EPI key terms

A measurement framework that tracks performance across four linked perspectives: financial, customer, internal process, and learning and growth. It connects day-to-day measures to strategy so improvement efforts are not judged on cost alone.

A documented set of long-range goals, priorities, and actions that directs where an organization invests effort and resources. Improvement projects should trace back to it so that gains support the mission rather than local preference.

The formal arrangement of committees, roles, decision rights, and reporting lines that oversees improvement work. It clarifies who approves projects, who allocates resources, and who is accountable for results.

A senior leader who champions an improvement initiative, removes organizational barriers, and secures funding and staff time. The sponsor owns the business case and is accountable for realizing the intended benefit.

A short authorizing document that states a project's problem, objective, scope, boundaries, team, timeline, and measures of success. It prevents scope creep by recording what is in and out of bounds before work begins.

The uncontrolled expansion of a project's boundaries as new requests are added without adjusting timeline, resources, or goals. It dilutes focus and often causes missed deadlines and incomplete benefits.

A ratio comparing the financial benefit gained from an improvement against its total cost, usually expressed as a percentage or payback period. It helps prioritize competing projects and justify continued funding.

A quantifiable measure chosen because it reflects progress toward a specific goal. Good indicators are few, clearly defined, owned by someone, and tied to a decision or action.

A precise written description of how a measure is calculated, including data source, inclusion rules, formula, and collection frequency. It ensures different people count the same thing the same way.

The recorded starting level of performance before an intervention, used as the comparison point for judging improvement. Without a credible baseline, claims of gain cannot be separated from normal variation.

A visual display that consolidates selected indicators, trends, and targets so leaders can see status at a glance. Effective dashboards use consistent scales and highlight exceptions rather than showing every available number.

A method using control charts to separate common-cause variation inherent in a stable process from special-cause variation triggered by a specific event. It guides whether to adjust the process or investigate a one-off anomaly.

A line graph of a measure over time with a center line, used to spot trends, shifts, and patterns without formal control limits. It is simple enough for frontline teams to build and interpret.

The routine, random fluctuation built into a stable process, produced by many small interacting sources. Reacting to it as if it were a problem typically makes performance worse, not better.

Variation that arises from an identifiable, atypical event rather than the normal working of the process. It signals a need to investigate the specific circumstance instead of redesigning the whole system.

A diagram showing the sequence of steps, decisions, inputs, and handoffs that turn a request into an output. It exposes redundancy, rework loops, and unclear ownership that are invisible in written procedures.

A visual technique that charts material and information flow plus the time spent in each step, separating value-adding activity from waiting and rework. It identifies where delay accumulates across the whole chain.

A process map organized into horizontal or vertical lanes, each assigned to a role, department, or system. Crossings between lanes reveal handoffs, which are frequent sources of delay and error.

The step in a process whose capacity is lowest and therefore limits the throughput of the entire system. Improving any other step produces little overall gain until the constraint is relieved.

A relationship stating that average work in progress equals average arrival rate multiplied by average time in system. It lets teams predict how reducing queues or inflow will change waiting time.

The transfer of work, information, or responsibility from one person, team, or system to another. Each handoff adds wait time and creates opportunity for lost or distorted information.

A structured, small-scale learning loop: plan a change and predicted result, run it, study what actually happened, then act to adopt, adapt, or abandon. Rapid cycles build evidence before large-scale rollout.

An improvement philosophy focused on removing waste, smoothing flow, and delivering value as defined by the customer. It emphasizes respect for people and continuous, incremental change by the people doing the work.

A data-driven methodology that reduces defects by identifying and controlling sources of variation, often using the DMAIC sequence of define, measure, analyze, improve, and control. It suits problems with measurable outcomes and unknown causes.

A practice of continuous, incremental improvement driven by the people who perform the work, often through small team events. It contrasts with occasional large redesigns by making improvement routine.

