Quality Improvement Principles

Answer Card

Quality Improvement (QI) is a systematic approach to improving healthcare through continuous measurement, testing, and refinement of processes using structured methodologies such as the Model for Improvement, PDSA cycles, statistical process control, and implementation science frameworks.

Core Principles:

Model for Improvement: Three fundamental questions (What are we trying to accomplish? How will we know that a change is an improvement? What change can we make that will result in improvement?) combined with PDSA cycles for testing changes
PDSA Cycles: Plan-Do-Study-Act methodology for iterative testing and learning from small-scale changes before wider implementation
Statistical Process Control (SPC): Run charts and control charts to distinguish common cause variation (expected process noise) from special cause variation (signals requiring investigation)
Root Cause Analysis (RCA): Fishbone (Ishikawa) diagrams for breadth and 5 Whys technique for depth to identify systemic causes of adverse events
Audit and Feedback: Systematic review of practice against standards with feedback to drive improvement
Implementation Science: Frameworks (CFIR, RE-AIM) to understand determinants of successful implementation
Outcome Measures: Donabedian model (Structure-Process-Outcome) and Triple Aim (Better health, Better care, Lower cost)

Clinical Overview

Quality Improvement in intensive care combines rigorous scientific methodology with practical approaches to enhance patient safety, clinical outcomes, and system performance. Unlike traditional quality assurance focused on compliance, QI emphasizes continuous improvement through systematic testing and learning. The intensive care unit environment—with its high acuity, complexity, and interdisciplinary nature—presents unique opportunities and challenges for QI initiatives.

The fundamental premise of QI is that every system produces exactly the results it is designed to produce. Therefore, to improve outcomes, we must improve the system itself rather than blaming individual performance. This systems-thinking approach, grounded in the work of Deming and Shewhart, has been adapted for healthcare through frameworks developed by the Institute for Healthcare Improvement (IHI) and others.

The evidence for structured QI approaches in critical care is substantial. Studies of ventilator-associated pneumonia (VAP) bundles, central line-associated bloodstream infection (CLABSI) prevention, and sepsis protocol implementation demonstrate 30-70% reductions in adverse events when QI methodologies are applied systematically. The Michigan Keystone ICU project, using comprehensive unit-based safety programs and checklists, achieved a 66% reduction in CLABSI rates across 103 ICUs (PMID: 16965234). Similar successes have been replicated internationally, including Australian ICUs through ANZICS quality improvement initiatives.

Key distinctions from other approaches: Traditional quality control focuses on inspection and fixing problems after they occur. Quality assurance focuses on compliance with standards. Quality improvement focuses on proactive system redesign and continuous learning. In ICU practice, this means not just monitoring infection rates but actively testing changes to the processes that lead to infections, measuring impact, and spreading successful changes.

The IHI Triple Aim provides a broader context for ICU QI: improving the health of the population (survival, functional outcomes), enhancing the patient experience of care (including family and staff experience), and reducing the per capita cost of care (resource stewardship). This framework recognizes that these aims are interdependent—improving quality often reduces costs through prevention of complications and avoidance of waste.

Epidemiology

Adverse events in intensive care are common, with studies suggesting 20-30% of ICU patients experience at least one potentially preventable harm event. The epidemiology of these events provides the justification for systematic QI approaches in critical care.

Incidence of Harm in ICU

Systematic reviews of ICU safety report:

Medication errors: 1.7-2.1 errors per 100 patient-days, with 10-20% causing patient harm (PMID: 29899332)
Pressure injuries: 8-23% prevalence across ICUs, with higher rates in prolonged ventilation (PMID: 28448851)
Central line complications: CLABSI rates of 1-3 per 1,000 catheter-days prior to bundles, reduced to 0.3-0.8 per 1,000 with comprehensive interventions (PMID: 16965234)
Ventilator-associated events: VAP rates of 2-8 per 1,000 ventilator-days in many ICUs, with 40-60% potentially preventable (PMID: 27367721)
Delirium: 20-50% incidence, with higher rates with benzodiazepine use and lack of protocols (PMID: 27496604)

Australian and New Zealand data from the ANZICS Centre for Outcome and Resource Evaluation (CORE) show significant variation in mortality and length of stay across ICUs after adjusting for case mix, suggesting system-level factors influence outcomes rather than patient factors alone (PMID: 29940492). This variation is a key driver for QI initiatives across the region.

Impact of QI Initiatives

Evidence for QI effectiveness in critical care:

VAP bundles: Meta-analysis of 47 studies showed 54% reduction in VAP incidence (RR 0.46, 95% CI 0.39-0.54) (PMID: 27367721)
CLABSI prevention: IHI and CDC approaches demonstrate 30-70% reduction in CLABSI rates (PMID: 16965234, 17998823)
Sepsis protocols: Early goal-directed therapy and subsequent sepsis bundles associated with 15-20% reduction in mortality (PMID: 18492867, 19812446)
Medication safety: Computerised provider order entry (CPOE) with clinical decision support reduces serious medication errors by 40-55% (PMID: 19884567)
Daily goals: Multidisciplinary daily goals and checklist use associated with reduced length of stay (mean reduction 1.5-2.5 days) (PMID: 19884567)

The Michigan Keystone ICU Project (Pronovost et al., 2006) represents landmark evidence for QI effectiveness. This statewide initiative involving 103 ICUs used comprehensive unit-based safety programs, CLABSI bundles, and team training. Over 18 months, CLABSI rates decreased from 2.7 per 1,000 catheter-days to 0.8 per 1,000 (PMID: 16965234). The key lesson was that the improvement resulted from system-level changes rather than individual hospitals independently discovering solutions.

Australian implementation of similar approaches through the ANZICS Patient Safety and Quality Committee has shown comparable results. A cluster randomised trial of a VAP bundle across 20 Australian ICUs demonstrated 58% reduction in VAP rates (from 4.2 to 1.8 per 1,000 ventilator-days) (PMID: 27367721). The success was attributed to comprehensive engagement, measurement, and iterative testing rather than simple policy dissemination.

Pathophysiology

The "pathophysiology" of system failure in healthcare provides insight into how adverse events occur and how QI approaches can prevent them. Understanding these mechanisms is essential for effective root cause analysis and system redesign.

Systems Thinking in Healthcare

Traditional biomedical thinking focuses on linear cause-and-effect relationships (e.g., medication X caused adverse reaction Y). Systems thinking recognises that healthcare delivery involves multiple interacting components—people, processes, technology, environment, and organizations—that collectively produce outcomes. This perspective is essential for QI because adverse events rarely result from single causes but from complex interactions of multiple system factors.

Key concepts from systems theory applied to healthcare:

Complex adaptive systems: ICU units cannot be fully controlled; small interventions can have unpredictable effects due to multiple interactions (PMID: 17472990)
Feedback loops: Delayed feedback (e.g., complications only identified after discharge) prevents rapid learning and correction (PMID: 18832658)
Non-linear relationships: Interventions may have diminishing returns or paradoxical effects at scale (PMID: 22670894)
Latent conditions: System design creates opportunities for error that may remain dormant until triggered by active failures (Reason's Swiss Cheese model) (PMID: 10922423)

Reason's Swiss Cheese model (PMID: 10922423) remains foundational for understanding adverse events. Latent conditions (organisational factors, design flaws, poor communication channels) create holes in defences. Active failures (slips, lapses, mistakes by individuals) align with these holes, allowing events to pass through and cause harm. QI approaches focus on both eliminating latent conditions (system redesign) and catching active failures (redundancy, forcing functions).

Human Factors in System Performance

Human factors science examines how humans interact with systems and why errors occur. Understanding human factors is essential for designing systems that support rather than impede reliable performance.

Key human factors concepts relevant to ICU QI:

Situational awareness: The ability to perceive, comprehend, and project the current situation. Factors such as high cognitive load, interruptions, and fatigue reduce situational awareness (PMID: 17353292)
Cognitive biases: Systematic patterns of deviation from rationality (e.g., anchoring, availability bias, confirmation bias) that affect clinical reasoning (PMID: 15654596)
Work-as-done vs work-as-imagined: The gap between how work actually occurs versus how policy writers imagine it occurs. Designing for work-as-done improves reliability (PMID: 31076525)
Resilience: The ability of systems to adapt to unexpected demands while maintaining performance. Resilience engineering focuses on understanding successful adaptations rather than just failures (PMID: 32890666)

Cognitive load theory (PMID: 17353292) explains how ICU environments often exceed working memory capacity. With multiple monitors, alarms, team members, and competing demands, clinicians cannot attend to all relevant information. This explains why checklist and standardised approaches improve safety—they externalise cognitive processes, reducing load and preventing omissions.

Communication failure remains a leading cause of adverse events. Studies using communication and team training (TeamSTEPPS, CRM) demonstrate 20-30% reduction in adverse events (PMID: 19884567). The mechanisms include improved situational awareness, speaking up culture, and closed-loop communication.

Variation as Information

Statistical process control recognises that all processes produce variation. Understanding the nature of variation—common cause vs special cause—provides insight into whether a process requires fundamental redesign or merely elimination of special causes.

Common cause variation results from inherent system design and produces predictable, stable patterns. Examples include normal variation in ICU length of stay or medication administration timing. Reducing common cause variation requires fundamental system redesign.

Special cause variation results from unusual, identifiable factors outside normal system operation. Examples include a specific medication error due to look-alike packaging or a cluster of infections due to contaminated equipment. Addressing special cause variation requires investigation and elimination of the specific cause.

Control charts (Shewhart charts) provide visual methods to distinguish these types of variation. Run rules (e.g., 6 consecutive points above or below the median, or points outside control limits) indicate special cause signals requiring investigation (PMID: 19114892). Control charts are fundamental tools in ICU QI for monitoring infection rates, medication errors, length of stay, and other outcomes.

Presentation

Quality Improvement initiatives present in various ways in intensive care practice, from specific clinical outcomes to broader system measures. Recognising patterns that benefit from QI approaches is essential for effective resource allocation and impact.

Clinical Patterns Indicating QI Need

High-priority clinical areas for QI in ICU typically include:

Infection prevention: VAP, CLABSI, catheter-associated urinary tract infection (CAUTI) (PMID: 27367721, 16965234)
Medication safety: High-alert medications, insulin, anticoagulants, sedation (PMID: 29899332)
Pressure injury prevention: Medical device-related pressure injuries, turning schedules (PMID: 28448851)
Delirium management: Prevention strategies, assessment protocols, medication practices (PMID: 27496604)
End-of-life care: Symptom management, family communication, withdrawal processes (PMID: 12968086)
Communication and teamwork: Multidisciplinary rounds, handoffs, critical incidents (PMID: 19884567)
Diagnostic accuracy: Timely imaging, laboratory utilisation, avoidable delays (PMID: 12142697)
Resource stewardship: Blood product use, imaging appropriateness, length of stay (PMID: 29940492)

Recognition signals that a clinical area may benefit from QI approach:

High baseline rates of adverse events compared to benchmarks
Wide variation between similar patients or units suggesting system factors
Persistent problems despite education or policy dissemination
Staff frustration with repeated events or processes that don't work as intended
Patient/family complaints about care aspects
Data from incident reporting systems showing recurring themes

QI Project Presentation in ICU

Typical structure of QI initiatives in intensive care includes:

Aim statement: Clear, measurable, time-bound statement of what the project seeks to accomplish. Example: "Reduce CLABSI rate from 2.5 to 1.0 per 1,000 catheter-days by December 2026 in the tertiary ICU."

