The Model for Improvement: An Essential Framework for Healthcare Improvement

Introduction

Every day in the NHS, dedicated professionals work to deliver the best possible care. However, even the most committed teams face challenges, from patient flow bottlenecks to variations in clinical practice. The Model for Improvement (MFI) offers a structured yet flexible approach to address these challenges, allowing teams to test changes on a small scale, learn from them, and then implement successful interventions more widely.

Developed by Associates in Process Improvement and adopted by organisations globally, including NHS England, the MFI provides a clear, actionable roadmap for initiating and sustaining improvement. It integrates seamlessly with existing clinical pathways and governance structures, making it an invaluable tool for anyone looking to make a positive impact in healthcare.

Why this topic matters

For clinicians and NHS teams, understanding and applying the Model for Improvement is crucial for several reasons:

• Structured Approach: It provides a common language and framework for QI, ensuring projects are well-defined and executed systematically.
• Efficiency and Effectiveness: By focusing on specific aims and using rapid-cycle testing (PDSA cycles), it helps teams learn quickly and avoid costly, large-scale failures.
• Empowerment: It empowers frontline staff to identify problems, develop solutions, and drive change within their own areas of practice.
• Evidence-informed Practice: It encourages data-driven decision-making, moving beyond anecdotal evidence to understand the impact of changes.
• Resilience and Adaptability: Its iterative nature allows for constant adjustment and refinement, which is essential in the dynamic environment of healthcare.
• Meeting National Priorities: Many national initiatives and CQC expectations increasingly rely on demonstrable QI activity, making proficiency in frameworks like the MFI highly beneficial.

Practical explanation: The Three Questions and PDSA Cycles

The Model for Improvement is built around two core components:

1. The Three Fundamental Questions

Before embarking on any improvement work, the MFI encourages teams to answer three critical questions:

• What are we trying to accomplish?

  • This question helps define the aim of the project. A good aim statement is specific, measurable, achievable, relevant, and time-bound (SMART). It should clearly state what you want to improve, by how much, and by when.
  • Example: "Reduce the average waiting time for specialist outpatient review from 12 weeks to 6 weeks for patients referred from Primary Care, within 6 months, without increasing clinic capacity."

• How will we know that a change is an improvement?

  • This question focuses on measurement. It requires identifying specific data points or metrics that will indicate whether your changes are leading to the desired improvement. These are often categorized as outcome, process, and balancing measures.
  • Example: Waiting list length, average waiting time (outcome measures); percentage of referrals triaged within 24 hours (process measure); clinic DNA rate (balancing measure to check for unintended consequences).

• What changes can we make that will result in improvement?

  • This question prompts the generation of ideas for change. This is an opportunity for creative thinking, brainstorming, and drawing on evidence, best practice, and local expertise. Start with small, actionable ideas that can be tested.
  • Example: Implementing a new standardised referral proforma, introducing a 'straight-to-test' pathway for specific conditions, or redesigning the triage process.

2. Plan-Do-Study-Act (PDSA) Cycles

Once the three questions are answered, the MFI guides teams through a series of rapid learning cycles known as Plan-Do-Study-Act (PDSA). This iterative process allows for testing changes on a small scale, learning from the results, and refining the change before broader implementation.

• Plan:

  • What to test: Clearly define the specific change you will test.
  • Who, what, where, when: Detail how the test will be carried out.
  • Prediction: State what you expect to happen.
  • Measures: How will you collect data during this specific test?
  • Example: "Test the new referral proforma with 5 GP practices next Monday. We predict it will reduce incomplete referrals by 50%. We will track incomplete referrals from these practices."

• Do:

  • Carry out the test: Implement the planned change.
  • Observe and collect data: Document what actually happens, noting any unexpected issues or observations. Stick to the plan as much as possible.
  • Example: Implement the proforma, collect feedback from practices, record the number of incomplete referrals received.

• Study:

  • Analyse the results: Compare the collected data to your predictions.
  • Summarise learnings: What went well? What didn't? Why? What did you learn?
  • Example: "The new proforma reduced incomplete referrals by 30%, not 50%. Some sections were unclear to GPs, and there was initial resistance to a new form."

• Act:

  • What next? Based on your learning, decide what to do next.
    • Adopt: If the change was successful, integrate it more widely.
    • Adapt: Modify the change based on learning and repeat the PDSA cycle.
    • Abandon: If the change was ineffective or caused new problems, discard it and try a different idea.
  • Example: "Adapt the proforma to clarify confusing sections and conduct a short Q&A session with another 5 practices. Plan a new PDSA cycle for next month."

Multiple, small PDSA cycles are far more effective than one large, complex cycle. They allow for rapid learning and minimise risk.

Common pitfalls

While the MFI is intuitive, common pitfalls can hinder progress:

• Vague Aim Statements: Without clear, measurable aims, it's impossible to know if you've achieved anything.
• No Measurement Strategy: Relying on 'gut feeling' rather than data means you can't objectively assess if a change is an improvement.
• Jumping to Solutions: Skipping the first two questions and implementing changes without understanding the problem or how to measure success.
• 'Big Bang' Implementations: Attempting to implement a large-scale change without prior small-scale testing. This significantly increases risk and potential for failure.
• Ignoring Balancing Measures: Focusing solely on the primary aim without considering potential negative consequences in other areas.
• Lack of Team Engagement: QI is a team sport. Without buy-in and active participation from those doing the work, changes are unlikely to stick.
• Confusing QI with Audit or Research: While related, QI focuses on improving a local system, not generating generalisable knowledge (research) or simply assessing current compliance (audit). Local governance advice should always be sought.