Any activity or resource use that consumes time or money without adding value the customer would pay for. Common forms include waiting, overproduction, rework, excess motion, and unnecessary transport or inventory.

A structured investigation that traces a problem back to the underlying conditions that allowed it to occur, rather than stopping at the immediate error. Techniques include the five whys and cause-and-effect diagrams.

An iterative questioning technique that asks why repeatedly to move from a surface symptom toward an underlying cause. It works best when each answer is checked against evidence rather than assumed.

A cause-and-effect diagram that arranges possible contributors to a problem into branches such as people, methods, machines, materials, measurement, and environment. It helps a team broaden its hypotheses before testing them.

An approach holding that every system has at least one limiting factor, and that improvement should focus there first. The sequence is to identify the constraint, exploit it, subordinate other steps, elevate it, then repeat.

A documented, current best method for performing a task, including sequence, timing, and quality checks. It provides a stable baseline that makes deviations visible and enables further improvement.

A structured review that identifies who affects or is affected by a change, and assesses their interest, influence, and likely position. It shapes engagement and communication plans for each group.

A documented schedule specifying what will be communicated to which audiences, by whom, through which channel, and how often. It reduces rumor and inconsistency during transitions.

Behavior that slows or blocks an intended change, often rooted in fear of loss, unclear rationale, or past failed initiatives. It is better treated as information to be understood than as disobedience to be overridden.

The degree to which an organization and its people have the awareness, desire, capability, and resources to adopt a change successfully. Assessing it early reveals gaps that can be addressed before rollout.

The shared assumptions, norms, and reward patterns that shape how people actually behave at work. Improvement efforts that conflict with culture tend to revert once attention moves elsewhere.

A shared belief that people can raise concerns, admit mistakes, or question practices without punishment or humiliation. It increases the likelihood that problems surface early enough to be fixed.

The extent to which an improvement persists after the project team disbands and attention shifts. It depends on embedded standard work, ownership, monitoring, and alignment of incentives.

A framework that distinguishes honest error, at-risk behavior, and reckless conduct, responding to each differently. It encourages reporting and learning while still holding people accountable for knowingly unsafe choices.

The obligation to limit access to sensitive information, such as patient or employee data, to those who need it for a legitimate purpose. It governs how improvement data are stored, shared, and reported.

A situation where a person's private interest could improperly influence decisions made in their professional role. It must be disclosed and managed, often by recusal, so that judgments remain credible.

Voluntary agreement by a person to participate after receiving and understanding relevant information about purpose, risks, benefits, and alternatives. It is a core requirement when improvement work also constitutes research.

A limited trial of a change on a small scale or in one setting before wider adoption. It generates evidence about feasibility and side effects while limiting risk if the change underperforms.

The checks that confirm data are complete, accurate, consistent, and drawn from the intended source before analysis. Unvalidated data can make a sound improvement look effective or a good process look broken.

The use of charts, graphs, and layout to make patterns and comparisons easy to grasp. Good design matches the chart to the question and avoids distortion through misleading scales or clutter.

A comparison that quantifies the expected costs of an option against its expected benefits, often over a defined time horizon. It supports prioritization but depends heavily on the assumptions used.

The structured collection of customer needs, expectations, and complaints used to define what counts as quality. It anchors improvement goals in what recipients actually value rather than internal convenience.

The planned development and verification of the knowledge and skills people need to perform a changed process reliably. Without it, new procedures decay into the old habits.

The elapsed time required to complete one unit of work from start to finish of a defined step or process. Reducing it usually shortens waits and lowers work in progress.

The total elapsed time from when a request or order arrives until the output is delivered to the requester. It includes all waiting and rework, so it reflects the customer's actual experience of delay.

A line of work items or people waiting for a resource to become available. Queues grow nonlinearly as utilization rises, so small capacity buffers can sharply cut waiting.

A formal assessment, often by a committee, of whether a proposed study or improvement project protects participants and meets regulatory and ethical standards. It determines whether oversight or consent is required.