Quality Improvement Projects in Nursing: Models, Measures, Examples

A quality improvement project in nursing takes an existing process in a defined setting, changes it deliberately, and measures whether the change improved things. It is one of the most common formats for a Doctor of Nursing Practice project, and one of the most commonly misclassified.

Quality improvement, evidence-based practice, and research

Programmes are strict about this distinction and blurring it is the fastest way to have a proposal returned.

Quality improvement changes a local process and measures the effect, usually through iterative cycles. Its aim is better local care, and it typically does not require full ethics review, though institutional determination is still needed.

Evidence-based practice applies existing external evidence to change practice. The emphasis is on appraising the evidence base and translating it.

Research generates new generalisable knowledge and requires ethics approval.

The practical test is what you intend to do with the result. If the aim is to improve this unit, it is quality improvement. If the aim is to produce findings others should adopt, it is research. Projects that intend to publish generalisable conclusions while claiming quality improvement status to avoid ethics review are the ones that cause problems later.

If your project is really about appraising and applying external evidence, see evidence-based practice and PICOT projects instead.

Choosing a project that can succeed

The projects that work share three features: the problem is measurable, the process is within your sphere of influence, and baseline data either exists or can be collected quickly.

The most common failure is scope. A project to "improve patient satisfaction" cannot be completed in a semester because satisfaction is downstream of everything. A project to reduce call bell response time on one unit during one shift pattern can be.

The second most common failure is choosing a problem the unit does not agree is a problem. Improvement work requires people to change behaviour, and staff who were not consulted rarely do.

Measures: outcome, process, and balancing

Every project needs three kinds of measure, and the third is almost always missing.

The outcome measure is what you ultimately want to change: the fall rate, the infection rate, the readmission rate.

Process measures track whether the change is actually happening. If hourly rounding is the intervention, the process measure is the proportion of rounds completed. Without it, a project that shows no outcome improvement cannot distinguish between an ineffective intervention and one that was never implemented.

The balancing measure tracks what might get worse elsewhere. Reducing length of stay could increase readmissions. Increasing rounding frequency could displace documentation time. Including a balancing measure signals that you understand the change has costs, and its absence is one of the most reliable markers of an inexperienced proposal.

Baseline data and PDSA cycles

Any claim of improvement requires knowing where you started. Collect baseline data before changing anything, over long enough to capture normal variation. A single week's baseline in a process that varies by day of week will mislead you.

A Plan-Do-Study-Act cycle is a test of change, not a project phase. Each cycle plans a small change, runs it, studies the data, and decides whether to adopt, adapt, or abandon. Projects that describe their entire semester as "one PDSA cycle" have misunderstood the tool. Several small fast cycles produce better learning than one long one.

Where data is tracked over time, a run chart is usually more informative than a before-and-after comparison, because it shows whether change coincided with the intervention or was already underway.

Getting the unit on side

Improvement work fails on people more often than on method. A project that is methodologically sound but arrives as an instruction from someone passing through the unit will not survive contact with a busy shift.

Three things reliably help. Involve the staff who do the work in defining the problem, because they usually know where the real bottleneck is and will say so if asked. Make the new behaviour easier than the old one rather than adding a step. And feed the data back to the people generating it, visibly and often, since audit results that disappear upwards stop being taken seriously.

Where a project needs formal stakeholder analysis, name who must agree, who must be informed, and who can block, and plan for each separately.

Writing it up

SQUIRE 2.0 is the reporting standard for quality improvement work and is what journals expect. It asks for the local problem, the rationale for the specific intervention, the context, the measures, the analysis, and an honest account of what did not work.

That last element matters. Quality improvement write-ups that report unbroken success are less credible than those documenting a cycle that failed and what was changed as a result. The learning is the contribution.

Where this fits

Quality improvement projects are frequently the vehicle for a culminating assessment. See the Doctor of Nursing Practice project and capstone project support. For the evidence appraisal that often precedes the change, see evidence-based practice and PICOT projects, and for the guide to evidence-based practice models covering Johns Hopkins, Iowa, ACE Star, Stetler, and PARIHS.

For programme-wide support across nursing coursework, see nursing writing.