Quality Improvement in Surgery: From Good Ideas to Changed Practice

Quality improvement in surgery has historically been conflated with audit — the measurement of current practice against a standard. Audit identifies gaps. Quality improvement closes them. These are related but distinct activities, and the distinction matters because the skills, methods, and change mechanisms required to close a gap are different from those required to identify it. A department that audits excellently but never implements lasting change has invested in measurement without investing in improvement.

The Model for Improvement: the most useful QI framework in clinical practice

The Institute for Healthcare Improvement's Model for Improvement is built around three questions and one iterative testing method. The three questions are: What are we trying to accomplish? (the aim statement — specific, measurable, time-bound, and focused on what matters to patients) How will we know that a change is an improvement? (the measurement plan — outcome, process, and balancing measures that distinguish genuine improvement from gaming) What changes can we make that will result in improvement? (the change ideas — drawn from evidence, from variation analysis, from patient experience data, and from the experience of teams who have already solved the same problem).

The iterative testing method — the Plan-Do-Study-Act cycle — allows small-scale testing of change ideas before system-wide implementation: changing the practice for one patient, one clinic, one day, to learn whether the change produces the intended effect before investing in the resource of implementing it at scale. This learning orientation — treating every implementation as a test from which evidence is gathered, not an intervention that is assumed to work — is the methodological discipline that distinguishes quality improvement from change management by optimistic assumption.

Measurement: the activity that makes QI honest

The measurement plan in quality improvement serves a different purpose from the outcome measurement in research — it is designed to support learning and decision-making in real time rather than to provide statistically valid evidence of effect. Run charts — simple time-series plots of a process or outcome measure tracked over time — provide a visual representation of whether a system is changing in the intended direction, and whether the change is sustained or temporary. Annotating run charts with the dates of specific interventions allows visual attribution of change to specific actions — the most accessible form of improvement evidence available without a trial design.