QI Project Guide

Quality Improvement Project Management in Australian Healthcare

Running a quality improvement project in healthcare is rarely a question of methodology. It is a question of project management discipline. Most QI ideas in Australian hospitals, general practices, and aged care facilities are sensible. Most of them fail to deliver lasting change. The reason is almost always the same: weak scoping, vague aims, no measurement plan, and no strategy for sustaining the improvement once the initial enthusiasm fades. This guide covers how to plan, run, measure, and sustain a quality improvement project in an Australian healthcare setting, from the initial problem statement through to the day a year later when the change is still in place because it has been embedded properly.

Most quality improvement projects fail for predictable reasons: an unclear problem statement, no measurable aim, a team that does not actually own the work, or no plan for sustaining the change once the project closes. The successful ones look boringly disciplined by comparison.

Five failure patterns recur across Australian healthcare quality programs. The first is starting with a solution rather than a problem. A QI project that begins with “we should roll out a new handover form” has skipped the diagnostic step entirely. The second is an aim that cannot be measured. “Improve communication” is not an aim; “reduce omitted medications at handover by 50% within six months on Ward 3B” is. The third is no ownership: a project where everyone is contributing but no one is accountable for the result. The fourth is starting without baseline data, which makes it impossible to demonstrate later that anything actually changed. The fifth, and the most common, is treating implementation as the finish line rather than the halfway mark.

Successful QI projects share five characteristics. They start with a clearly defined problem grounded in incident data, audit findings, or staff and consumer feedback. They have a specific, measurable, time-bound aim. They have a small multidisciplinary team with named roles and clear accountability. They collect baseline data before any intervention. And they have an explicit plan for embedding the change into policy, training, and the audit schedule so it survives the people who ran the project.

For an overview of the methodologies used to structure the improvement work itself, see our guide to quality improvement methodologies in Australian healthcare. This article focuses on the project management discipline that wraps around whichever methodology you choose.

A complete QI project moves through five phases: identify the problem, plan the project, implement the change, evaluate the results, and sustain the improvement. Each phase has distinct outputs and a clear handover point to the next. Skipping a phase is the most common cause of projects that finish on paper but never produce lasting improvement in practice.

The phases do not need to be rigidly sequential in every project. Plan and Implement frequently interleave, particularly when PDSA cycles surface new information that reshapes the plan. What does not change is the requirement that each phase produces its own output and that none of them is skipped. A project without a charter is a project that will drift; a project without an evaluation is a project that cannot prove its results; a project without a sustainability plan is a project whose improvements will quietly evaporate within twelve months.

For most QI projects in a hospital ward, general practice, or aged care facility, the full cycle from problem identification through to confirmed sustainability runs 6 to 12 months. Larger system-level projects affecting multiple services or sites typically run longer.

The project charter is the foundation document that turns a QI idea into a project the clinical governance committee can sponsor. A well-written charter sets out the problem, the aim, the scope, the team, the timeline, and the measures of success in two or three pages. A charter that runs longer than that usually has a scope problem that needs fixing before the project starts, not after.

The problem statement is the most important section. It describes the gap between current practice and the desired state, supported by data. “Medication errors at handover are an ongoing concern” is not a problem statement; it is a hand-wave. “Audit of 60 medication handovers across Ward 3B in March identified omitted medications in 28% of cases, against a target of less than 5% set by the medication safety committee” is a problem statement. It names the issue, quantifies it, anchors it to a source, and references the standard it is failing.

The aim statement specifies what the project will achieve, by how much, by when, and where. Use the SMART format adapted for healthcare QI: specific, measurable, achievable, relevant, and time-bound. A workable aim looks like this: “Reduce omitted medications at nursing handover on Ward 3B from 28% to under 10% by 30 November 2026, sustained for three consecutive monthly audits.” The specificity is what makes the aim useful both for guiding the team during the project and for evaluating success at the end.

Scope defines what is in and what is out. A QI project that tries to fix medication handover, discharge summary quality, and admission medication reconciliation in the same six months is not one project; it is three, and it will deliver none of them. Define scope tightly: one ward or one practice, one process, one defined patient population. Expansion to other sites comes later, once the change has proven itself in the original pilot.