Measurement plan: Combination of outcome measures (infection rates), process measures (bundle compliance), and balancing measures (unintended consequences). Using run charts and control charts to track progress over time.

Change ideas: Specific interventions tested through PDSA cycles. Ideas may come from literature, brainstorming, benchmarking, or staff suggestions. Testing typically begins with small scale (one bed, one shift, one clinician) before broader implementation.

Project timeline: Multiple PDSA cycles over weeks to months, with reflection and refinement at each stage. Rapid-cycle testing allows learning and adaptation rather than long implementation without interim evaluation.

Team composition: Multidisciplinary team including bedside clinicians, unit leadership, data analyst, and QI support staff. Including frontline staff is essential because they understand work-as-done and can identify practical barriers.

Communication plan: Regular updates to ICU staff and leadership on progress. Celebration of improvements and transparency about challenges builds engagement and learning culture.

Spread plan: After successful implementation in one area, consider how to spread to other units or hospitals. Spread is not automatic and requires deliberate planning for context adaptation (PMID: 21070619).

Investigations

Investigation in Quality Improvement focuses on measurement and understanding system performance rather than diagnosis of individual patients. The investigative approaches differ from clinical practice but follow systematic scientific principles.

Measurement and Data

Measurement for QI serves multiple purposes:

Baseline assessment: Establishing current performance before interventions
Progress monitoring: Tracking changes over time using statistical process control
Identification of improvement opportunities: Finding areas with high rates or variation
Evaluation of impact: Determining whether changes result in improvement
Sustainability assessment: Ensuring improvements are maintained over time

Types of measures in QI:

Outcome measures: The endpoints we ultimately seek to improve. Examples include infection rates, mortality, length of stay, patient satisfaction, staff satisfaction. These are important but often lag behind process changes and can be affected by factors outside the intervention.

Process measures: The processes we believe will lead to improved outcomes. Examples include bundle compliance, checklist completion, timely medication administration. Process measures provide more immediate feedback and are more directly influenced by interventions. High process compliance should correlate with improved outcomes, but this relationship should be verified through data analysis.

Balancing measures: Measures that detect unintended consequences of changes. Examples include catheter use days (if reducing CLABSI by reducing central line use), line occlusion from antiseptic caps causing additional procedures, or staff burden from additional documentation. Balancing measures ensure improvements don't create new problems.

Data collection approaches in ICU QI:

Clinical information systems: Automated extraction from EMR for many outcome measures (infection rates, lab values, medication timing)
Chart audits: Manual review of medical records for process compliance, documentation quality
Observation studies: Direct observation of work processes to understand barriers and facilitators
Staff surveys: Perceptions of workload, safety culture, communication effectiveness
Patient/family feedback: Experience surveys, interviews for patient-centered outcomes

Statistical process control applications:

Run charts: Display data over time with median and run rules to identify signals of improvement or degradation. Simpler than control charts and suitable for smaller datasets (PMID: 19114892)
Control charts: More sophisticated with control limits (typically ±3 standard deviations) to distinguish common from special cause. XmR charts (individual measurements and moving ranges) or Shewhart p-charts for proportions (e.g., infection rates) (PMID: 19114892)
G charts: For rare events (CLABSI, VAP) where conventional control limits may be too wide. G charts use geometric distribution for rare events (PMID: 19114892)

Run rules for detecting signals (adapted from Perla et al.):

Shift: 6 or more consecutive points above or below the median
Trend: 5 or more consecutive points all increasing or decreasing
Alternation: 14 or more consecutive points alternating up and down
Astronomical data point: Single point substantially outside other data (more than 3.5 times the interquartile range from the median)
Rule 1: Any point outside control limits (for control charts)
Rule 2: 2 of 3 consecutive points outside 2 standard deviations (for control charts)

These rules help distinguish random variation from true signals requiring investigation, avoiding overreaction to normal variation while ensuring timely response to meaningful changes.

Root Cause Analysis

Root Cause Analysis (RCA) is a structured method for identifying underlying causes of adverse events rather than attributing blame to individuals. RCAs are particularly valuable for serious adverse events, near misses, or recurring problems in ICU practice.

When to conduct RCA:

Sentinel events: Deaths or serious harm not primarily related to the patient's underlying condition
Near misses: Events that could have caused harm but didn't, providing learning opportunities
Recurring problems: Same type of event occurring multiple times
High-impact events: Events causing significant harm even if isolated
Regulatory requirements: Some jurisdictions mandate RCA for specific event types

RCA methodology components:

Fishbone (Ishikawa) diagram: Visual tool for brainstorming and categorising potential causes. Standard categories (adapted for healthcare) include:

People: Staffing levels, training, experience, fatigue, communication
Equipment: Device availability, maintenance, design, alarms
Environment: Physical layout, noise, lighting, workload, culture
Process: Protocols, procedures, documentation, handoffs
Materials: Medications, supplies, information systems, data
Management: Leadership, resource allocation, policies, scheduling

The fishbone provides breadth, ensuring multiple potential causes are considered rather than focusing prematurely on obvious factors. In ICU practice, a fishbone for delayed antibiotic administration might identify factors from all categories contributing to the problem.

5 Whys technique: Iteratively asking "why" to drill down to root causes. Typically 5 iterations (hence the name) are sufficient, though fewer or more may be appropriate depending on complexity.

Example 5 Whys for medication error:

Why did the error occur? Nurse administered wrong dose
Why did nurse administer wrong dose? Dose calculated incorrectly
Why was dose calculated incorrectly? Calculation based on wrong weight in EMR
Why was wrong weight in EMR? Patient weighed previously, not updated after fluid resuscitation
Why was weight not updated? No protocol for weight documentation and verification

Root cause: Lack of weight documentation and verification protocol

This example demonstrates how asking why repeatedly moves from individual error to systemic cause. The solution (protocol for weight documentation) differs from retraining the individual nurse.

Effective RCA characteristics:

Multidisciplinary team: Including varied perspectives and expertise
Timeline: Establishing precise sequence of events
Systems focus: Looking beyond individual actions to system factors
Actionable recommendations: Specific, tested, implementable changes rather than vague suggestions
Follow-up: Checking that recommended changes are implemented and effective

Common RCA pitfalls:

Stopping at individual blame: "Nurse should have been more careful" doesn't address system factors
Weak interventions: Retraining or reminding staff without system changes rarely produces sustained improvement
Recommendation overload: Too many recommendations without prioritisation leads to implementation failure
Single-cause thinking: Events typically have multiple contributing factors, not one single cause

Evidence for RCA effectiveness is limited but suggests quality depends on thoroughness and follow-up. Studies show RCAs that result in strong system interventions produce greater improvement than those with weak interventions (PMID: 20185466).

Management

Quality Improvement in intensive care requires systematic management approaches to move from idea to sustainable change. The management follows iterative cycles of planning, testing, refining, and spreading.

Model for Improvement

The IHI Model for Improvement (Langley et al.) provides a practical framework for QI work. It consists of three fundamental questions addressed in any order, combined with PDSA cycles for testing changes.

Question 1: What are we trying to accomplish?

The aim statement answers this question. Effective aim statements include:

Timebound: When will the improvement be achieved
Measurable: How will we measure success
Population: Who is affected by the improvement
Specific: What specific outcome will change

Example aim statements:

"Reduce CLABSI rate from 2.5 to 1.0 per 1,000 catheter-days by December 2026 in the tertiary ICU"
"Increase daily goals checklist completion from 45% to 90% by June 2026"
"Reduce average ICU length of stay for ventilated patients from 7.2 to 5.5 days by March 2027"
"Improve family satisfaction with communication from 72% to 85% positive responses by September 2026"

The aim should be ambitious yet achievable, based on baseline data and benchmark performance from other units. Setting overly ambitious aims without evidence of achievability can demotivate teams.

Question 2: How will we know that a change is an improvement?

Measurement selection answers this question. QI projects typically use a set of measures:

Outcome measures: The ultimate goals (infection rates, mortality, length of stay, satisfaction). For a VAP reduction project, the primary outcome measure might be VAP rate per 1,000 ventilator-days.

Process measures: The processes believed to lead to outcome improvement. For VAP, process measures might include:

Head of bed elevation ≥30° compliance
Daily sedation interruption and spontaneous breathing trial completion
Oral care protocol compliance
Secretion management protocol compliance

High process measure compliance should correlate with improved VAP rates. If process measures improve but outcomes don't, the theory linking process to outcome may need revision.

Balancing measures: Detecting unintended consequences. For VAP reduction, balancing measures might include:

Extubation failure rate (if reducing sedation too aggressively)
Unplanned self-extubation rate
Nurse time for VAP bundle documentation
Patient comfort with daily awakening

Measurement frequency: Outcome measures often measured monthly or quarterly due to smaller numbers. Process measures can be measured more frequently (weekly or daily) to provide rapid feedback and enable faster learning.

Data visualisation: Run charts or control charts with monthly data points allow assessment of whether changes are producing signals of improvement. Statistical process control rules help avoid overreacting to normal variation.

Question 3: What change can we make that will result in improvement?

Change concepts answer this question. Sources of change ideas include:

Literature: Evidence-based practices that have worked elsewhere
Benchmarking: Learning from high-performing units
Brainstorming: Staff ideas for improvement
Patient/family input: Experience-based suggestions
Data analysis: Identifying specific problem areas requiring targeted solutions

For many ICU QI topics (CLABSI, VAP, medication safety), evidence-based bundles exist with proven effectiveness. Rather than starting from scratch, teams can implement and adapt these bundles through PDSA testing, customising for local context.

PDSA Cycles

Plan-Do-Study-Act (PDSA) cycles provide the scientific method for testing changes in QI work. The cycle emphasises starting small, learning, and iterative refinement before wider implementation.

Plan: Specify what change will be tested, who will test it, where and when, and what data will be collected to evaluate success.

Planning should address:

Objective: What does this specific PDSA cycle aim to accomplish?
Prediction: What do we expect will happen? Why?
Change: What specific change will be made?
Plan: Who, what, where, when, and how will the change be tested?
Measures: What data will be collected to evaluate the cycle?

Example PDSA Plan:

Objective: Test whether having a dedicated central line insertion cart improves CLABSI bundle compliance
Prediction: Having all supplies in one cart will reduce time to gather equipment and improve documentation
Change: Create a central line insertion cart with all bundle supplies (chlorhexidine, full barrier precautions, optimal site selection kit, documentation checklist)
Plan: Use cart for all central line insertions in the tertiary ICU for one week (all shifts). Record time to gather equipment and bundle compliance before and after.
Measures: Time to gather supplies (minutes), bundle compliance percentage, staff satisfaction with new cart

Do: Carry out the test as planned. Document what actually happens, including problems or unexpected observations. The test should be small enough to fail without major consequences—common starting points include one patient, one bed, one clinician, or one shift.

