
The Model for Improvement and the PDSA Cycle
The Model for Improvement, developed by Associates in Process Improvement (Langley and colleagues) and promoted by the Institute for Healthcare Improvement (IHI), starts with three questions:
- What are we trying to accomplish? A specific, measurable, time-bound aim, such as "reduce falls with injury on 6 West by 50% within six months."
- How will we know that a change is an improvement? A small family of measures.
- What change can we make that will result in improvement? Ideas drawn from evidence, front-line insight and other organizations.
Changes are then tested in rapid Plan-Do-Study-Act (PDSA) cycles:
| Phase | What happens | Executive expectation |
|---|---|---|
| Plan | Define the change, who will test it, where and when; predict what will happen; plan data collection | A written prediction makes learning explicit |
| Do | Run the test on a small scale (one nurse, one shift, five patients) and record problems | Start small so failures are cheap and safe |
| Study | Compare results with the prediction; analyze what was learned | Use run charts rather than single before-and-after numbers |
| Act | Adopt, adapt or abandon the change; plan the next cycle | Increase the scale of testing as confidence grows |
Successful tests are linked into ramps: a change moves from one patient to one unit to several units, then into standard work. W. Edwards Deming's earlier Plan-Do-Check-Act (PDCA) cycle is the same idea; Deming later preferred "Study" to emphasize learning over inspection.
The Three Types of Measures
| Measure type | Question answered | Fall-prevention example |
|---|---|---|
| Outcome | Did the system's results improve for patients? | Falls with injury per 1,000 patient days |
| Process | Are the changed steps being done reliably? | Percentage of high-risk patients with bed-exit alarms on and toileting rounds documented |
| Balancing | Did the change cause problems elsewhere? | Restraint use, nurse overtime, patient sleep complaints |
Balancing measures are a frequent exam target: a falls program that quietly increases restraint use has not improved care.
Lean: Eliminating Waste From the Patient's Perspective
Lean grew from the Toyota Production System and was adapted to health care by organizations such as Virginia Mason and ThedaCare. It defines value from the patient's perspective and treats everything else as waste to be removed. A common memory aid for the eight wastes is DOWNTIME:
| Waste | Health care example |
|---|---|
| Defects | Medication errors, rework on incomplete documentation |
| Overproduction | Ordering tests or supplies "just in case" |
| Waiting | Patients waiting for beds, transport or discharge medications |
| Non-utilized talent | Nurses performing tasks that clerks, transporters or pharmacy could do |
| Transportation | Moving specimens, supplies or patients farther than necessary |
| Inventory | Overstocked supply rooms, expired items |
| Motion | Nurses walking repeatedly to distant supply or medication rooms |
| Extra-processing | Duplicate charting and redundant assessments |
Core Lean tools include value stream mapping (charting every step and its wait time from request to completion), 5S (sort, set in order, shine, standardize, sustain) for workspaces, standard work, visual management and daily huddle boards, A3 problem solving (a single-page structured analysis from background through follow-up) and kaizen or rapid process improvement events. Leaders practice gemba walks—going to where the work happens to observe and ask questions.
Six Sigma and Lean Six Sigma
Six Sigma, developed at Motorola, aims to reduce variation and defects using the DMAIC sequence: Define the problem and customer requirements, Measure current performance, Analyze root causes, Improve the process and Control it to sustain gains. "Six sigma" performance corresponds to about 3.4 defects per million opportunities. Lean Six Sigma combines Lean's focus on flow and waste with Six Sigma's statistical control of variation, and both are part of the robust process improvement toolkit described by Chassin and Loeb.
| Method | Best suited to | Typical tools |
|---|---|---|
| Model for Improvement / PDSA | Testing changes quickly where the solution is not yet known | Aims, measures, run charts, PDSA ramps |
| Lean | Slow, wasteful flow (throughput, discharge, supply) | Value stream maps, 5S, standard work, A3 |
| Six Sigma (DMAIC) | Variation and defects in a defined, measurable process | Process capability, statistical analysis, control charts |
Gap Analysis: Deciding Whether Change Is Needed
A gap analysis compares the current state with a desired state—a regulation, benchmark, evidence-based standard or strategic goal—and defines the actions needed to close the difference.
- Define the desired state precisely (for example, Joint Commission NPG requirements, a top-quartile HCAHPS target or an evidence-based protocol).
- Measure the current state with data, observation and staff input.
- Quantify and describe the gap (how large, where, for whom).
- Identify root causes of the gap.
- Plan and prioritize actions, owners, resources and timelines.
- Monitor until the gap closes.
A gap analysis differs from a SWOT analysis (section 7.4): SWOT scans internal strengths and weaknesses and external opportunities and threats to shape strategy, while a gap analysis measures the distance to a specific target. Executives often use SWOT to choose priorities and gap analysis to plan how to reach them.
Building Improvement Into the Organization
Executives make improvement science routine rather than episodic by:
- Building capability through improvement training at several levels, from frontline basics to advanced coaches.
- Linking projects to strategy so improvement work addresses board-level aims (see Hoshin Kanri in section 8.1).
- Creating escalation paths: tiered huddles move a problem from the unit to the department to the executive safety huddle when it cannot be solved locally, so performance gaps reach leaders quickly.
- Protecting time and resources for testing, data support and coaching.
- Spreading what works with standard work, training and monitoring across units.
A nurse manager wants to test a new hourly toileting routine to reduce falls. According to the Model for Improvement, what is the best first step after setting an aim and measures?
Run a small PDSA test with one nurse on one shift, predicting the result and recording what happens
Launch the routine on every inpatient unit at once so the results can reach statistical significance
Ask the board to approve a new hospital-wide fall policy before any testing begins on the units
Wait for a randomized controlled trial in the literature to prove the routine works elsewhere first
A value stream map of the discharge process shows patients waiting an average of three hours after the discharge order for medications from the pharmacy. In Lean terms, which type of waste does this primarily represent?
Extra-processing
Waiting
Overproduction
Non-utilized talent
A hospital cuts patient falls with injury by 40% after introducing frequent toileting rounds. Which measure should leaders also track to confirm the change has not created a new problem?
The percentage of scheduled toileting rounds completed
The number of staff who attended fall prevention training
Falls with injury per 1,000 patient days on the units
Restraint use and nurse overtime on the same units
Sections you finish are checked off in the contents.