The baseline has to exist before anything changes
It is the one part of a QI project that cannot be reconstructed. Once the intervention has run, a pre-intervention measure either exists or the project cannot demonstrate anything, however well the change itself went.
So the first question is always what you can measure now, from data that already exists, over a period long enough to be stable. Frequently that determines the project, because a promising idea with no obtainable baseline is not a project you can complete.
Process and outcome measures answer different questions
An outcome measure tells you whether the thing you cared about improved. A process measure tells you whether the change was actually implemented. Without the second, a null result is uninterpretable — you cannot tell whether the intervention failed or never happened.
Most programs want both, and the process measure is the one students forget. It is also usually the easier of the two to collect.
A change somebody could keep doing
Interventions that depend on a person being unusually diligent stop the day that person moves. What survives is a change to a default, a form, a workflow or a prompt, and QI rubrics increasingly ask about sustainability directly.
Say what would keep it running after your project ends, and be honest where the answer is that nothing would. Naming that is a stronger finding than claiming a durability the design cannot support.
One further note on scope. QI projects fail far more often from being too ambitious than from being too modest, and a small change measured properly is worth considerably more academically than a large one measured loosely. Programs are assessing whether you can run a cycle correctly, not whether you can transform a department. A completed small cycle also gives you something to write up honestly, which an abandoned large one never does.
Three steps, then the record keeps itself.
Send the problem and the setting
What you want to improve, where, and what data already exists that could serve as a baseline.
Fix the measures first
One outcome measure, one process measure, and a baseline captured before anything changes.
Plan for after
What keeps the change running once the project ends, stated honestly rather than assumed.
Queries on record.
Does a QI project need review board approval?
Often not, and the determination belongs to your board rather than to you. Many institutions have a short form for exactly this question. Getting a determination on file before collecting anything protects you if the project is later written up for publication, which is when the distinction between QI and research suddenly matters.
How long does the intervention need to run?
Long enough for the measure to show something beyond normal variation, which depends entirely on how often the thing you are measuring happens. A weekly event needs months; a daily one may need weeks. This calculation belongs at the design stage because it determines whether the project fits your program's timeline.
What if the measure does not improve?
Report it and interpret it, because that is a legitimate result and rubrics recognise it. This is exactly where a process measure earns its place: if the change was implemented as intended and the outcome did not move, that says something about the intervention. If it was never implemented, that says something else entirely.
Can I use data my organisation already collects?
Usually yes and it is the fastest route to a baseline, provided you have permission to access it for this purpose. Existing quality dashboards, incident reports and audit data are often exactly right. Confirm what governance applies before extracting anything, because access for your role does not automatically mean access for a project.
What is the most common reason these fail?
No baseline. Everything else is recoverable — a measure can be changed, an intervention adjusted, a timeline extended — but a pre-intervention measurement that was never taken cannot be created afterwards. It is the first thing to settle and the thing most often left until the change is already underway.
Elsewhere in the hall.
Send the assignment and the deadline.
A specialist credentialed in your own field reads what you actually have and replies within two hours with a plan and a price. The reading costs nothing, and if the honest answer is that we are not the right hall for it, you will get that instead.