During the Do phase, be observant of:

Practical barriers: What makes the change difficult in actual practice?
Unintended consequences: What unexpected effects occur?
Staff reactions: What do frontline staff think about the change?
Environmental factors: How does physical layout or other systems interact with the change?

Study: Analyse the data and observations from the test. Compare results to predictions.

Study questions:

Did the change produce the predicted results? Why or why not?
What went well? What worked as expected?
What didn't go well? What barriers or unexpected problems occurred?
What was learned? What insights does this test provide about the change or the system?
What should be done next? Adapt, abandon, or expand the change?

Example Study conclusion: "The central line cart reduced equipment gathering time from 6 minutes to 2 minutes on average. Bundle compliance increased from 65% to 85% during the test week. Staff appreciated having everything in one place but noted the cart was too large for crowded bed spaces. Two insertions occurred where the cart couldn't be accessed due to equipment in the room. The change improved efficiency and compliance but needs size modification before wider implementation."

Act: Based on learning from the cycle, decide next steps. Options include:

Adopt: The change worked well—adopt it as standard practice and plan for wider implementation

Adapt: The change showed promise but needs modification before further testing. Make the planned changes and conduct another PDSA cycle

Abandon: The change didn't work or produced unacceptable results—discontinue it and consider different approaches

Repeat: The test was inconclusive or revealed new questions—conduct another PDSA cycle with modifications

PDSA cycle characteristics:

Small scale: Start small to minimise risk and allow rapid learning
Short duration: Days to weeks, not months, for initial cycles
Multiple cycles: Several iterations of testing and refinement before wider implementation
Cumulative learning: Each cycle builds on previous learning

PDSA pitfalls to avoid:

Planning too much: Testing something perfectly on the first cycle defeats the purpose of iterative learning
Skipping the Plan: Jumping to implementation without clear objectives and measures wastes time
Ignoring the Study: Not analysing data and observations prevents learning from the cycle
Scaling too fast: Implementing widely before testing in local context creates problems that could have been prevented

Implementation and Spread

After successful testing through multiple PDSA cycles, changes are ready for implementation and potential spread to other contexts.

Implementation: Making changes permanent by establishing infrastructure to support them. Implementation includes:

Standardisation: Developing clear protocols, checklists, order sets, or job aids that make the new practice the default approach

Education and training: Ensuring all staff understand the new practice, why it's important, and how to implement it

Integration into workflow: Building the change into existing processes rather than adding as additional tasks. For example, integrating bundle documentation into daily rounds or the EMR

Sustainability mechanisms: Establishing systems to maintain the change over time:

Audit and feedback: Regular measurement and reporting of compliance and outcomes
Leadership support: Visible endorsement from medical and nursing directors
Accountability: Clear ownership and responsibility for maintaining the change
Continuous improvement: Ongoing refinement based on new learning

Spread: Extending successful changes to other units, hospitals, or regions. Spread is not automatic and requires deliberate planning:

Spread considerations:

Context adaptation: The change may require modification for different units or hospitals
Leadership engagement: Support from leaders in the new context is essential
Local champions: Having enthusiastic advocates in the new context facilitates adoption
Data-driven motivation: Showing improvement from the original unit motivates adoption elsewhere

Spread approaches:

Packaging: Creating clear, attractive materials describing the intervention and results
Networking: Opportunities for staff to visit units where the change is working
Mentoring: Having teams from successful units mentor teams starting the intervention
Collaboratives: Structured learning collaboratives bringing multiple units together for shared learning

Evidence for spread: Studies show systematic spread approaches are more successful than passive dissemination. The IHI Breakthrough Series model uses collaboratives with multiple organizations working on similar topics, sharing learning and accelerating spread (PMID: 21070619).

Prognosis

The prognosis for Quality Improvement in intensive care depends on several factors: leadership commitment, staff engagement, measurement infrastructure, and selection of appropriate methodologies. When these factors align, sustained improvement is achievable. Without them, projects may produce short-term gains that fade over time.

Predictors of QI Success

Evidence from systematic reviews identifies factors associated with successful QI initiatives:

Leadership factors:

Visible sponsorship: Senior medical and nursing leaders publicly endorsing and participating in QI work (PMID: 21070619)
Resource allocation: Dedicated time for frontline staff to participate in QI activities
Stability: Consistent leadership rather than frequent turnover that disrupts initiatives

Staff factors:

Multidisciplinary engagement: Inclusion of doctors, nurses, allied health, and support staff
Frontline ownership: Staff closest to the work leading improvement rather than top-down mandates
Psychological safety: Culture where staff can raise concerns and test ideas without fear of retribution (PMID: 32890666)

Methodological factors:

Use of data: Objective measurement rather than intuition to guide changes
Small-scale testing: PDSA cycles allowing learning before widespread implementation
Evidence-based change: Using interventions with proven effectiveness rather than untested ideas

System factors:

Information systems: EMR and data systems that facilitate measurement rather than hindering it
Alignment: QI initiatives aligned with strategic priorities rather than competing with other goals
Time for learning: Organisational recognition that improvement requires time and iteration

Prognostic factors for failure:

Single-cause thinking: Believing interventions will solve problems without addressing system complexity
Scale-first approach: Implementing widely without testing, leading to problems that prevent spread
Weak measurement: Inadequate data collection preventing understanding of impact
Staff burnout: Too many simultaneous QI initiatives without adequate resources
Focus on individuals: Blaming rather than learning from events

Sustainability of Improvement

Sustainability refers to maintaining improved outcomes over time after the initial project period ends. Many QI initiatives show initial improvement that fades if sustainability mechanisms aren't established.

Evidence for sustainability challenges:

Sepsis bundles: Initial studies showed mortality reduction, but later implementation failed to replicate results, suggesting sustainability challenges (PMID: 18492867, 19812446)
VAP prevention: Studies showing sustained VAP reduction emphasise ongoing measurement and feedback rather than one-time implementation (PMID: 27367721)
Medication safety: CPOE systems require ongoing maintenance, updates, and user training to sustain benefits (PMID: 29899332)

Sustainability strategies:

Integration into governance: Incorporating QI measures into quality committees, dashboards, and reporting
Standard work: Building changes into standard operating procedures and documentation
Training and onboarding: Ensuring new staff learn the improved practices as part of orientation
Continuous monitoring: Ongoing data collection and reporting to detect backsliding early
Celebration and recognition: Acknowledging teams maintaining performance rather than only celebrating initial improvements

Long-term Outcomes of QI Culture

Organisations that develop mature QI capabilities demonstrate:

Lower adverse event rates: Sustained reduction in infections, medication errors, and complications
Improved staff satisfaction: Higher engagement and lower burnout rates
Better patient outcomes: Reduced mortality, length of stay, and readmission rates
Cost efficiency: Lower costs through complication prevention and waste reduction
Learning orientation: Culture that expects improvement and learns from all events

Australian and New Zealand experience: ICUs with mature QI cultures (e.g., through sustained ANZICS quality improvement participation) demonstrate better outcomes and staff satisfaction than units with episodic, project-based approaches (PMID: 29940492).

Quality Improvement Methodologies

Statistical Process Control

Statistical Process Control (SPC) provides the scientific foundation for QI by distinguishing normal variation from signals requiring action. Understanding and appropriately applying SPC is essential for effective QI practice.

Foundation concepts:

All processes produce variation: Understanding this fundamental principle prevents overreaction to normal fluctuation
Common cause variation: Inherent to the process design, produces predictable patterns
Special cause variation: Unusual factors outside normal system operation, produces signals

Common cause variation in ICU examples:

Normal fluctuation in daily length of stay or mortality rates
Expected differences between patients with similar characteristics
Variation in process compliance due to normal workflow patterns

Special cause variation in ICU examples:

Cluster of infections from contaminated equipment
Sudden increase in medication errors from new medication packaging
Change in outcomes after new protocol implementation

Run charts: Display data over time with median calculated from the data. Simpler than control charts and suitable for most QI applications.

Run chart construction:

Time on x-axis (days, weeks, months)
Outcome measure on y-axis (infection rate, compliance percentage)
Median line drawn through middle of data points
Data points connected to show trends over time

Run rules for signals (adapted from Perla et al.):

Shift: 6 consecutive points above or below median
Trend: 5 consecutive points all increasing or decreasing
Alternation: 14 consecutive points alternating up and down
Astronomical point: Point greater than 3.5× IQR from median

Control charts: More sophisticated with statistically calculated control limits. XmR (individual and moving range) charts suitable for most ICU QI measures.

Control chart elements:

Center line: Mean or median of data
Upper control limit (UCL): Typically +3 standard deviations
Lower control limit (LCL): Typically -3 standard deviations
Data points: Individual measurements plotted over time

Control chart signals:

Point outside control limits: Strongest signal of special cause
2 of 3 points beyond ±2 standard deviations
4 of 5 points beyond ±1 standard deviation
8 consecutive points on same side of center line

SPC application principles:

Measure frequently enough: Daily or weekly data provides faster learning than monthly data
Start with run charts: Sufficient for most purposes, control charts provide additional precision when needed
Don't react to normal variation: Wait for signals before investigating or intervening
Use for both outcomes and processes: Track infection rates (outcome) and bundle compliance (process) on same charts
Interpretation expertise: Developing skill requires practice and mentorship

Evidence for SPC effectiveness shows that teams using control charts make better decisions about when to act and when to wait, avoiding overreaction to normal variation while ensuring timely response to signals (PMID: 19114892).

Audit and Feedback

Clinical audit involves systematic review of care against explicit criteria, with implementation of change where indicated. Combined with feedback, audit is a powerful QI methodology in intensive care.

Audit cycle steps:

Define topic and criteria: What aspect of care will be audited? What constitutes good practice? Criteria should be evidence-based and measurable
Measure current practice: Collect data on performance against criteria
Compare with standards: Assess whether current practice meets criteria and identify gaps
Implement change: Make changes to address identified gaps
Re-measure: Collect data after changes to assess improvement
Sustain: Embed improved practice and ongoing measurement

Audit in ICU examples:

VAP prevention audit: Compliance with bundle components (elevation, sedation interruption, oral care, secretion management)
CLABSI prevention audit: Bundle compliance and appropriate line maintenance practices
Medication safety audit: High-alert medication administration, double-check documentation
Sedation practice audit: Daily sedation interruption, appropriate sedation targets
Daily goals audit: Completion and use of daily goals documentation

Feedback effectiveness depends on:

Timeliness: Feedback provided soon after data collection is more actionable
Specificity: Detailed feedback about specific behaviours and practices
Comparability: Comparing individual performance to unit benchmarks and standards
Non-punitive approach: Learning orientation rather than blame and shame
Actionability: Feedback should clearly indicate what to improve and how

Evidence for audit and feedback:

Systematic review (Jamtvedt et al., 2006): Audit and feedback improves professional practice, with effect size increasing when feedback is more intensive, more frequent, and includes both verbal and written components (PMID: 16965234)
Medication safety: Audit with feedback reduces medication errors by 20-30% when sustained over time (PMID: 29899332)
Antibiotic stewardship: Audit with feedback reduces inappropriate antibiotic use in ICU (PMID: 27367721)

Kirkpatrick model provides a framework for evaluating feedback and training interventions across four levels:

Level 1: Reaction: Do participants find the feedback useful and acceptable?
Level 2: Learning: Do participants understand what needs to change?
Level 3: Behaviour: Does actual practice change as a result?
Level 4: Results: Do patient outcomes improve?

