A Portfolio Pathway QI project is a senior-led test of change with a measurable aim, planned PDSA cycles, balanced measurement and an honest write-up. It is not the same as audit. One well-documented project with completed cycles and visible change beats three thin posters every time. Most applicants build their project around the Model for Improvement, with a driver diagram, run-chart data and a short SQUIRE-style write-up in the evidence pack.
What QI proves in a Portfolio Pathway application
Quality Improvement evidence does something the other evidence categories cannot. Clinical audit shows you can hold practice up against a known standard and act on a gap. MSF shows how colleagues experience your work. Workplace-Based Assessments show direct observation of you doing the job. A QI project shows something different: that you can identify a problem in your service, frame it in measurable terms, recruit a team to test changes, manage the inevitable mess of trying things that fail, and bring something better out the other side.
That is Consultant-shaped work. It is one of the few evidence types that pushes squarely into the leadership and systems-thinking domains of Good Medical Practice, and into the higher-order Capabilities in Practice that distinguish a senior doctor from a competent registrar. It is also one of the few evidence types that demonstrably involves other people: nurses, allied health professionals, managers, patients. Done well, it is unmistakable senior practice.
How heavily QI is weighted varies by specialty. The Royal College of Radiologists treats QI as a specific curriculum requirement, with one to three projects across the application window. Other Specialty Specific Guidance documents describe QI more loosely as part of broader improvement and governance expectations. Before planning a project, check your specialty SSG for the exact framing.
QI vs audit vs service evaluation
The single most common QI mistake at panel level is mislabelling. A doctor presents an audit as a QI project, or a service evaluation as either. Assessors notice. The methodology has to fit the question. All three are legitimate evidence types in their own right, and the Portfolio Pathway can accommodate all three. They are not interchangeable.
Are we meeting the standard?
"Are we doing what national guidance says we should be doing?"
- Compares current practice against an agreed external standard
- Standard usually from NICE, Royal College or national audit
- Closed loop: measure, act, re-measure
- Outputs: gap analysis, action plan, re-audit data
- HQIP is the lead body
What change can we test?
"How can we make this safer, faster, more reliable?"
- Tests change ideas iteratively, usually via PDSA cycles
- Aim statement, driver diagram, balanced measures
- Local, continuous, often multi-disciplinary
- Outputs: run charts, learning from each cycle, sustained change
- IHI and NHS England QSIR are the methodology references
How is this service performing?
"What does our current service look like, descriptively?"
- Describes how a service is functioning right now
- No comparison to external standard, no testing of change
- Useful for scoping or as the precursor to audit or QI
- Outputs: descriptive report, activity data, recommendations
- Often the genuine activity hiding behind a mislabelled "audit"
A useful internal check before you start: write down the question your project is answering, in one sentence. If the sentence is "are we doing what the guideline says?", it is audit. If it is "what change can we make to improve this?", it is QI. If it is "what does our service actually look like right now?", it is service evaluation. Be honest with yourself early. Mislabelling does not make the work stronger and it confuses the assessor.
The Model for Improvement
The framework most NHS QI work uses is the Model for Improvement, developed by Associates in Process Improvement and adopted as the foundation of the Institute for Healthcare Improvement's methodology. The Royal College of Physicians' QI Hub, the East London NHS Foundation Trust QI programme, and the NHS England Quality, Service Improvement and Redesign (QSIR) toolkit all build on it. If a panel sees you have used the Model for Improvement, they know the language and they know the structure.
The model has two parts. The first is a thinking part: three questions to answer before you start. The second is a doing part: the PDSA cycle, run repeatedly to test changes.
The PDSA cycle, the doing part, sits inside the model. Plan a small test. Do it. Study the data and what you learned. Act on the learning by adapting, abandoning or adopting the change. Then run the next cycle. The cycles are deliberately small, especially at the start. The goal is rapid learning, not rapid implementation. A project with one large untested change is usually weaker evidence than a project with three small linked PDSAs, even if the headline result looks similar.
The senior-doctor twist: what assessors expect from you
Most NHS QI training is pitched at FY and core trainee level. The textbook examples are the hand-hygiene project, the early-warning-score project, the discharge-summary template project. Useful work, but at trainee scope. A Portfolio Pathway applicant trying to evidence Consultant-shaped practice should not submit a trainee-shaped project as their flagship QI evidence. The shape of senior-led QI is different.