The team section names the project lead, the clinical sponsor, the data and measurement lead, and the frontline staff who will actually run the changes. Every named role has a defined contribution and a time commitment. Vague references to “the ward team” or “the clinical group” are a warning sign that ownership has not been worked through.

Timeline and milestones translate the aim into a sequence of checkpoints: baseline data collection by week 2, PDSA cycle 1 by week 6, mid-project review by week 12, post-intervention measurement by week 18, sustainability review by week 24. The timeline is not a guarantee, but it gives the clinical governance committee a basis for tracking progress and intervening if the project stalls.

Success measures define how the team will know the project has worked. The primary outcome measure is the headline number tied to the aim. Process measures track whether the intervention is being delivered as designed. Balancing measures track unintended consequences elsewhere in the system. We cover the measurement framework in detail later in this article.

The completed charter should be signed off by the clinical sponsor and approved by the clinical governance committee before implementation begins. That sign-off is not a formality. It commits the organisation to the project, allocates the resources, and creates the line of accountability back to the governance structure. For broader context on how QI projects sit within governance, see our overview of what is clinical governance.

QI projects in healthcare touch more people than they appear to at first glance, and the speed at which a project moves is determined less by the technical change than by how well the stakeholders have been engaged. A change that the affected staff understand, agree with, and have helped shape will roll out smoothly. The same change imposed without consultation will stall regardless of how good it is on paper.

Start with a stakeholder map covering four groups: the people whose practice will change, the people whose work will be affected downstream, the people who hold the resources or authority needed to make the change happen, and the people who can credibly block the change if they choose to. In a medication handover project, group one is the nurses doing the handover. Group two includes pharmacy, the medical team writing the orders, and the morning shift inheriting the patients. Group three is the nurse unit manager, the director of nursing, and the medication safety committee. Group four includes any senior clinician on the ward whose buy-in others will follow.

The engagement strategy then matches the stakeholder. People whose practice is changing need genuine consultation early, before the intervention is designed. They are also the source of insight into why current practice looks the way it does, and what will and will not work in their environment. People affected downstream need to be informed and consulted on interface points. People with authority need a clear case for action and confidence the project is being run responsibly. Potential blockers need to be brought into the design conversation early enough to either resolve their concerns or surface them transparently to governance.

Consumer and carer engagement is a specific requirement of the National Safety and Quality Health Service (NSQHS) Standards under Standard 2 (Partnering with Consumers). Where a QI project affects the patient experience or the way care is delivered to or with a patient, consumer representation in the project team is expected, not optional. The Australian Commission on Safety and Quality in Health Care publishes practical guidance on consumer partnership that translates well into the QI project context.

The most reliable mistake in stakeholder engagement is engaging too late. A project that has already designed its intervention and is now “rolling it out for input” has skipped the part where input was actually useful. Engagement that starts at the problem statement and runs through the design process produces interventions that are usually more practical and always more durable.

Once the project charter is approved, the team chooses a methodology to structure the improvement work itself. The three methodologies most commonly used in Australian healthcare are PDSA (Plan, Do, Study, Act), Lean, and Six Sigma. Each has a place; the choice depends on the type of problem and the organisational context.

PDSA cycles are the workhorse methodology for most healthcare QI projects, particularly in clinical settings. A PDSA cycle is a small-scale test of a change. The team plans a specific change to test, does it on a small sample (one patient, one shift, one clinic session), studies what happened against the prediction, and acts on what they learned. Successful changes are adopted, expanded, and tested at larger scale; unsuccessful changes are adapted or abandoned. PDSA suits situations where the right intervention is not yet known and learning is needed before scaling. Multiple PDSA cycles over a project allow rapid iteration without committing the whole organisation to an untested change.

Lean focuses on identifying and eliminating waste in a process. Waste in a healthcare context includes waiting time, unnecessary motion, duplicate documentation, defects requiring rework, and overproduction. Lean tools such as value stream mapping and standard work suit projects where the process flow is the problem: clinic wait times, admission delays, theatre turnover, and discharge process redesign are typical Lean applications. Lean projects tend to involve more upfront mapping and a larger initial intervention than PDSA-driven projects.