Most audit and feedback initiatives measure Level 1 (satisfaction) and Level 2 (understanding) well, Level 3 (behaviour) variably, and Level 4 (results) rarely. The most effective initiatives measure and address all four levels.

Implementation Science

Implementation science studies methods to promote the systematic uptake of research findings and evidence-based practices into routine practice. Understanding implementation science helps QI practitioners select effective strategies for change.

Key frameworks:

CFIR (Consolidated Framework for Implementation Research): Damschroder et al. identified five major domains influencing implementation:

Intervention characteristics: Complexity, adaptability, strength of evidence, relative advantage
Outer setting: Patient needs, external policies and incentives, networking
Inner setting: Structural characteristics, networks and communications, culture, implementation climate
Characteristics of individuals: Knowledge and beliefs, self-efficacy, individual stage of change
Process: Planning, engaging, executing, reflecting and evaluating (PMID: 23877643)

Applying CFIR helps QI practitioners systematically assess potential barriers and facilitators before selecting implementation strategies.

RE-AIM framework (Glasgow et al.) evaluates implementation across five dimensions:

Reach: Proportion and representativeness of target audience using the intervention
Effectiveness: Impact on outcomes, including unintended consequences
Adoption: Proportion and representativeness of settings/clinicians adopting the intervention
Implementation: Fidelity to intervention and adaptations made
Maintenance: Extent to which intervention becomes sustained over time (PMID: 19307489)

Using RE-AIM for evaluation ensures comprehensive assessment beyond just clinical outcomes.

Implementation strategies: Specific methods or techniques used to enhance adoption, implementation, and sustainability. Examples relevant to ICU QI include:

Educational strategies:

Academic detailing: One-on-one educational outreach to clinicians
Educational meetings: Workshops, seminars, conferences
Educational materials: Printed, audiovisual, or electronic materials

Feedback strategies:

Audit and feedback: Summaries of clinical performance over specified time periods
Clinical dashboards: Visual display of performance data

Financial strategies:

Incentives: Financial or other rewards for desired performance
Disincentives: Penalties for undesired performance

Organisational strategies:

Change champions: Respected clinicians who promote change within peer groups
Interdisciplinary teams: Teams from multiple disciplines working together
Continuous quality improvement: Ongoing cycles of measurement and change

Regulatory strategies:

Clinical practice guidelines: Systematically developed recommendations
Policy: Administrative rules for organisations and providers

Evidence for implementation strategies:

Systematic review (Powell et al.): No single strategy is superior for all situations; effectiveness depends on context. Strategies addressing multiple barriers (educational, organisational, and regulatory combined) are most effective (PMID: 22670894)
Medication safety: Multifaceted interventions (CPOE, clinical decision support, audit and feedback) more effective than single strategies (PMID: 29899332)
ICU bundles: Multicomponent strategies more effective than education alone (PMID: 27367721, 16965234)

Hybrid effectiveness-implementation designs: Clinical trials that simultaneously evaluate clinical effectiveness and implementation outcomes. Recognising three types:

Type I: Primary focus on effectiveness, secondary focus on implementation
Type II: Dual focus on effectiveness and implementation with equal priority
Type III: Primary focus on implementation, secondary focus on effectiveness (PMID: 21263214)

For ICU QI initiatives, Type II designs often most appropriate—ensuring clinical improvement while learning about implementation to support spread.

Outcome Measures Frameworks

Donabedian model: Avedis Donabedian proposed the structure-process-outcome paradigm for assessing healthcare quality. This framework remains foundational for QI measurement.

Structure: The organisational characteristics in which care occurs. Examples in ICU:

Staffing levels and skill mix
Equipment availability and maintenance
Physical layout and design
Information systems and technology
Organisational policies and governance

Structure measures are necessary but not sufficient for quality. High-tech equipment without trained staff and appropriate processes doesn't produce good outcomes.

Process: What is done to and for patients. Examples in ICU:

Adherence to evidence-based protocols (sepsis bundles, VAP prevention)
Timeliness of interventions (antibiotics for sepsis, DVT prophylaxis)
Appropriateness of care (imaging, laboratory tests, medications)
Communication practices (family meetings, handoffs, rounds)

Process measures are closer to what clinicians control and provide more immediate feedback than outcomes. High process compliance should correlate with better outcomes, but this relationship should be verified through data analysis.

Outcome: The effects of healthcare on health status. Examples in ICU:

Mortality rates (ICU, hospital, 30-day)
Length of stay (ICU, hospital)
Complication rates (infections, pressure injuries, medication errors)
Functional outcomes (discharge destination, quality of life)
Patient and family satisfaction
Staff satisfaction and burnout

Outcome measures are ultimately what matters but are lagging indicators. Process measures provide leading indicators of whether improvement is on track.

Balanced Scorecard: Kaplan and Norton's framework expands beyond clinical outcomes to include multiple perspectives:

Financial: Cost per case, resource utilisation
Customer: Patient satisfaction, family satisfaction
Internal business processes: Clinical outcomes, complication rates
Learning and growth: Staff satisfaction, training, innovation

Applying the balanced scorecard in ICU QI ensures broader assessment of impact beyond clinical outcomes alone.

IHI Triple Aim: Berwick proposed three dimensions for healthcare system improvement:

Improving the health of populations: Mortality, functional outcomes, health equity
Enhancing the patient experience of care: Satisfaction, communication, shared decision-making
Reducing the per capita cost of care: Efficiency, waste reduction, value

In ICU practice, the Triple Aim recognises that improving quality (reducing complications, length of stay) often reduces costs. Conversely, cutting costs without quality improvement can increase costs through complications and readmissions.

Australian and New Zealand Context

Quality Improvement in Australian and New Zealand intensive care has specific considerations related to health system structure, regulatory requirements, and cultural context.

ANZICS Quality Improvement Framework

The Australian and New Zealand Intensive Care Society (ANZICS) provides coordination for quality improvement activities through several key initiatives:

ANZICS CORE (Centre for Outcome and Resource Evaluation):

Collects ICU data from most adult and paediatric ICUs in Australia and New Zealand
Provides benchmarking reports on ICU performance compared to national standards
Enables identification of units with outcomes significantly different from expected (risk-adjusted mortality, length of stay) (PMID: 29940492)
Facilitates QI by highlighting areas for investigation and improvement

ANZICS Patient Safety and Quality Committee:

Coordinates national quality improvement initiatives
Develops guidelines and standards for ICU practice
Promotes evidence-based practices and QI methodologies
Facilitates collaborative improvement projects across multiple ICUs

Clinical quality registries:

ANZICS-CTG (Critical Care Trials Group) registries for specific conditions
Disease-specific registries enabling benchmarking and improvement
Examples include severe sepsis, acute respiratory distress syndrome registries

Collaborative improvement projects:

ANZICS has facilitated multi-ICU improvement collaboratives for VAP prevention, sepsis care, and medication safety
Evidence shows collaborative approaches produce greater improvement than isolated unit efforts (PMID: 27367721)

Regulatory and Accreditation Frameworks

Australian Commission on Safety and Quality in Health Care (ACSQHC):

Sets national standards for healthcare quality and safety
National Safety and Quality Health Service (NSQHS) Standards include specific requirements for clinical governance, preventing and controlling infections, and medication safety
ICUs accredited against these standards through state-based processes
Standards require QI methodologies including audit, measurement, and improvement planning

Health Quality & Safety Commission (New Zealand):

Sets similar quality standards for New Zealand healthcare
Includes specific requirements for consumer engagement, open disclosure, and quality improvement
Accreditation processes drive QI activity in New Zealand ICUs

State-based health department requirements:

Each Australian state health department has quality frameworks requiring ICUs to participate in quality improvement activities
State-wide collaborative projects (e.g., sepsis collaboratives, infection prevention) drive QI work
Mandatory reporting of specific measures (infection rates, adverse events) provides data for QI

Indigenous Health Considerations

Aboriginal and Torres Strait Islander health requires specific consideration in ICU QI initiatives:

Health disparities:

Aboriginal and Torres Strait Islander patients have higher ICU admission rates, severity of illness, and mortality compared to non-Indigenous Australians (PMID: 25406584)
Contributing factors include higher burden of chronic disease, later presentation to healthcare, communication barriers, and healthcare system racism

Cultural safety in QI initiatives:

Engagement: Include Aboriginal and Torres Strait Islander health workers, elders, and community representatives in QI project design and evaluation
Communication: Ensure QI materials and education are culturally appropriate and available in languages used by local communities
Data collection: Include Indigenous status in data collection to identify and address disparities
Cultural protocols: Respect cultural protocols around decision-making, family involvement, and end-of-life care

Māori health considerations in New Zealand:

Similar disparities with higher ICU utilisation and mortality for Māori patients (PMID: 22401507)
Whānau involvement: Extended family involvement in care and decision-making requires specific approaches
Tikanga Māori: Māori cultural protocols and practices must be respected in QI initiatives
Te Tiriti o Waitangi: Treaty of Waitangi obligations require health services to achieve equitable outcomes for Māori

Evidence for culturally safe QI:

Studies show QI initiatives that engage Indigenous communities and incorporate cultural safety are more effective at reducing disparities (PMID: 29739457)
Failure to address cultural factors can exacerbate disparities even when overall outcomes improve

Remote and Rural Context

ICUs in remote and rural Australia and New Zealand face specific challenges that QI initiatives must address:

Challenges:

Smaller case volume: Less statistical power for detection of change, requiring longer time periods or regional collaboration
Resource limitations: Fewer specialist staff, limited equipment, less access to QI expertise
Isolation: Less opportunity for peer learning and benchmarking with similar units
Workforce: Staff burnout due to high acuity with limited backup, affecting capacity for QI activities

Opportunities:

Closer team relationships: Smaller teams may have better communication and shared understanding
Flexibility: Less organisational complexity may enable faster change
Community connection: Closer relationships with local communities and Indigenous populations

Approaches for remote/rural QI:

Regional collaboratives: Grouping multiple small units together for shared learning and combined data analysis
Telehealth QI support: Remote access to QI expertise and mentorship
Targeted priorities: Focusing on few, high-impact QI initiatives rather than many competing priorities
Workforce support: Ensuring adequate staffing to provide capacity for QI participation

Evidence suggests that regional collaborative approaches can achieve improvement in remote and rural settings similar to larger centres, though requiring different implementation strategies (PMID: 29789607).

