Skip to main content

Segments

Overview​

A segment is a reusable group of patients defined by the conditions they meet — for example, "patients with an HbA1c over 7 who also have an open diabetes care order." You build segments visually in the Segment Builder, see a live estimate of how many patients match, save them, and reuse them as the audience for a campaign.

Segments are live: the conditions are evaluated against your organization's current data every time they're used. When a campaign runs, it re-counts the audience, so newly matching patients are picked up automatically and patients who no longer match are skipped.

Opening the Segment Builder​

  1. Open Outreach → Segments. This shows every saved segment with its estimated patient count, when it was last modified, and whether it's currently used in a campaign.
  2. Click New segment to start a fresh segment, or click an existing segment to edit it.

From the list you can also search, rename, duplicate, or delete a segment, and start a campaign from one with Use. A segment that is part of an active or scheduled campaign is marked as in use and cannot be deleted until it's removed from that campaign.

Building a segment​

The builder has two halves: the rule canvas on the left where you define conditions, and the estimate panel on the right that summarizes the audience.

Conditions​

A condition is a single rule, made of three parts that adapt to what you pick:

  1. Resource — the kind of clinical or administrative record to look at: Patient, Observation, Service Request, Questionnaire Response, or Payment Reconciliation.
  2. Field / code — which value to test. For coded resources (like Observations) you pick a code using the code picker; for others you pick a field such as gender, state, or status.
  3. Operator and value — how to compare it: greater than, is, is not, contains, within the last N days, and so on. The available operators change to match the field type (number, money, date, text, or choice).

Code picker​

When a condition needs a clinical code, the code picker opens with two ways to find it:

  • Current data — the codes that actually appear in your organization's data, with how many records use each one. This is the quickest way to build a segment that returns results, because every code listed is guaranteed to exist in your data.
  • Code systems — search standard vocabularies (such as LOINC) by name or code.

Custom fields​

Fields your organization has configured on the patient record appear in the builder under their human-readable label (not a technical code), so you can filter on the data your team already captures.

Counting and presence​

Two condition types make it easy to target patients by activity:

  • Count of records — match on how many matching records a patient has, e.g. "at least 3 visits" or "more than 2 active orders."
  • Has / does not have — match patients who do or do not have a record of a given type, e.g. "has a completed intake questionnaire" or "has no recent lab result."

AND / OR groups​

Conditions live inside groups. A group set to ALL matches patients who satisfy every condition in it (AND); a group set to ANY matches patients who satisfy at least one (OR). You can nest groups to express more precise logic, such as "in the last 30 days AND (overdue for a screening OR missing a care plan)."

Joining resources​

Because conditions can reference different resource types in the same segment, the builder automatically links them by patient — so a segment can combine, for example, a Patient attribute, an Observation value, and a Service Request status, and only return patients who match across all of them.

The live estimate​

As you add or change conditions, the estimate panel recalculates the estimated patient count out of your total population. While it's recalculating, an Updating… indicator appears and the number dims so it's clear a new estimate is on the way. The panel also shows a plain-language summary of the active conditions and a guidance message when no patients match, suggesting ways to broaden the filters.

The estimate is useful for:

  • Resource planning — gauge how many patients an outreach will reach.
  • Targeting — confirm the segment is neither too broad nor too narrow before you use it.
  • Performance expectations — set a baseline for how a campaign should perform.

Saving and leaving​

  • Save is the primary action and is enabled only after you've made a change. Once a segment is saved with no further edits, the button shows Saved.
  • Cancel (or navigating away) while you have unsaved changes prompts you to confirm before discarding them, so edits are never lost by accident.
  • Use in Campaign hands the segment to the campaign builder as its audience.

Example use case​

Suppose you want to reach patients in your state who registered in the last month but haven't completed an intake form:

  1. Name the segment, e.g. "New patients — intake incomplete."
  2. Add a Patient condition: state is your state.
  3. Add a Patient condition: registered within the last 30 days.
  4. Add a Questionnaire Response condition: does not have a completed intake form.
  5. Keep the group on ALL so all three must be true, watch the estimate settle, and Save.

Duplicating segments​

To start from an existing segment:

  1. From the Segments list, open the menu (the three dots) on the segment's row.
  2. Choose Duplicate.
  3. The copy opens with all conditions carried over; rename it and adjust as needed, then Save.

The copy is independent of the original — useful for creating variations of a segment without affecting the one already in use.

Using a segment in a campaign​

When you choose Use in Campaign, the segment becomes the campaign's audience. The campaign re-resolves the audience from the segment each time it runs, and a patient's eligibility is re-checked at the start of each run and again after any Wait step, so messages go out based on the patient's eligibility as of that most recent check.