Decide what to improve

Name the thing that should change before building anything: a direction, a number that defines it, a target, and the kinds of outcome a care flow can move.

Care flows get built to change something, and telling whether the change worked depends on having named it first.

This page covers naming it; Know whether it worked covers proving it moved.

A goal names a measure, a metric and a target

"Improve patient engagement" is a direction, not a goal. A goal also needs a metric and a target, and different people usually set each part.

The measure is the plain-English direction, and it belongs to leadership. Reduce the cost of delivering care. See more patients with the same team. Meet a reporting requirement.

The metric is the number that defines the measure precisely enough to count. It usually belongs to the people closer to the work: directors, operations leads, clinicians. "Reduce the cost of delivering care" becomes staff time per episode of care, multiplied by what that time costs.

The target is the value and the timeframe. Hold the current standard. Reach a number by the end of the year. Change how the service works entirely. Each demands a different amount of upheaval, which is worth knowing before committing to one.

A target expressed only as a percentage hides a problem: a team that wants to improve something by a fifth but does not know today's figure has not set a target, only an ambition. Establishing the baseline is the first job, and often the most useful thing a first care flow does.

Use the metric someone else already defined

Where a regulator, a payer, or an accreditation body has already defined the metric, adopt theirs literally instead of inventing a proxy.

Adopting an existing definition removes the argument about what counts. It also makes the translation into money close to mechanical, since the payment or penalty attached to that metric is already written down.

What this looks like in practice depends on where a team operates:

Where care is paid for partly on quality, a national quality measure set already defines what counts and what each point of improvement is worth.
Where a health service or ministry mandates reporting, the requirement itself is the metric, and meeting it protects funding or accreditation.
Where a payer contract sets performance conditions, the contract names the number.

Not every metric is an outcome

CareOps practice watches several kinds of number at once. A clear goal says which kind it is aiming at.

Kind	Examples
Care flow performance	Completion rate, drop-off rate, time to finish a step
Financial	Cost per episode of care, cost per patient per clinician, revenue per patient
Clinical outcomes	Vital signs, lab results, control rates for a condition
Patient-reported outcomes	Validated questionnaires the patient completes themselves

All of them are worth monitoring, but only financial, clinical and patient-reported numbers answer "did care or cost improve".

Care flow performance metrics are easy to produce, which is why they get mistaken for the goal. Completion and drop-off rates are genuinely useful, because they usually explain why something is or is not working. But they describe how the process ran, not whether the patient or the budget is better off. Track them to diagnose, and report the outcome they serve.

The outcomes a care flow can move, and where the value shows up

Quality and compliance, care team capacity, and speed of change each translate into value differently.

Kind	What changes	Where the value shows up
Quality and compliance	The right care action happens for the right patient at the right time, consistently, with a record of it	Quality-linked payment, avoided penalties, protected accreditation
Care team capacity	Less time per task, so the same team carries more patients	Cost avoided, or more patients served without hiring
Speed of change	The gap between deciding a program is needed and patients being in it	Revenue that arrives sooner, and the ability to respond when rules change

What each looks like:

Capacity. Intake that staff used to collect by phone is collected by the care flow instead, so the same team takes on more patients without adding people.
Speed of change. The operations team adjusts the intake questions itself, rather than raising a ticket and waiting for an engineering cycle.
Quality and compliance. Every eligible patient gets the reminder, the appointment, and the follow-up, so a reporting requirement is met without anyone chasing it by hand.

Improve at the bottleneck, not next to it

When several things could be better, choose wherever work is piling up.

Improving a step downstream of the bottleneck only gives people more time to wait, while improving the bottleneck itself moves everything behind it.

In most care organizations the constraint sits around the clinician, because their time is the scarcest and most expensive thing in the system. Taking non-clinical work off clinicians is therefore the change worth making first.

Next steps

Next: Know whether it worked, to set up the measurement before the change goes live.