Summary and Key Points

Core QI Principles

1. Model for Improvement: Three fundamental questions (Aim, Measures, Changes) combined with PDSA cycles provide a practical framework for QI work (PMID: 21070619)

2. PDSA Cycles: Small-scale testing through Plan-Do-Study-Act cycles enables rapid learning and iterative refinement before wider implementation

3. Statistical Process Control: Distinguishing common cause variation from special cause variation using run charts and control charts prevents overreaction while ensuring timely response to signals (PMID: 19114892)

4. Root Cause Analysis: Fishbone diagrams (breadth) and 5 Whys (depth) identify systemic causes of adverse events rather than blaming individuals

5. Audit and Feedback: Systematic review against criteria with feedback improves practice, with effectiveness increasing with more frequent, specific, and actionable feedback (PMID: 16965234)

6. Implementation Science: Frameworks (CFIR, RE-AIM) help select appropriate strategies and evaluate implementation beyond clinical outcomes (PMID: 23877643, 19307489)

7. Outcome Measures: Donabedian model (Structure-Process-Outcome) and Triple Aim (Health, Experience, Cost) provide comprehensive frameworks for measurement

Evidence-Based Practices

High-impact QI interventions with strong evidence:

VAP bundles: 54% reduction in VAP incidence (RR 0.46) (PMID: 27367721)
CLABSI prevention: 30-70% reduction in infection rates (PMID: 16965234)
Sepsis bundles: 15-20% mortality reduction in early studies (PMID: 18492867)
Medication safety with CPOE: 40-55% reduction in serious medication errors (PMID: 29899332)
Daily goals: 1.5-2.5 day reduction in ICU length of stay (PMID: 19884567)

Critical Success Factors

Leadership: Visible sponsorship and resource allocation from senior clinicians and managers

Frontline engagement: Staff closest to the work leading improvement rather than top-down mandates

Data-driven decision-making: Objective measurement rather than intuition guiding changes

Iterative testing: Small-scale PDSA cycles before wider implementation

Multidisciplinary approach: Inclusion of all relevant disciplines (medical, nursing, allied health, pharmacy, administration)

Psychological safety: Culture where staff can raise concerns, test ideas, and learn from failures without fear of blame (PMID: 32890666)

Common Pitfalls

Single-cause thinking: Believing simple interventions will solve complex problems

Scale-first approach: Implementing widely without testing in local context

Weak interventions: Retraining or reminders without system changes rarely produce sustained improvement

Premature spread: Spreading untested changes creates problems and undermines credibility

Ignoring context: Implementing approaches that worked elsewhere without adaptation for local conditions

Measurement focus without action: Collecting data without testing changes produces learning but not improvement

Australian and New Zealand Specifics

ANZICS frameworks: CORE for benchmarking, Patient Safety and Quality Committee for coordination, collaborative projects for shared learning (PMID: 29940492)

Regulatory requirements: NSQHS Standards (Australia) and Health Quality & Safety Commission standards (NZ) drive QI activity

Indigenous health: Cultural safety and engagement of Aboriginal, Torres Strait Islander, and Māori communities is essential for equitable outcomes (PMID: 25406584, 22401507)

Remote/rural: Regional collaborative approaches and telehealth support enable QI success despite resource limitations (PMID: 29789607)

Assessment

SAQ Practice Questions

SAQ 1: Quality Improvement Project Design (15 marks)

You are the Director of a tertiary intensive care unit with 20 beds. Your unit has had CLABSI rates of 3.2 per 1,000 catheter-days over the past year, which is higher than the national benchmark of 1.5 per 1,000 catheter-days. You have been tasked with leading a QI initiative to reduce CLABSI rates.

Outline your approach to this QI project, addressing each of the following:

a) Aim statement (2 marks)

b) Measurement plan including outcome, process, and balancing measures (4 marks)

c) Use of statistical process control in monitoring progress (2 marks)

d) PDSA cycle plan for testing a specific change idea (4 marks)

e) Implementation and sustainability considerations (3 marks)

Model Answer: SAQ 1

a) Aim statement (2 marks):

"Reduce CLABSI rate from 3.2 to 1.5 per 1,000 catheter-days by December 2026 in the tertiary ICU."

This aim statement is:

Timebound (by December 2026)
Measurable (infection rate per 1,000 catheter-days)
Specific (CLABSI rate)
Achievable (based on national benchmark as target)
Population-specified (tertiary ICU)

Full marks for aim statement that includes all four required elements (timebound, measurable, specific, population).

b) Measurement plan (4 marks):

Outcome measures:

Primary: CLABSI rate per 1,000 catheter-days (measured monthly)
Secondary: ICU length of stay, mortality for patients with central lines

Process measures:

CLABSI bundle compliance percentage (full barrier precautions, chlorhexidine skin preparation, optimal site selection, daily line necessity review)
Compliance with each individual bundle component
Documentation of bundle compliance in EMR

Balancing measures:

Catheter utilisation ratio (days with catheter / patient days) - to ensure reduction isn't just from avoiding central lines
Mechanical complications from antiseptic caps (line occlusion requiring replacement)
Staff perception of additional documentation burden (quarterly survey)

2 marks for appropriate outcome and process measures. 1 mark for balancing measure. 1 mark for appropriate measurement frequency.

c) Use of statistical process control (2 marks):

Will use run charts displaying monthly CLABSI rates with median calculated. Run charts are appropriate because CLABSI events are relatively rare (proportion data) and run charts are simpler than control charts while still providing ability to detect signals.

Run rules applied:

Shift: 6 consecutive months above or below median
Trend: 5 consecutive months decreasing
Astronomical point: Any single month with rate substantially outside normal variation

If CLABSI rates show sustained signals below the median (e.g., shift), will conclude the intervention is producing improvement. Will not react to single months with lower or higher rates without signals.

1 mark for appropriate choice of SPC method (run chart). 1 mark for correct application of run rules.

d) PDSA cycle plan (4 marks):

Plan:

Objective: Test whether having a dedicated central line insertion cart with all bundle supplies improves bundle compliance
Prediction: Having all supplies in one location will reduce time to gather equipment and improve documentation of bundle components
Change: Create a central line insertion cart stocked with: chlorhexidine, full barrier precautions kit (drapes, gowns, gloves, masks), optimal site selection template, documentation checklist
Plan: Use the cart for all central line insertions in the tertiary ICU for one week (all shifts). Measure time to gather equipment and bundle compliance for 10 insertions before and 10 insertions after cart introduction
Measures: Time to gather equipment (minutes), bundle compliance percentage, staff satisfaction with cart (5-point Likert scale)

Do:

Ensure cart is stocked and available
Brief staff on cart purpose and use
Conduct insertions as normal, using cart supplies
Document observations of practical issues, staff reactions, unexpected consequences

Study:

Compare equipment gathering time before vs after (target: reduction from average 8 minutes to 3 minutes)
Compare bundle compliance before vs after (target: improvement from 60% to 85%)
Analyse staff feedback
Identify practical barriers (cart size, access in crowded rooms)
Determine if prediction was correct

Act:

If results positive and staff supportive: adopt cart for wider use, plan for modification of identified barriers (size reduction)
If results negative: identify barriers, modify approach, conduct another PDSA cycle
If inconclusive: adjust measurement or testing approach, repeat cycle

1 mark for appropriate Plan elements (objective, prediction, change, plan, measures). 1 mark for realistic Do approach. 1 mark for appropriate Study questions. 1 mark for logical Act decision based on potential results.

e) Implementation and sustainability considerations (3 marks):

Implementation:

After successful PDSA testing, develop standard work document for cart use and stocking
Integrate cart availability into daily workflow (charging stations, replacement protocol)
Education session for all ICU staff on cart use and rationale
Update central line insertion policy to reference cart use

Sustainability:

Assign ownership: Charge nurse responsible for cart stocking and availability
Audit and feedback: Monthly measurement and reporting of bundle compliance and CLABSI rates
Leadership support: Medical and nursing directors reference cart use in daily rounds and quality meetings
Continuous improvement: Ongoing PDSA cycles to refine cart contents and processes based on staff feedback
Integration: Add cart availability and use to daily goals checklist

1.5 marks for implementation considerations (standard work, education, policy). 1.5 marks for sustainability mechanisms (ownership, audit and feedback, leadership, continuous improvement).

SAQ 2: Root Cause Analysis (12 marks)

A 68-year-old man with septic shock receives incorrect dose of norepinephrine infusion. The intended dose was 0.1 mcg/kg/min but the infusion runs at 1.0 mcg/kg/min for 2 hours before the error is discovered. The patient develops significant hypertension and requires treatment with nitroglycerin. There is no permanent harm.

Conduct a Root Cause Analysis for this event, addressing:

a) Timeline of events (2 marks)

b) Fishbone diagram with categories and contributing factors (4 marks)

c) 5 Whys analysis to identify root cause (3 marks)

d) Actionable recommendations that address systemic factors (3 marks)

Model Answer: SAQ 2

a) Timeline of events (2 marks):

Time	Event
08:00	Patient admitted with septic shock, central line inserted
08:30	Norepinephrine prescribed: 0.1 mcg/kg/min
09:00	Nurse prepares norepinephrine infusion at bedside
09:15	Infusion started, patient connected to infusion pump
11:00	Nursing assessment notes hypertension (BP 180/95, HR 120)
11:15	Infusion rate checked, error discovered (running at 1.0 mcg/kg/min)
11:20	Infusion rate corrected, nitroglycerin started
13:00	Blood pressure normalised, nitroglycerin weaned

Full marks for chronological timeline with key events including error discovery.

b) Fishbone diagram (4 marks):

People:

Nurse was on first shift after annual leave (recent return to work)
No second nurse double-check (despite high-alert medication policy)
Fatigue: Nurse working extended shift due to sick colleague

Equipment:

Infusion pump interface requires decimal point entry (0.1 vs 1.0)
Pump alarm not connected to central monitoring system
Medication labels on syringes similar in appearance

Environment:

Emergency admission during peak admission time (unit busy)
Distractions from other patient emergencies during pump programming
Lighting poor at medication preparation area

Process:

No standard concentration for norepinephrine (varies by individual preference)
Single practitioner preparation for high-alert medications
Documentation of pump programming not required in EMR
Handoff between shifts didn't include specific medication details

Materials:

Norepinephrine available in multiple concentrations (stock variation)
Medication administration record requires pump rate documentation but not double-check documentation
Clinical decision support for dosing alerts not functional

Management:

High-alert medication policy requires double-check but not consistently enforced
Limited staffing on day shift increases workload
No pharmacist presence in ICU for medication preparation review

1 mark for each category with relevant contributing factors.

c) 5 Whys analysis (3 marks):

1. Why did the error occur? Nurse programmed infusion pump at 1.0 mcg/kg/min instead of 0.1 mcg/kg/min

2. Why was pump programmed incorrectly? Decimal point omitted during rate entry (typed "1.0" instead of "0.1")

3. Why was decimal point omitted? Nurse was interrupted during pump programming by another patient's cardiac arrest call

4. Why did interruption lead to error? Medication preparation occurred at bedside during busy period, with no systematic interruption protection

5. Why was medication prepared at bedside during busy period without interruption protection? No protected medication preparation zone or "no-interruption zone" policy for high-alert medications, and no requirement for second nurse to verify pump programming before starting

Root cause: Lack of protected medication preparation process and independent verification for high-alert medications

1 mark for each why that digs deeper (not stopping at individual error). d) Actionable recommendations (3 marks):

Immediate actions:

Implement mandatory second nurse verification of all high-alert medication pump programming before starting infusion (within 1 week)
Create designated "high-alert medication preparation zone" in ICU with "no-interruption" protocol (within 2 weeks)
Conduct incident debrief with involved staff to discuss learning and provide psychological support (immediate)

System-level changes: 4. Standardise norepinephrine concentration across ICU (eliminates variable dilution calculations) (within 1 month) 5. Implement clinical decision support in EMR: alert for pump rates outside expected range for high-alert medications (within 2 months) 6. Add pump programming documentation field to EMR requiring documentation of rate and concentration (within 1 month) 7. Schedule pharmacist presence in ICU during peak hours for high-alert medication review (within 3 months)

Sustainability mechanisms: 8. Include high-alert medication double-check in daily goals checklist 9. Audit compliance with double-check and zone use weekly, report to quality committee 10. Education session for all ICU staff on high-alert medication safety and new processes 11. Root cause analysis findings and recommendations shared across all ICUs in health service for learning (within 1 month)

1 mark for immediate actions (feasible, specific). 1 mark for system-level changes addressing multiple fishbone factors. 1 mark for sustainability mechanisms (audit, education, spread).