Step-by-step approach to using the Model for Improvement

1. Form your team: Assemble a small, multidisciplinary team of frontline staff, managers, and relevant stakeholders.
2. Define your aim: Use the first MFI question to craft a SMART aim statement. Involve your team in this.
3. Identify measures: Determine what data you will collect (outcome, process, balancing) to track progress towards your aim. Establish baseline data.
4. Brainstorm change ideas: Generate potential solutions to achieve your aim. Consider evidence, 'what works elsewhere', and local expertise.
5. Prioritise and select a change: Choose one small, manageable change to test first.
6. Plan your first PDSA cycle: Detail who, what, where, when, and what to expect from this specific test.
7. Execute the 'Do' phase: Carry out the test as planned, collecting data and observing.
8. 'Study' the results: Analyse the data, compare with predictions, and discuss learnings with your team.
9. 'Act' on your learning: Decide whether to adopt, adapt, or abandon the change. Document your decision.
10. Repeat PDSA cycles: Continue to iterate, refining changes and expanding their scope as you gain confidence and evidence of improvement.
11. Sustain and spread: Once improvement is consistently demonstrated, embed the changes into routine practice and consider spreading successful interventions to other relevant areas.

This resource supports, but does not replace, clinical judgement. Local policy, formulary and specialist advice should be followed. Always ensure relevant local governance and clinical safety processes are adhered to.

Example in clinical practice: Reducing Post-Operative Nausea and Vomiting (PONV)

A surgical ward team identified that patients often experienced significant Post-Operative Nausea and Vomiting (PONV), leading to prolonged recovery, patient dissatisfaction, and occasional readmissions.

• What are we trying to accomplish?

  • Within 3 months, reduce the incidence of moderate to severe PONV on Ward 7 from 30% to 15% in adult elective surgical patients, without increasing antiemetic expenditure.

• How will we know that a change is an improvement?

  • Outcome Measure: Percentage of patients reporting moderate to severe PONV using a validated local scale post-operatively (tracked daily).
  • Process Measure: Percentage of high-risk patients receiving PONV prophylaxis according to local guidelines (tracked per shift).
  • Balancing Measure: Total antiemetic drug costs per patient, length of hospital stay.

• What changes can we make that will result in improvement?

  • Initial ideas included improving risk stratification for PONV, standardising antiemetic prescribing, implementing new patient education, or using alternative antiemetics.

PDSA Cycle 1: Enhanced Risk Assessment

• Plan: Test providing a laminated 'PONV Risk Assessment' checklist to two anaesthetists for 5 elective patients each during their pre-operative assessment. Predict it will improve compliance with local guidelines by 20%. Collect feedback from anaesthetists.
• Do: Anaesthetists used the checklist. Data showed some initial confusion on scoring, but overall, 80% used it. Two patients reported issues.
• Study: Compliance improved by 15%, but the scoring system was not user-friendly. Anaesthetists felt it added time to the assessment.
• Act: Adapt the checklist to simplify the scoring, incorporate it directly into the electronic pre-assessment form, and plan to test with all anaesthetists next week.

PDSA Cycle 2: Standardised Prophylaxis Protocol

• Plan: Test the revised PONV risk assessment integrated into the EPR, which triggers a standardised antiemetic prescribing protocol. Trial for 10 high-risk patients on one shift. Predict improved prescribing compliance from 60% to 85% and a reduction in PONV for these patients.
• Do: Implemented. Prescribing compliance for identified high-risk patients reached 80%. One patient experienced mild PONV, but overall, the intervention appeared positive.
• Study: Prescribing compliance improved significantly. Early data suggests a reduction in PONV, but more patients are needed.
• Act: Adopt the revised risk assessment and prescribing protocol for all elective patients on the ward. Begin measuring the overall PONV rate and continue monthly data collection for three months to confirm sustained improvement.

How Lazomis can help

Lazomis offers practical tools that complement the Model for Improvement, helping NHS teams to execute their QI projects effectively:

• Project Setup Templates: Use structured templates to define your aim statements, identify key measures, and plan your PDSA cycles efficiently.
• Data Collection and Visualisation: Input your measurement data directly into Lazomis. Our dashboards can then help you visualise trends over time, making the 'Study' phase of your PDSA cycles clearer and more insightful.
• Learning and Documentation: Our platform provides a central repository for documenting your PDSA cycles, capturing learnings, and tracking the evolution of your improvement journey.
• Reporting Tools: Generate concise reports on your QI progress, ideal for sharing with wider teams, governance committees, or for educational portfolios.

By centralising your QI project elements within Lazomis, teams can spend less time on administration and more time on actual improvement work, making the iterative nature of the MFI more manageable.

Key takeaways