Six Sigma applies statistical methods to reduce variation in a process and is most useful where defects, errors, or unwanted variation are the central problem and the data volume is high enough to support statistical analysis. Pathology turnaround times, surgical site infection rates, and medication dispensing accuracy are areas where Six Sigma methods can add value. Six Sigma requires more statistical training and a longer project timeline than PDSA.

In practice, many Australian QI projects use a blended approach. The overall structure follows the project lifecycle; PDSA cycles drive the iterative testing within the implementation phase; Lean thinking informs how the team looks at the underlying process; statistical tools borrowed from Six Sigma support the measurement and evaluation. The choice of label matters less than the discipline behind it.

For a deeper look at how these methodologies compare and when each is best suited to particular healthcare contexts, see our companion article on quality improvement methodologies in Australian healthcare. For context on how QI projects intersect with the clinical audit cycle that often surfaces them in the first place, see our guide to clinical audit methodology. Where the trigger for a QI project is a specific incident or adverse event, the investigation typically uses root cause analysis first, with the resulting recommendations feeding into a QI project.

A QI project without a measurement plan is an opinion exercise. The measurement plan is what turns the project into something the clinical governance committee can act on and that other services can learn from. Three categories of measure belong in every QI project: outcome, process, and balancing.

Outcome measures track whether the headline aim is being achieved. In the medication handover example, the outcome measure is the rate of omitted medications. Outcome measures are typically expressed as a percentage or a rate per defined denominator (per 100 patient days, per 1,000 handovers, per admission). They are sampled at baseline, during implementation, and after the intervention has stabilised. Outcome measures usually lag the intervention by weeks because they measure the system effect rather than the immediate intervention.

Process measures track whether the intervention itself is being delivered as designed. If the intervention is a structured handover checklist, the process measure is the percentage of handovers where the checklist was used correctly. Process measures are essential because they let the team distinguish between an intervention that did not work and an intervention that was never properly implemented. A poor outcome with strong process measures means the intervention is wrong; a poor outcome with weak process measures means the team has an implementation problem to solve before drawing any conclusion about the intervention.

Balancing measures track unintended effects. Reducing one problem sometimes creates a new one elsewhere. Improving documentation may add to nursing workload; reducing readmissions may increase outpatient demand; tightening one infection control practice may displace risk to another setting. A balancing measure is whatever you would not want to get worse as a side effect of the project. Identifying balancing measures during planning forces the team to anticipate unintended consequences and acts as a credibility check during evaluation.

Run charts are the standard visual format for tracking QI measures over time. A run chart plots the measure on the vertical axis and time on the horizontal axis, with a median line showing central tendency. Run charts make it easy to see whether change has occurred, whether it has held, and whether the variation pattern has shifted. They are inexpensive to produce, easy for clinical teams to read, and effective at communicating progress to governance committees.

Control charts add statistical context. They include upper and lower control limits calculated from baseline data, allowing the team to distinguish common-cause variation (normal noise in the system) from special-cause variation (a meaningful shift requiring investigation). Control charts are particularly useful in projects where statistical significance matters and the data volume is sufficient. For most ward-level or practice-level QI projects, run charts are adequate; control charts add value where the project is system-wide or the measures feed into formal reporting.

Baseline data must be collected before any intervention. Two to three months of baseline measurement is typical for outcome measures with monthly sampling; daily or weekly process measures need a corresponding baseline window. Without baseline data, the project cannot demonstrate change at the end, regardless of how good the intervention actually was.

Sustaining a QI improvement is harder than achieving it. The literature on healthcare improvement is consistent on this point: a substantial proportion of QI gains erode within twelve to eighteen months of project closure, and the reasons are almost never technical. They are organisational. The change was not embedded into the way new staff are trained, the practice was not written into policy, no one was responsible for monitoring it after the project ended, and gradually the system drifted back to its previous behaviour.

Sustaining an improvement requires a deliberate plan, built into the project from the start, that addresses four embedding mechanisms.

The first is policy. The new practice should be written into the organisation’s policy or procedure documents, with version control and a review date. A change that exists only in the project team’s memory will not survive their next rotation. Updating the documented policy is the lowest-cost embedding mechanism and the easiest to skip.