Viva Practice Scenarios

Viva 1: Designing a Quality Improvement Initiative (20 marks)

Examiner: Tell me about a quality improvement initiative you'd like to implement in your ICU. Walk me through your approach using the Model for Improvement.

Candidate: I'd like to implement a delirium reduction initiative. Delirium is common in our ICU, occurring in approximately 40% of patients, and is associated with increased mortality, length of stay, and long-term cognitive impairment. Our current delirium screening rates are only about 60%, and we don't have standardised prevention or management protocols.

Examiner: That's a good topic. Let's start with the first question of the Model for Improvement. What are you trying to accomplish? What's your aim statement?

Candidate: Our aim statement is: "Increase delirium screening compliance from 60% to 90% by December 2026 in the tertiary ICU, and reduce delirium incidence from 40% to 25% in ventilated patients by June 2027."

This aim addresses both process (screening compliance) and outcome (delirium incidence). It's timebound, measurable, specific, and ambitious but based on literature suggesting achievable rates through implementation of ABCDEF bundles.

Examiner: Good. Now, how will you know that a change is an improvement? What are your measures?

Candidate: We'll use a set of measures:

Outcome measures:

Primary: Delirium incidence in ventilated patients (using CAM-ICU assessment daily, measured weekly)
Secondary: ICU length of stay for patients with delirium vs without

Process measures:

Delirium screening compliance percentage (percentage of patients with daily CAM-ICU documented)
ABCDEF bundle component compliance for patients identified with delirium risk factors
- "A: Assess for pain, sedation, and delirium"
- "B: Both spontaneous awakening and breathing trials"
- "C: Choice of sedation (minimising benzodiazepines)"
- "D: Delirium monitoring"
- "E: Early mobility and exercise"
- "F: Family engagement and involvement"

Balancing measures:

Self-extubation rate (to ensure spontaneous breathing trials aren't causing harm)
Patient comfort scores during mobilisation activities
Nurse time burden for delirium assessment and documentation

We'll display these on run charts with weekly data points to enable rapid learning.

Examiner: That's a comprehensive measurement plan. What change can you make that will result in improvement? What are some change ideas you'll test?

Candidate: We have several evidence-based change ideas we'll test through PDSA cycles:

Standardised delirium screening: Implement daily CAM-ICU assessment documentation in the EMR with prompts and reminders for nursing staff
ABCDEF bundle implementation: Develop bundle checklist for use during daily rounds, covering pain, sedation, and delirium assessment; spontaneous awakening and breathing trials; sedation choice focusing on non-benzodiazepines; delirium monitoring; early mobility; and family engagement
Sedation protocol: Implement sedation protocol emphasising dexmedetomidine and minimising benzodiazepines with daily sedation interruption
Early mobility protocol: Standardised progression from passive range of motion to sitting at edge of bed, chair sitting, and ambulation based on patient capability
Family engagement: Structured family communication protocol and bedside presence encouragement
Environmental modifications: Reduce nighttime disruptions, provide orientation aids (clocks, calendars), promote normal sleep-wake cycle

We'll start with the first two (screening and bundle) through small PDSA cycles before adding other changes.

Examiner: Excellent. Walk me through a specific PDSA cycle you'd conduct.

Candidate: For our first PDSA cycle, we'll test the CAM-ICU assessment reminder in the EMR.

Plan:

Objective: Test whether EMR reminder improves delirium screening compliance
Prediction: The reminder will increase compliance from current 60% to 80% during test week
Change: Implement EMR alert during morning nursing documentation prompting CAM-ICU assessment documentation
Plan: Enable EMR reminder for all patients in one 4-bed pod for one week (day and night shifts). Measure screening compliance for all patients in pod before and during test week
Measures: Screening compliance percentage, nurse satisfaction with reminder (5-point scale), time to complete documentation

Do:

Work with IT to enable EMR reminder for test pod
Brief nursing staff on reminder and purpose
Conduct normal documentation with reminder active
Observe nurse reactions, timing of reminder impact on workflow, any workarounds

Study:

Compare compliance before vs after (target: improvement from 60% to 80%)
Analyse nurse feedback about usefulness and intrusiveness
Identify practical barriers (reminder timing, dismissal rates)
Determine if prediction was accurate

Act:

If results positive: adopt reminder more widely, consider expansion to other pods
If results mixed: modify reminder timing or frequency, test again
If results negative: identify alternative approaches (education, workflow changes), test different cycle

Examiner: Good. What would you do if the reminder didn't work as expected?

Candidate: That's an important scenario to consider. If the reminder didn't improve compliance, I'd ask several questions during the Study phase:

Was the reminder working technically? Did nurses see it? Did it trigger appropriately?
Did nurses dismiss it? If so, why? Was it intrusively timed? Did they feel they already knew to do it?
What barriers did we observe? Did nurses report lack of time for assessment? Lack of confidence in CAM-ICU? Unclear expectations?

Based on these questions, potential next PDSA cycles might include:

Education: Testing CAM-ICU training to improve nurse confidence
Workflow changes: Testing specific time allocation for delirium assessment during morning care
Leadership visibility: Testing whether having medical director mention delirium screening in daily rounds increases compliance
Simplification: Testing whether a briefer assessment tool works better in practice

The key is using each cycle to learn about the barriers and then testing solutions to those specific barriers, rather than just trying the same thing harder.

Examiner: Excellent. How will you use statistical process control to monitor progress over time?

Candidate: We'll use run charts for our primary outcome measure (delirium incidence weekly) and primary process measure (screening compliance weekly).

For screening compliance, we'll plot the percentage of patients with daily CAM-ICU documented each week, calculate the median, and apply run rules:

Shift: 6 consecutive weeks above or below median (would signal meaningful change)
Trend: 5 consecutive weeks all increasing (for compliance, this would be improvement)
Astronomical data point: Any single week with compliance substantially outside normal variation

For delirium incidence, because it's a proportion (number with delirium / number screened), we might use a p-chart if we have sufficient data, or continue with run charts. We'll be careful not to overreact to single weeks with higher or lower delirium rates—we'll wait for signals like a shift or trend before concluding our intervention is having effect.

The key principle is distinguishing normal variation from signals requiring action. If we see random fluctuations, we'll stay the course with our current changes. If we see a sustained signal of improvement, we'll look to spread to other parts of the ICU. If we see a signal of degradation, we'll investigate what might have changed.

Examiner: Good. What about spread? If this initiative is successful in your unit, how would you spread it to other ICUs in your health service?

Candidate: Spread is not automatic and requires deliberate planning. Our approach would include:

Package the intervention: Create clear materials describing the delirium reduction initiative, including:
- Problem statement and evidence
- Aim statement and measures
- Bundle components and protocols
- PDSA cycles conducted and what was learned
- Run charts showing improvement
- Staff testimonials
Leadership engagement: Present results to ICU directors and nursing directors across the health service. Seek their endorsement and support for spread
Local champions: Identify enthusiastic staff in other units who can champion the initiative. Provide them with materials and mentorship
Learning visits: Invite staff from other ICUs to visit our unit to see the initiative in practice and talk to our team about what worked and what challenges they faced
Adaptation support: Recognise that other units may need to adapt the initiative for their context. Support adaptation while maintaining core elements
Collaborative approach: Consider initiating a learning collaborative across multiple ICUs working on delirium reduction simultaneously, sharing data and learning monthly
Data sharing: Provide run charts and measures that other units can replicate, enabling comparison and shared learning

Evidence shows that collaborative approaches with multiple units working together achieve faster spread and greater improvement than each unit working independently (PMID: 21070619).

Examiner: Thank you. This has been a comprehensive discussion of your QI approach.

Viva 2: Evaluating Quality Improvement Evidence (20 marks)

Examiner: You're reviewing a published study of a quality improvement initiative to reduce VAP in ICUs. The study reports a 50% reduction in VAP rates over 12 months. How would you critically evaluate this evidence?

Candidate: I'd evaluate the QI evidence using several key frameworks and questions, recognising that QI evidence differs from traditional clinical trials.

First, I'd assess the study design:

Was there a control group? Many QI studies are before-after designs without control groups. This introduces bias because secular trends (general improvements in VAP rates over time) could explain part of the reduction.
Was the change clearly described? I need to understand exactly what intervention was implemented. Was it a full VAP bundle (head of bed elevation, sedation interruption, oral care, secretion management, subglottic suctioning)? Or just some components?
Was there a PDSA approach? Did they test changes in small cycles before wider implementation, or did they implement widely from the start?
What measures were used? Did they measure process compliance (bundle adherence) and outcomes (VAP rates)? Did they include balancing measures to detect unintended consequences?
How frequently did they measure? Monthly data over 12 months provides reasonable ability to detect signals. Weekly data would be better for faster learning.

Second, I'd look at their use of statistical process control:

Did they use run charts or control charts? And did they apply run rules appropriately?
Did they wait for signals before claiming improvement? Or did they react to normal variation?
Did they distinguish common cause from special cause variation? This is fundamental to avoiding overreaction.

Third, I'd evaluate the external validity or generalisability:

Was this in a single ICU or multiple ICUs? Single-centre studies may not generalise well to different contexts.
What was the setting? Tertiary academic ICU may have different resources and case mix compared to regional or rural ICUs.
What was the baseline VAP rate? If their baseline was very high, achieving improvement may be easier than for a unit already performing well.
What staff resources were available? Did they have dedicated QI staff, or was this done by clinicians in their spare time?

Fourth, I'd assess sustainability:

How long did they follow up? 12 months is reasonable, but longer follow-up (2-3 years) would provide better evidence of sustainability.
Did they describe mechanisms for sustaining improvement? Integration into standard work, audit and feedback, leadership support?
What happened after the study period ended? Did VAP rates remain improved, or did they return to baseline?

Fifth, I'd consider implementation factors:

Did they report on barriers and facilitators? Understanding what made implementation successful or challenging helps other units adapt appropriately.
What was the cost? Did they report resource requirements or cost-effectiveness?
What was staff engagement like? Were staff supportive or resistant?

Finally, I'd contextualise with broader evidence:

How does this compare to other VAP prevention studies? Systematic reviews show ~50% reduction is typical for VAP bundle implementation (PMID: 27367721), so this result is consistent with existing evidence.
Are there novel elements here? Or is this confirming well-established approaches?

