Professional Role · Topic 12 of 21

Quality Improvement Methods

Professional role · quality improvement

Quality Improvement Methods

Ask three questions, test changes in small PDSA cycles, and track outcome, process and balancing measures; RCA looks back at harm, FMEA looks ahead.

Key numbers

Six Sigma: 3.4 defects per million3.4 defects / million

FMEA: risk priority number ranks failuresRPN = S × O × D

Run chart shift: 6 points past median≥ 6 points one side

01 · What it isModel for Improvement: three questions plus PDSA

The IHI Model for Improvement asks three questions — What are we trying to accomplish? How will we know a change is an improvement? What change will result in improvement? — then tests changes in PDSA cycles.

02 · What it isTrack outcome, process and balancing measures

Outcome measures show whether patients are better off; process measures show whether the system works as planned; balancing measures catch new problems elsewhere — for example, readmissions rising when length of stay is cut.

03 · Don't confuseRCA looks back; FMEA looks ahead

Root cause analysis is retrospective: after an adverse event it finds the system causes. Failure modes and effects analysis is prospective: it maps a process to find where it could fail before harm occurs.

The rules

Know these

Aim statement: measurable, time-bound, defined populationA good aim statement is time-bound and measurable (how much, by when), names the population whose care will change, and says where the improvement takes place.

Prefer forcing functions over training and policiesAction hierarchy: stronger actions are forcing functions, architectural changes, simplifying and standardizing processes; intermediate are checklists, read-back and staffing; weaker are double checks, warnings, new policies and training.

Run chart shift: 6 points past medianA run chart plots data over time around the median; six or more consecutive points on one side of the median (a shift) signals nonrandom change; points on the median are skipped.

What it is

Core idea

Lean tools: value stream map and 5SLean defines value from the customer's point of view and removes waste through continuous flow; its common tools are value stream mapping, standardized work and 5S.

Control chart: limits at 3 standard deviationsA control (Shewhart) chart adds control limits 3 standard deviations from the process mean; a point beyond a limit signals nonrandom variation that needs investigation.

Donabedian triad: structure, process, outcomeDonabedian's triad views quality through structure (how care is organized — staffing, credentials), process (what is done to the patient — e.g., VTE prophylaxis given) and outcome (what happens to the patient).

Diagnosis & monitoring

Work-up

Six Sigma: 3.4 defects per millionSix Sigma targets a defect rate of 3.4 per million opportunities — performance 6 standard deviations from the process average.

FMEA: risk priority number ranks failuresIn FMEA each failure mode is scored 1–10 for severity, likelihood of occurrence and likelihood of detection; their product is the risk priority number, and the highest RPNs are fixed first.

Management

Orders

PDSA: test a change on a small scale must knowA PDSA cycle — Plan, Do, Study, Act — tests a change on a small scale (one patient, one day) in the real setting, repeating cycles under varied conditions before implementing and spreading it.

DMAIC: Define, Measure, Analyze, Improve, ControlDMAIC = Define, Measure, Analyze, Improve, Control; unlike PDSA, change is proposed only at step 4 after baseline measurement and analysis, and the Control step sustains gains with ongoing control charts.

RCA²: multidisciplinary search for system causesRoot cause analysis and action (RCA²) uses a multidisciplinary team to reconstruct the event from records and interviews, identify active and latent system causes rather than blame individuals, and produce strong corrective actions.

Find causes with fishbone, 5 Whys, ParetoTo find causes, use an Ishikawa (fishbone) diagram, a 5-Whys analysis and Pareto charts, which rank causes by frequency so the team targets the few that drive most defects.

Root cause analysis or FMEA?

Commonly confused

Timing	Retrospective — after an adverse event or near miss
Typical trigger	Sentinel event or serious harm
Core question	Why did this happen? (active and latent causes)
Output	Corrective actions ranked by the action hierarchy

Commonly confused

Tell apart

Lean cuts waste; Six Sigma cuts variation must knowLean, from the Toyota Production System, improves efficiency by eliminating waste (steps that add no value); Six Sigma reduces process variation and defects using statistics through the DMAIC steps.

More detail

Reference

Show 5 reference facts

Use a small family of measures (typically 4 to 10) and gather 'just enough' data in small sequential samples rather than everything 'just in case'.

Stratify measures by race, ethnicity, language, sexual orientation and gender identity so improvement narrows rather than widens equity gaps.

FMEA is most useful for evaluating a new process before implementation or a proposed change to an existing one; teams list steps, failure modes, causes and effects.

PDSA grew from Shewhart's statistical process control work with Deming; Six Sigma was coined at Motorola in the 1980s, and Lean Six Sigma combines waste and variation reduction.

An institutional review board exemption can usually be obtained for data used for quality improvement (see Research Design & Research Ethics for when a project becomes research).

Pearl

Test small, measure three ways — outcome, process and balancing — and spread only what the data show works.

Sources · Institute for Healthcare Improvement — Model for Improvement · Institute for Healthcare Improvement — Model for Improvement: Establishing Measures · Anhøj J, Olesen AV. Run charts revisited: a simulation study of run chart rules for detection of non-random variation in health care processes. PLoS One 2014;9:e113825 · Monday LM. Define, Measure, Analyze, Improve, Control (DMAIC) methodology as a roadmap in quality improvement. Glob J Qual Saf Healthc 2022;5:44–46 · AHRQ PSNet Glossary — Six Sigma · AHRQ PSNet Glossary — Lean · de Barros LB et al. Lean healthcare tools for processes evaluation: an integrative review. Int J Environ Res Public Health 2021;18:7389 · AHRQ PSNet Primer — Root Cause Analysis · AHRQ CANDOR Toolkit — Event Investigation and Analysis Guide, Appendix H (action hierarchy adapted from the VA National Center for Patient Safety) · IHI — RCA²: Improving Root Cause Analyses and Actions to Prevent Harm (NPSF 2015) and Action Hierarchy Tool · IHI — Failure Modes and Effects Analysis (FMEA) Tool · Li T et al. An integrated FMEA-RCA risk-management approach for DVT prevention after total knee arthroplasty. Risk Manag Healthc Policy 2026;19 (RPN = S × O × D, each 1–10) · AHRQ PSNet Glossary — Donabedian Model · AHRQ PSNet Primer — Measurement of Patient Safety · Starter sheet built from the verified AGACNP fact map (facts reviewed 2026-10-15).