Trainee-shaped QI
- Single ward, single intervention
- One person doing most of the data work
- Process-only measures, short window
- Outcome: poster at local meeting
- Reflection focused on individual learning
- Project ends when the rotation ends
Senior-led QI
- Service- or pathway-level scope
- Multi-disciplinary team you assembled and led
- Balanced measures over six to twelve months
- Outcome: change embedded in clinical governance
- Reflection focused on leadership, systems and culture
- Plan for sustainability after you step back
None of this means a senior project has to be vast. It means a senior project should look like a senior doctor ran it. The visible signs of that are usually: a defined team with named roles, clinical governance registration, sign-off from a clinical lead or service manager, balanced measurement rather than process-only counts, more than one PDSA cycle, and a plan for what happens to the change after the project formally ends.
Driver diagrams: making the theory of change visible
A driver diagram is a one-page picture of how your project will work. It puts the aim on the left, lists the primary drivers that will move the aim in the middle, the secondary drivers underneath, and the specific change ideas you will test on the right. NHS England, the IHI and most NHS QI teams treat it as the standard way to externalise a theory of change before any data are collected. Including a driver diagram in your evidence pack is one of the cleanest signals to an assessor that your project sits inside a recognised methodology rather than being assembled retrospectively.
Reduce time to first dose of sepsis antibiotics on the AMU.
From median 92 minutes to under 60 minutes within 6 months, for all NEWS 5+ admissions.
A good driver diagram does three things. First, it makes the project legible to anyone joining the team mid-flight. Second, it forces you to articulate the causal chain from change idea to aim, which exposes weak links before you waste a cycle on them. Third, it gives the panel a single image that says "this is a structured QI project". You can sketch your first version on a whiteboard with your team in an hour.
Measurement: process, outcome, balancing
The Model for Improvement is explicit that a QI project needs at least one measure from each of three types. Most weak QI projects fail here, by collecting only process data and presenting it as if it answered an outcome question. The QSIR Measurement for Improvement guide makes the typology clear.
Did the change actually happen?
What proportion of eligible patients got the new pathway? Did the bundle get used? Process measures tell you whether the test was carried out, not whether it worked.
Did the thing you cared about improve?
Did time to antibiotics fall? Did the readmission rate drop? Outcome measures are usually the slowest to move, and the ones panels and patients care about most.
What might you have broken?
Did downstream waits get worse? Did another team's workload spike? Balancing measures catch the unintended consequences that audit alone tends to miss.
Display measurement over time rather than as before-and-after points. A run chart with weekly or monthly data is the QSIR-recommended display because it shows variation, trends and the effect of cycles. Two bar charts at month zero and month six can hide the most important features of a real improvement, particularly the wobbly stretch after each change is introduced.
Choose the smallest measure set that still lets you tell the story. One process, one outcome, one balancing is enough for most projects. A spreadsheet with twelve metrics that nobody updates is worse than three metrics tracked weekly. The work of measurement compounds. Start small.
Running PDSA cycles in practice
The discipline of running good PDSAs is mostly about being deliberate. Each cycle should have a hypothesis, a small scope, a data plan and a date. The cycle is over when you know whether the change moved the measure, not when the rotation moves on.
State the hypothesis
Write the change idea as a testable sentence. "If we add a NEWS-triggered sepsis prompt to triage, time to first antibiotic will fall in patients screening positive at triage." If you cannot write the sentence, the test is not ready.
Plan the smallest credible test
One nurse, one shift, three patients. The goal of an early PDSA is learning, not proof. Many projects waste months on the first cycle because they tried to run it on the whole unit.
Do the test, collect the data
Run the change. Capture not only the planned data but also what surprised you. Surprises are usually where the learning is.
Study what happened
Plot the data on your run chart. Compare to your prediction. Where did the team push back, and why? Where did the workflow break?
Act on the learning
Choose one of three: adapt the change for the next cycle, abandon it because the theory was wrong, or adopt it more widely. Adopting is rarely the right choice after a single cycle.
Plan the next cycle
Set the date now. Linked cycles are what turn a single test into a project. A QI write-up with three linked PDSAs reads as a methodology. One large cycle reads as a one-off intervention.