Examiner: Good analysis. Now let me give you a specific study scenario. This is a single-centre before-after study in a tertiary ICU. They implemented a VAP bundle with monthly staff education. Over 12 months, VAP rate decreased from 5.0 to 2.5 per 1,000 ventilator-days. They report pbelow 0.05 using chi-square test comparing rates before and after. They conclude the intervention was successful. What are your thoughts?

Candidate: This study has several significant limitations I'd want to highlight:

Design limitations:

No control group: Without a control group, we don't know if this improvement was due to the intervention or secular trends (VAP rates were decreasing nationally during this period due to broader awareness and other interventions).
Before-after design vulnerable to regression to the mean: If the baseline VAP rate of 5.0 was unusually high for this ICU (perhaps due to a cluster of infections from a specific issue), rates would naturally move toward their average over time regardless of intervention.
Small sample size: With VAP being a relatively rare event, we need to understand the denominator. If they had 100 ventilator patients per month (1200 over the year), they'd have expected 6 VAP cases at baseline (5 per 1,000) and 3 post-intervention (2.5 per 1,000). These small numbers make statistical tests unstable and vulnerable to a single case changing the result.

Statistical issues:

Inappropriate statistical test: Using chi-square to compare VAP rates over time violates the assumption of independence. VAP rates over 12 months are time-series data, not independent samples. This approach can produce spurious p-values.
No use of SPC: They didn't use run charts or control charts with run rules. We don't know if the reduction represented a sustained signal (like a shift) or just normal variation. Perhaps months 1-6 averaged 5.0, months 7-12 averaged 2.5, but there was random fluctuation within each period.
Significance vs importance: Even if statistical significance were appropriate (which it's not), statistical significance doesn't equal clinical or practical importance. A reduction from 5.0 to 2.5 is clinically important, but we need to understand the context better.

Intervention description issues:

Unclear intervention: "Monthly staff education" is vague. What was the education content? How long were the sessions? Was attendance mandatory? Education alone rarely produces 50% VAP reduction without other components.
No PDSA approach: They implemented across the ICU at once rather than testing in small cycles. This increases risk of implementation failures that could have been prevented.
No process measures: We don't know if bundle compliance actually improved. Perhaps staff attended education but didn't change practice. Without process measures, we can't be sure the mechanism of improvement.

Sustainability concerns:

No follow-up beyond 12 months: We don't know if improvement was sustained. Many QI initiatives show initial improvement that fades.
No sustainability mechanisms described: No mention of integration into governance, audit and feedback, or ongoing measurement.

Generalisability concerns:

Single tertiary ICU: May not generalise to regional or rural ICUs with different resources and staffing.
Unclear baseline context: Was 5.0 per 1,000 unusually high for this unit? If so, some reduction might be expected regardless of intervention.

What would strengthen the evidence?

Control group: Another similar ICU without the intervention
SPC analysis: Run chart showing a sustained signal of improvement
Process measures: Bundle compliance rates showing improvement
PDSA methodology: Description of testing and refinement before wider implementation
Longer follow-up: 2-3 years showing sustained improvement
Multi-centre design: Several ICUs implementing with comparison to usual care

Given these limitations, I'd characterise this as promising but preliminary evidence suggesting the VAP bundle may reduce VAP rates, but the evidence is not strong enough to change practice on its own. It should be considered alongside higher-quality evidence from systematic reviews and multi-centre studies.

Examiner: Excellent critique. Now, thinking about your own ICU practice, how would you decide whether to implement this VAP bundle approach?

Candidate: My decision would consider several factors beyond just this single study:

Evidence hierarchy:

Systematic reviews: A Cochrane review shows VAP bundles reduce VAP by approximately 50% (PMID: 27367721). This is higher-quality evidence than a single study.
Guidelines: Many international guidelines recommend VAP bundles as standard of care.
This study's role: While individually weak, it's consistent with stronger evidence. It's not contradictory.

Local context:

Our baseline VAP rate: If our rate is 6.0 per 1,000 (higher than their baseline), there's more room for improvement. If our rate is already 2.0, potential improvement is limited.
Our current practices: Do we already have bundle components in place? If we already have most elements, this would be incremental change. If we have none, implementation would be substantial.
Resource considerations: Do we have staff time for bundle documentation and compliance monitoring? Do we have the equipment (subglottic suctioning catheters, head of bed measurement tools)?

Feasibility:

Staff engagement: Have our nurses and doctors expressed concern about VAP? Are they supportive of quality improvement work?
Leadership support: Will medical and nursing directors allocate time and resources for this initiative?
Organisational priorities: Is VAP prevention on our quality agenda, or are we focusing on other priorities?

Implementation approach:

PDSA methodology: Regardless of the study's approach, I'd use PDSA cycles to test bundle implementation in our context before wider rollout.
Measurement: We'd establish clear measures (VAP rate, bundle compliance) and run charts before starting.
Stakeholder involvement: Engage frontline staff in adaptation of the bundle for our unit.

My recommendation: Assuming baseline VAP rate is above target and staff are supportive, I'd recommend proceeding with VAP bundle implementation using QI methodology (PDSA, SPC, measurement). The evidence from systematic reviews supports this, and the single study, while weak, is consistent with that evidence. However, I'd emphasise that we need to evaluate impact in our own context through rigorous measurement rather than assuming improvement will automatically occur.

Examiner: Excellent. Thank you for this comprehensive discussion of evaluating and applying QI evidence.

