--- module: 073-03 language: en chapter: 73 title: "Patient Safety, Quality Improvement, Human Factors, and Clinical Systems" module_title: "Improvement methods, teams and escalation, governance, patient partnership, and sustaining change" source_sha256: 2c721eba6327bc870b63b332352255de30bc979aa5093985d3af3af78cc0b741 --- # Quality improvement and safer systems ## Aim and theory of change ### Iterative system change, not motivation alone ### Aim: population, outcome, magnitude, timeframe #### "Improve safety" is too vague ### Ambitious yet realistic target ### Driver diagram links aim to changes #### Theory explains why each change works ### Outcome, process, and balancing measures #### Baseline data before declaring improvement ## Disciplined iterative testing ### Plan-do-study-act: predict, test, compare, adapt #### Learning, not merely doing ### Vary shifts, staff, patients, pressure ### Small tests must not become endless pilots ### Scale needs core elements and ownership #### Different context may cause failure ### Education alone is weak #### Strong: eliminate, force, standardise, simplify #### Weak actions rely on vigilance ## Checklists and bundles ### Checklists for critical, often omitted steps #### Short, timed to workflow, shared pause ### Bundles: small set of evidence-based practices #### Aggregate compliance can hide one omission #### Documentation is not bedside delivery ### Checklist fatigue from overlapping tools ### Routine bypass: examine the rule first ## Medication and diagnostic systems ### Standard concentrations, barcodes, decision support ### Low-value alerts teach overriding ### Track pending results with explicit owners ### Show trends and unresolved abnormalities ### Portals cannot transfer responsibility ### Trigger tools select cases for review ## Teams and escalation ### Briefings, huddles, debriefings ### Graded assertiveness needs safe leader response ### Escalation policy: whom, when, alternatives #### Patients and families can activate review ### Emergency roles and closed-loop orders #### Hot debrief, then later system review ## Sociotechnical change ### Stakeholders affected, controlling, working ### Resistance may be legitimate safety concern #### Listening improves design ### Computerised orders add new error types #### Wrong patient, defaults, copy-forward ### Automation shows uncertainty, allows challenge ### Cybersecurity is patient safety #### Backups, segmentation, usable downtime plans ## Governance for learning and accountability ### Easy, confidential reporting with feedback ### Voluntary reports cannot estimate incidence #### Combine with review, triggers, observation ### Just-culture responses matched to behaviour #### System contribution keeps personal duty ### Boards need data, stories, closed actions ### Documentation can become performative compliance ## Patients as partners ### Patients confirm, but safety stays with service ### Speaking up limited by illness and hierarchy ### Co-design includes those most affected ### Support harmed patients and involved staff #### Staff support must not silence accountability ## Sustain, spread, and evaluate ### Control plans with owners and review dates ### Improvement decays when attention ends #### Embed in orientation, equipment, budgets ### Stratify to avoid widening inequity ### Spread keeps causal core, adapts form #### Adaptation without theory loses active ingredient ## Ethical use of resources ### Quality improvement keeps ethical obligations #### Minimise data, protect confidentiality ### Opportunity cost of new protocols #### Nursing time, delays, work shifted to patients ### De-implement ineffective practices ### Complete when safety persists without extra effort