The second is training and orientation. New staff joining the ward, practice, or service need to learn the new practice as the default, not the previous one. That means updating orientation packages, competency assessments where they exist, and any clinical education materials that touch the area. If the change is significant, building it into the relevant clinical update or annual mandatory training cycle is worth the additional effort.

The third is the audit schedule. The new practice should be included in the clinical audit program at a frequency proportionate to risk. Quarterly audits during the first year post-implementation, moving to annual once the change has demonstrated stability, is a reasonable default for moderate-risk practices. High-risk areas should remain on a tighter cycle indefinitely. The audit results feed back to the clinical governance committee as part of standard reporting, which keeps the practice visible.

The fourth is ownership transfer. The QI project team is, by definition, temporary. Before the project closes, ownership of the sustained practice should be formally handed over to a permanent role: the nurse unit manager, the practice manager, the clinical lead for the service, or the relevant governance committee. That handover should be documented, with the new owner accepting accountability for monitoring and for triggering action if compliance drifts.

A sustainability review at 6, 12, and 18 months after project closure is the final check. The review confirms whether the outcome measure is still tracking at the post-project level, whether the process measure compliance has held, and whether any balancing measures have changed. A review that surfaces drift triggers a corrective action plan rather than another full project; minor drift is normal and is the reason sustainability monitoring exists.

The complement to this article on the audit side is our guide to clinical audit methodology, which describes the audit cycle that monitors the sustained improvement. Many QI projects originate from audit findings and end with their improvements feeding back into the audit schedule, closing the quality loop.

QI results need to be presented in a format that clinical governance committees can act on, not in the academic-paper format that often gets produced by default. Three artefacts cover most reporting needs: a one-page project summary, a structured committee report, and a results visualisation for staff communication.

The one-page project summary is what gets shared widely. It includes the problem statement, the aim, the headline outcome (baseline measure, post-intervention measure, percentage change, confidence the change is sustained), the intervention in two or three sentences, and a single chart, usually a run chart, showing the trend. Anyone reading the one-pager should understand within two minutes what the project did and what it achieved.

The committee report is more detailed. It covers the same material as the one-pager but adds the methodology used, the process and balancing measures, the cost of the project, the sustainability plan, and any unresolved issues or risks. The committee report is what supports a governance decision about whether to scale the change to other services, transfer the practice formally into policy, or accept the project as closed. A committee report should be no more than four to six pages; if it runs longer, the team has usually included material better placed in an appendix or working document.

The staff communication artefact is what gets pinned to a noticeboard, included in a team meeting, or distributed in the service newsletter. It is the version designed to be read in 30 seconds by a nurse on a break or a GP between consultations. Visual clarity matters more than completeness: one chart, one or two sentences of context, the names of the team to thank, and a clear “what changes for you” message if the audience is staff whose practice has been affected by the project.

Data visualisation in QI follows a few simple rules. Run charts are the default. Two-axis charts and three-dimensional effects almost always reduce clarity. Pie charts are rarely the right choice for QI data, which is usually time-series. Colour should support, not decorate. A run chart with a median line, a clearly marked intervention point, and a few annotations explaining what happened at key turns is more useful than any amount of dashboard polish.

For practice managers tracking QI projects against operational metrics, our guide to GP practice KPIs and dashboards covers the broader measurement context within which most general practice QI work sits. For practices using QI activity as part of their accreditation evidence, our overview of the GP practice accreditation cycle in Australia shows where QI fits in the RACGP accreditation requirements. And for organisations preparing for hospital accreditation, the link between QI activity and the NSQHS Standards is direct: Standard 1 (Clinical Governance) and the seven clinical standards all reference quality improvement as part of the evidence base assessors will look for.

TalentMed (RTO 22151) offers the BSB50920 Diploma of Quality Auditing as a nationally recognised, 100% online qualification. The program covers project management for QI, NSQHS Standards, audit methodology, risk management, and the quality improvement frameworks discussed in this article. It runs over 12 months with daily intakes throughout the year.

TalentMed Pty Ltd, RTO 22151. The BSB50920 Diploma of Quality Auditing is nationally recognised on the National Register. Confirm current course duration, fees, and intake details on the course page before enrolling. Information in this article reflects published guidance of the Australian Commission on Safety and Quality in Health Care; refer to safetyandquality.gov.au for authoritative NSQHS Standards documentation.