The 6-month QI plan
Most senior-led NHS QI projects take six to twelve months from registration to a meaningful write-up. Six months is a working minimum for a project with three linked PDSAs and balanced measurement. Twelve months is more typical when the change needs to embed into governance and survive a few rota cycles.
The rhythm matters more than the speed. A project that runs three deliberate cycles across six months is better evidence than one that runs nine cycles across two months and then stops. The pattern an assessor wants to see is a project that breathes: data, learning, adapting, data, learning, adapting.
Writing it up for the portfolio
The international reporting standard for QI work is SQUIRE 2.0 (Standards for Quality Improvement Reporting Excellence). It was developed through formal consensus and is used by BMJ Quality & Safety and other journals as the standard for QI publications. You do not need to publish to use the SQUIRE structure. Adopting its headings for your portfolio write-up signals methodological literacy and gives assessors the predictable shape they are used to reading.
A practical SQUIRE-style structure for a portfolio write-up is roughly: background and problem; the local context; the specific aim statement; the change ideas and why they were chosen; measures and how they were collected; results, ideally with a run chart; what you learned, including what did not work; what was sustained; and a short reflection on your role and what you would do differently. Three to four sides of A4 is usually enough. The evidence pack itself, including driver diagram, run chart and PDSA logs, sits behind the write-up.
The strongest signal of a weak QI write-up is a strong headline result with no narrative about what failed along the way. Real QI projects fail in places. Cycles get abandoned. Teams push back on changes. If your write-up reads as a perfectly executed plan, assessors will assume it has been retrospectively tidied. Honest, balanced reflection on the messy bits is part of what makes the evidence credible.
Common mistakes that weaken QI evidence
Most weak QI evidence fails for predictable reasons. The pattern is consistent across specialties and at every level of seniority.
- Mislabelled methodology. The project is labelled as QI but is actually a service evaluation or a closed-loop audit. The mismatch undermines everything else in the pack.
- No driver diagram or aim statement. Without a written aim and a theory of change, the project reads as a sequence of unrelated activities rather than a structured improvement effort.
- Only process measures. Counting how often the new bundle was used, with nothing on what it achieved or what it broke, fails the second question of the Model for Improvement.
- Single huge cycle. One large change, implemented at the start, with a before-and-after data summary at the end. This is not PDSA. It is a small intervention study presented as QI.
- No balancing measures. The project shows the target metric improved but says nothing about whether anything else got worse. Senior assessors are particularly alert to this gap.
- Junior-shaped scope. A ward-level project run by you alone, with no team, no governance touchpoint and no sustainability plan. Acceptable as an early piece but weak as flagship senior evidence.
- Hidden failures. Polished narrative with no acknowledgement that anything failed or pushed back. Assessors read this as retrospective tidying.
- Disappearing change. The project ended, you moved on, and the improvement quietly reverted. No write-up should claim sustained change without evidence of it persisting beyond your direct involvement.
Mapping a QI project to the GMC domains
One project, well-documented, contributes evidence across all four GMC domains. This is part of why QI is one of the highest-yield evidence categories. A modest project mapped properly can populate more of your evidence map than several thinner pieces in other categories.
Domain 1 (Knowledge, skills and performance) is evidenced by the methodology itself: using the Model for Improvement, designing measures, interpreting run-chart data, adapting based on results. Domain 2 (Safety and quality) is evidenced by the choice of project, the balancing measures, and any patient-safety improvement that resulted. Domain 3 (Communication, partnership and teamwork) is evidenced by the multi-disciplinary team you led, the stakeholders you engaged, and the governance dialogue you maintained. Domain 4 (Maintaining trust) is evidenced by transparency about what failed, your willingness to abandon ideas that did not work, and the honesty of your reflective write-up. Make these links explicit in your evidence pack rather than leaving the assessor to infer them.
QI projects you led overseas can be included in a Portfolio Pathway application, but they are typically weighed alongside UK methodology rather than as a direct substitute. The pragmatic route for most successful applicants is to add a UK QI project to the file from inside an NHS post, using the Model for Improvement so the methodology is recognisable to assessors. Overseas projects sit as supporting evidence of longitudinal QI capability.