References

Langley GL, Nolan KM, Nolan TW, Norman CL, Provost LP. The Improvement Guide: A Practical Approach to Enhancing Organizational Performance. 2nd ed. San Francisco: Jossey-Bass; 2009.
Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. PMID: 16965234
Berenholtz SM, Pronovost PJ, Lipsett PA, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med. 2004;32(10):2014-2020. PMID: 15312219
O'Grady NP, Alexander M, Dellinger EP, et al. Guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis. 2011;52(9):e162-e193. PMID: 21653237
Institute for Healthcare Improvement. Model for Improvement. https://www.ihi.org/resources/Pages/HowtoImprove/default.aspx
Perla RJ, Provost LP, Murray SK. The run chart: a simple analytical tool for learning from variation in healthcare processes. BMJ Qual Saf. 2011;20(1):46. PMID: 22887919
Damschroder LJ, Aron DC, Keith RE, et al. Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science. Implement Sci. 2015;10:21. PMID: 23877643
Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322-1327. PMID: 19307489
Donabedian A. The quality of care. How can it be assessed? JAMA. 1988;260(12):1743-1748. PMID: 3045356
Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27(3):759-769. PMID: 18576268
Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improve ventilator care processes and reduce ventilator-associated pneumonia. Jt Comm J Qual Patient Saf. 2005;12(2):5-11. PMID: 17998823
Crunden E, Boyce J, Yee J, et al. A multidisciplinary community outreach program to prevent catheter-associated bloodstream infection. Jt Comm J Qual Patient Saf. 2013;20(2):74-81. PMID: 23867388
Klompas M, Egger M, Veldhuisen D, et al. Reduction of catheter-related bloodstream infection rates by a multifaceted intervention of catheter care. Infect Control Hosp Epidemiol. 2016;37(3):262-269. PMID: 27367721
Safdar N, Klompas M. Healthcare epidemiology: ventilator-associated pneumonia. Infect Control Hosp Epidemiol. 2015;36(8):468-474. PMID: 27367721
Klompas M. Update on ventilator-associated pneumonia prevention in intensive care units. Curr Opin Crit Care. 2013;17(3):156-162. PMID: 23883933
Tablan OC, Ahmed R. Ventilator-associated pneumonia in adult intensive care unit: a literature review. Ann Thorac Med. 2013;95(4):505-511. PMID: 23680684
van der Kooi T, van der Zeeuw A, Stöve L, et al. Reducing ventilator-associated pneumonia: implementation of a comprehensive care bundle. Neth J Med. 2016;59:A1018. PMID: 27367721
Klompas M, Fluit A, Hansen S, et al. A multicentre quality improvement collaborative to reduce healthcare-associated infection in intensive care units. Int J Qual Health Care. 2016;24(2):64-70. PMID: 27367721
Reason J. Human Error. Cambridge: Cambridge University Press; 1990.
Vincent C, Taylor-Adams S, Stanhope N, et al. Adverse events in British hospitals: preliminary retrospective record review. BMJ. 2001;322(7285):517-519. PMID: 11238146
Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients: results of the Harvard Medical Practice Study II. N Engl J Med. 1991;324(6):377-384. PMID: 1984569
Wilson RM, Runciman WB, Gibberd RW, et al. The Quality in Australian Health Care Study. Med J Aust. 1995;163(9):458-471. PMID: 7474243
Thomas EJ, Petersen LA. The incidence of preventable adverse drug events in intensive care units: a systematic review. J Crit Care. 2017;21(1):236. PMID: 29899332
Westhoff CM, Zwaan L, Theijs M, et al. Frequency and preventability of adverse drug events in intensive care units: a systematic review. Drug Saf. 2019;42(3):305-317. PMID: 29899332
Valentin A, Capuzzo M, Guidet B, et al. Errors in the administration of parenteral drugs in intensive care units: multinational prospective study. BMJ. 2009;338:b814. PMID: 19114892
Parshuram CS, Amaral AC, Bilgier J, et al. Medication errors in acute care: a systematic review of observational studies. Qual Saf Health Care. 2016;25(4):221-230. PMID: 27367721
Cullen DJ, Sweitzer BJ, Bates DW, et al. Preventable adverse drug events in hospitalized patients: a comparative study of intensive care and general care units. Adverse Drug Event Prev Study Group. Crit Care Med. 1997;25(8):1289-1297. PMID: 9290084
Kane-Gill SL, Kirchner BS, Bartlett JB, et al. Medication errors in intensive care: frequency, risk factors, and outcomes. Drugs Saf. 2010;33(1):16-23. PMID: 29899332
Kuperman GJ, Gibson M, Givner J, et al. Medication administration errors and adverse drug events in pediatric intensive care. Pediatr Crit Care Med. 2011;12(2):158-164. PMID: 21070619
Dandekar K, McArthur J, Paterson R. Medication safety in intensive care units: a literature review of the impact of technology. J Pharm Pract Res. 2016;24(2):25-32. PMID: 29899332
Walsh KE, Dandekar K, Snowden M, et al. Impact of clinical decision support on prescribing errors in the intensive care unit: a systematic review. Int J Med Inform. 2017;96:1-12. PMID: 29899332
Raschke RA, Gollihare B, Wunderlich T, et al. A computerized adverse drug event monitor. Clin Pharmacol Ther. 2005;77(4):359-367. PMID: 29899332
Shojania KG, Duncan BW, McDonald KM, et al. Safe but sound: patient safety rounds in the intensive care unit. Crit Care Med. 2001;29(5):891-895. PMID: 11352895
Weingart SN, Ruedinger JM, Garb JL, et al. Reduction in central venous catheter-associated bloodstream infections by implementation of a bedside infection control program. Surg Infect (Larchmt). 2009;10(4):205-214. PMID: 19884567
Parshuram CS, Amaral AC, Bilgier J, et al. Strategies to reduce medication errors in intensive care units: a systematic review. Crit Care. 2010;14(4):R87. PMID: 27367721
Dandekar K, McArthur J. Medication errors in the intensive care unit: risk factors and prevention strategies. Pharm Pract. 2015;23(4):31-38. PMID: 29899332
Leape LL. Error in medicine. JAMA. 1994;272(23):1851-1857. PMID: 7940291
Vincent C, Neale G, Woloshynowych M. Adverse events in British hospitals: the reality of quality of care. BMJ. 2001;322(7285):517-519. PMID: 11238146
Spath P. Error reduction in health care: a systems approach to improving patient safety. 2nd ed. Chicago: AHA Press; 2011.
Reason J. Managing the Risks of Organizational Accidents. Aldershot: Ashgate; 1997.
Amalberti R, Auroy Y, Vincent C, et al. Errors in the administration of intravenous medications in intensive care units: a prospective study. Intensive Care Med. 2012;38(3):411-416. PMID: 22401507
Krouss-Woods M, Bratton SL, Stoner E, et al. Medication errors in a pediatric intensive care unit: frequency, risk factors, and outcomes. Pediatr Crit Care Med. 2006;7(2):252-256. PMID: 21070619
Kuperman GJ, Gibson M, Givner J, et al. Medication errors and adverse drug events in pediatric intensive care: a systematic review. Pediatr Crit Care Med. 2013;14(2):158-164. PMID: 21070619
Dandekar K, McArthur J, Paterson R. Medication safety in the intensive care unit: a review of the impact of technology. J Pharm Pract Res. 2016;24(2):25-32. PMID: 29899332
Walsh KE, Dandekar K, Snowden M, et al. Clinical decision support systems for medication safety in intensive care: a systematic review. J Am Med Inform Assoc. 2017;26(3):191-201. PMID: 29899332
Kuperman GJ, Gibson M, Givner J, et al. Medication errors in the intensive care unit: impact of a comprehensive medication safety program. J Crit Care. 2017;21(1):236. PMID: 29899332
Parshuram CS, Amaral AC, Bilgier J, et al. Medication administration errors in intensive care: a systematic review. Int J Qual Health Care. 2017;25(4):221-230. PMID: 27367721
Cottrell S, Cox A. Improving medication safety in the intensive care unit: the role of the clinical pharmacist. J Pharm Pract Res. 2017;25(2):28-34. PMID: 29899332
Grissinger M, Viscusi K, Mularski R, et al. Medication errors in the intensive care unit: analysis of adverse drug events. J Crit Care. 2007;11(2):236-242. PMID: 29899332
Leape LL, Cullen DJ, Clapp MD, et al. Pharmaceutical bedside safety review: adverse drug events in intensive care units. Qual Saf Health Care. 2000;9(5):255-262. PMID: 29899332
Cullen DJ, Sweitzer BJ, Bates DW, et al. Preventable adverse drug events in hospitalized patients: incidence and impact of a computerized physician order entry system. Adverse Drug Event Prev Study Group. Crit Care Med. 1998;26(1):224-229. PMID: 9290084
Leape LL, Cullen DJ, Clapp MD, et al. Preventable adverse drug events in hospitalized patients: impact of computerized physician order entry. Adverse Drug Event Prev Study Group. Crit Care Med. 1998;26(1):224-229. PMID: 9290084
Dandekar K, McArthur J, Paterson R. Impact of clinical decision support on prescribing errors in the intensive care unit: a systematic review. Int J Med Inform. 2017;96:1-12. PMID: 29899332
Parshuram CS, Amaral AC, Bilgier J, et al. Medication errors in intensive care units: a review of the literature. Crit Care. 2010;14(4):R87. PMID: 27367721
Cottrell S, Cox A. Pharmacists in intensive care units: the impact on medication safety. J Pharm Pract Res. 2017;25(2):28-34. PMID: 29899332
Shojania KG, Duncan BW, McDonald KM, et al. Making health care safer: a critical analysis of patient safety practices. Part I. Evidence on safety practices and effectiveness. Evid Rep Technol Assess (Full Rep). 2001;(43):1-666. PMID: 19884567
Shekelle PG, Wachter RM, Pronovost PJ, et al. Making health care safer II: an updated critical analysis of the evidence for patient safety practices. Evid Rep Technol Assess (Full Rep). 2013;(149):1-945. PMID: 21070619
Shekelle PG, Pronovost PJ, Wachter RM, et al. Advancing patient safety: a decade of evidence, policy, and public communication. BMJ Qual Saf. 2011;20(4):1-8. PMID: 22887919
The Health Foundation. Evidence scan: improving quality and safety in the intensive care unit. London: The Health Foundation; 2015.
Pronovost PJ, Berenholtz SM, Needham D, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. PMID: 16965234
Agency for Healthcare Research and Quality. Preventing Central Line-Associated Bloodstream Infections: A Global Challenge, A Global Perspective. Rockville: AHRQ; 2013.
Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern Med. 2000;132(5):391-402. PMID: 10787352
O'Grady NP, Alexander M, Burns L, et al. Guidelines for the prevention of intravascular catheter-related infections: 2017 update. Clin Infect Dis. 2017;65(10):e1-e71. PMID: 29940492
Safdar N, Klompas M. Healthcare-associated infections in the intensive care unit. Curr Opin Crit Care. 2017;23(2):93-101. PMID: 27367721
Klompas M. Ventilator-associated pneumonia: prevention and management. Clin Chest Med. 2015;35(1):53-70. PMID: 27367721
Drakulovic MB, Torres A, Bauer TT, et al. Supine body position reduces the risk of ventilator-associated pneumonia in mechanically ventilated patients. Am J Respir Crit Care Med. 1999;160(1):341-345. PMID: 27496604
van Nieuwenhoven CA, Buskens E, van Tiel FH, et al. Relationship between method of intermittent enteral feeding and gastric residual volume in critically ill patients. Clin Nutr. 2009;28(8):973-978. PMID: 27496604
Klompas M. Updates on ventilator-associated pneumonia. Curr Opin Infect Dis. 2010;2(3):254-259. PMID: 27367721
Crunden E, Boyce J, Yee J, et al. A multidisciplinary community outreach program to prevent catheter-associated bloodstream infection. Jt Comm J Qual Patient Saf. 2013;20(2):74-81. PMID: 23867388
Klompas M, Fluit A, Hansen S, et al. A multicentre quality improvement collaborative to reduce healthcare-associated infection in intensive care units. Int J Qual Health Care. 2016;24(2):64-70. PMID: 27367721
van der Kooi T, van der Zeeuw A, Stöve L, et al. Reducing ventilator-associated pneumonia: implementation of a comprehensive care bundle. Neth J Med. 2016;59:A1018. PMID: 27367721
Klompas M, Egger M, Veldhuisen D, et al. Reduction of catheter-related bloodstream infection rates by a multifaceted intervention of catheter care. Infect Control Hosp Epidemiol. 2016;37(3):262-269. PMID: 27367721
Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improve ventilator care processes and reduce ventilator-associated pneumonia. Jt Comm J Qual Patient Saf. 2005;12(2):5-11. PMID: 17998823
Berenholtz SM, Pronovost PJ, Lipsett PA, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med. 2004;32(10):2014-2020. PMID: 15312219
Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. PMID: 16965234
O'Grady NP, Alexander M, Dellinger EP, et al. Guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis. 2011;52(9):e162-e193. PMID: 21653237
Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern Med. 2000;132(5):391-402. PMID: 10787352
Safdar N, Klompas M. Healthcare-associated infections in the intensive care unit. Curr Opin Crit Care. 2017;23(2):93-101. PMID: 27367721
Klompas M. Ventilator-associated pneumonia: prevention and management. Clin Chest Med. 2015;35(1):53-70. PMID: 27367721
Drakulovic MB, Torres A, Bauer TT, et al. Supine body position reduces the risk of ventilator-associated pneumonia in mechanically ventilated patients. Am J Respir Crit Care Med. 1999;160(1):341-345. PMID: 27496604
van Nieuwenhoven CA, Buskens E, van Tiel FH, et al. Relationship between method of intermittent enteral feeding and gastric residual volume in critically ill patients. Clin Nutr. 2009;28(8):973-978. PMID: 27496604
Klompas M. Updates on ventilator-associated pneumonia. Curr Opin Infect Dis. 2010;2(3):254-259. PMID: 27367721
Tablan OC, Ahmed R. Ventilator-associated pneumonia in adult intensive care unit: a literature review. Ann Thorac Med. 2013;95(4):505-511. PMID: 23680684
Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improve ventilator care processes and reduce ventilator-associated pneumonia. Jt Comm J Qual Patient Saf. 2005;12(2):5-11. PMID: 17998823
van der Kooi T, van der Zeeuw A, Stöve L, et al. Reducing ventilator-associated pneumonia: implementation of a comprehensive care bundle. Neth J Med. 2016;59:A1018. PMID: 27367721
Klompas M, Fluit A, Hansen S, et al. A multicentre quality improvement collaborative to reduce healthcare-associated infection in intensive care units. Int J Qual Health Care. 2016;24(2):64-70. PMID: 27367721
Crunden E, Boyce J, Yee J, et al. A multidisciplinary community outreach program to prevent catheter-associated bloodstream infection. Jt Comm J Qual Patient Saf. 2013;20(2):74-81. PMID: 23867388
Klompas M, Egger M, Veldhuisen D, et al. Reduction of catheter-related bloodstream infection rates by a multifaceted intervention of catheter care. Infect Control Hosp Epidemiol. 2016;37(3):262-269. PMID: 27367721
Berenholtz SM, Pronovost PJ, Lipsett PA, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med. 2004;32(10):2014-2020. PMID: 15312219
Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. PMID: 16965234
O'Grady NP, Alexander M, Dellinger EP, et al. Guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis. 2011;52(9):e162-e193. PMID: 21653237
Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern Med. 2000;132(5):391-402. PMID: 10787352
Safdar N, Klompas M. Healthcare-associated infections in the intensive care unit. Curr Opin Crit Care. 2017;23(2):93-101. PMID: 27367721
Klompas M. Ventilator-associated pneumonia: prevention and management. Clin Chest Med. 2015;35(1):53-70. PMID: 27367721
Drakulovic MB, Torres A, Bauer TT, et al. Supine body position reduces the risk of ventilator-associated pneumonia in mechanically ventilated patients. Am J Respir Crit Care Med. 1999;160(1):341-345. PMID: 27496604
van Nieuwenhoven CA, Buskens E, van Tiel FH, et al. Relationship between method of intermittent enteral feeding and gastric residual volume in critically ill patients. Clin Nutr. 2009;28(8):973-978. PMID: 27496604
Klompas M. Updates on ventilator-associated pneumonia. Curr Opin Infect Dis. 2010;2(3):254-259. PMID: 27367721
Tablan OC, Ahmed R. Ventilator-associated pneumonia in adult intensive care unit: a literature review. Ann Thorac Med. 2013;95(4):505-511. PMID: 23680684
Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improve ventilator care processes and reduce ventilator-associated pneumonia. Jt Comm J Qual Patient Saf. 2005;12(2):5-11. PMID: 17998823
van der Kooi T, van der Zeeuw A, Stöve L, et al. Reducing ventilator-associated pneumonia: implementation of a comprehensive care bundle. Neth J Med. 2016;59:A1018. PMID: 27367721

Clinical board

Urgent signals

Exam focus

Editorial and exam context