Download the PDSA project planner
A two-page A4 planner. Page 1 is a blank driver diagram template with prompts. Page 2 is a PDSA cycle planner with a run-chart grid and a short SQUIRE-style debrief block. Print, fill in with your team, file in your portfolio.
The PDSA project planner
Driver diagram template, PDSA cycle planner, run-chart grid and SQUIRE-style debrief prompts.
Where this sits in the evidence cluster
QI is one of the four cornerstone evidence types alongside audit, MSF, WBAs and reflective practice. None of these is sufficient on its own. Read them as a set, and build evidence in parallel rather than sequentially.
If you are a non-substantive Consultant or senior SAS doctor in a department where QI is informal or under-resourced, a credible project is harder to build than it should be. That is a real constraint, not a personal failure. The pragmatic path is usually to find one colleague who runs QI well, register a project jointly with their team, and use the structure of the Model for Improvement to make sure the methodology is on the page even if the local infrastructure is thin.
Official sources used
| Publisher | Source |
|---|---|
| Institute for Healthcare Improvement | The Model for Improvement (three questions and PDSA) |
| NHS England / Aqua | Quality, Service Improvement and Redesign (QSIR) toolkit |
| NHS England / Aqua | QSIR: Driver diagrams (PDF) |
| NHS England | QSIR: Measurement for Improvement (PDF) |
| HQIP | Guide for clinical audit, research and service review |
| SQUIRE | SQUIRE 2.0 reporting guidelines for QI publications |
| Royal College of Physicians | RCP Quality Improvement Hub |
| GMC | Good Medical Practice (2024) |
| GMC | Quality Improvement Activity (revalidation supporting information) |
| Royal College of Radiologists | RCR Audit and Quality Improvement |
| Future Healthcare Journal | The RCP Quality Improvement Hub: how it can help physicians (Choudry et al., 2016) |
Frequently asked questions
What is a quality improvement project in healthcare?
A QI project is a structured attempt to improve a clinical process or outcome by testing changes and measuring whether they help. Most NHS QI projects use the Model for Improvement, which combines three questions (What are we trying to accomplish? How will we know that a change is an improvement? What changes can we make that will result in improvement?) with the Plan-Do-Study-Act cycle. QI is distinct from clinical audit, which measures current practice against a known standard, and from research, which seeks to generate generalisable new knowledge.
Can a QI project replace an audit for the Portfolio Pathway?
Sometimes, but check your Specialty Specific Guidance before assuming this. Some Royal Colleges and Faculties accept QI projects in place of clinical audit, others expect at least one completed audit cycle or clear evidence of implemented change. The safest approach for senior applicants is to include both: a closed-loop audit and a QI project with clear PDSA cycles. They evidence different competencies, which is the point.
How many QI projects do I need for a Portfolio Pathway application?
There is no single universal number across specialties. Royal College guidance varies. The Royal College of Radiologists, for example, suggests one to three clinical audits or QI projects across the application window. For most senior applicants, one well-documented project with completed PDSA cycles and measurable change is stronger evidence than several shallow project summaries. Quality of evidence beats volume.
What is a PDSA cycle?
A PDSA cycle (Plan, Do, Study, Act) is the iterative test of change that defines the Model for Improvement. You plan a small test, carry it out, study the data, and act on what you learned by adapting, abandoning or adopting the change. Most credible QI projects involve several linked PDSA cycles rather than a single attempt. Each cycle should generate data, learning and a clear next step.
Do I need ethical approval for a QI project?
QI projects typically do not require formal NHS research ethics committee approval, because they aim to improve local care rather than generate generalisable new knowledge. However, they do need local governance: project registration with your clinical governance or quality team, agreement on data handling and patient confidentiality, and discussion with your clinical lead. If the design is testing a treatment, randomising patients or comparing care models in a way that produces transferable evidence, that may cross into research and need formal ethics review.
What is a driver diagram and do I need one?
A driver diagram is a one-page visual that links your project aim to the primary drivers that will move the result, the secondary drivers underneath them, and the specific change ideas you plan to test. It is a recommended NHS England and IHI tool for building and communicating your theory of change. You do not strictly need one for every project, but including one in your evidence pack tells assessors you used a recognised QI methodology and thought clearly about how change would happen. For a senior-led project the answer is usually yes, build one.