All use cases

Prevent / Care

Lab reminders and PRO surfacing to the care team

Your patient was meant to get labs done weeks ago and the care team has no idea how they're really feeling. The agent closes both loops without ever reading the results itself.

Get aheadTime-based triggerBehavioral triggerPushIn-appChatEmailGLP-1 & PeptidesHealth & Telehealth

What it does

Monitoring is where weight-medication programs quietly leak: clinician-ordered labs go undone, and the things that actually predict whether a patient stays, like mood and a shifting sense of self, never reach the care team. The agent reminds the patient to complete the labs their clinician ordered and chases the ones that slip, and it surfaces patient-reported outcomes such as mood and identity to the care team as structured signal. It deliberately does not interpret any result. Reading labs is a clinical act, so interpretation stays clinician-side and the agent holds the monitoring loop open around it.

How it works

  1. 1
    Trigger. a clinician-ordered lab comes due (or stays incomplete past its window), or a patient submits a check-in that includes mood or identity responses.
  2. 2
    Decision. the agent reads which labs are ordered and outstanding and which patient-reported outcomes are new, checks consent, and decides who to remind and what to route to the care team. It does not score, read, or interpret any lab value. Every patient-facing message passes the quality judge.
  3. 3
    Action. the agent reminds the patient to book or complete the lab on their preferred channel, and packages the patient-reported outcomes into a structured summary for the care team.
  4. 4
    Follow-through. outstanding labs are chased on a cadence until completed or the clinician stands them down; a patient-reported outcome that reads as a clinical or mental-health concern is flagged to a human rather than handled by the agent, and any acute self-harm signal triggers an immediate stop, a helpline plus a human, and loop closure on the patient's safety. The loop writes back so reminders stop once results are in.
Example
AgentYour clinician asked for a lab check this month and it's still open. Want me to pull up the nearest place to get it done, or send the order to your inbox? It keeps your care team able to look after you safely.
PatientI keep putting it off. Honestly I've been feeling pretty low lately too.
AgentThanks for telling me, that matters and I'll make sure your care team sees it. I'm not able to read lab results myself, but I'll get this booked so your clinician has what they need. Here's the closest slot this week.

Configuration

How the agent is wired for this use case.

Triggera lab_due or lab_overdue event from the lab system, or a submitted check-in carrying mood/identity patient-reported outcomes
Tools & actions
  • Clinical/EHR system · read which labs the clinician ordered and their status
  • Lab system · track outstanding draws and confirm completion
  • Scheduling system · help the patient book or locate a draw
  • Messaging channel · send reminders and capture check-in responses
  • Care-team queue · surface structured patient-reported outcomes (mood, identity) to clinicians
Autonomyreminders and PRO surfacing run unattended behind the quality judge. By design the agent never interprets a lab value; under the detect-then-escalate policy class it surfaces results and outcomes for the clinician and routes any concerning report to a human, and a self-harm signal halts the flow for an immediate helpline-plus-human escalation. No diagnosis, no result interpretation.
Channelspush · in-app · chat · email
Escalationa patient-reported outcome signalling a clinical or mental-health concern, or a repeatedly missed safety-relevant lab, hands off to the care team; any acute self-harm signal is a mandatory immediate hand-off to the crisis pathway and a human.

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

lab_due / lab_overdue events and submitted check-ins carrying mood and identity responses

Data

clinician-ordered lab list and status, patient channel preference, consent state, memory of prior reminders, the region-specific crisis-pathway resources

Guardrails

judge gating on every message; strictly no lab interpretation (a deliberately device-safe posture); detect-then-surface, with concerning outcomes escalated to a human; acute self-harm cues route immediately to the crisis pathway and a human; reminders cancel once results are in

Metrics it moves

  • adherenceup: chased labs get completed, keeping the monitoring loop closed and the patient safely on therapy
  • time-to-resolutiondown: care teams see the labs and the mood and identity signal sooner, without reading raw threads
  • csatup: the patient feels watched over without being handed numbers they can't read

Related use cases

Engage / Lifecycle

Test-result interpretation agent

the deliberate opposite: this card surfaces, that one interprets

Engage / Lifecycle

Scheduled health-questionnaire check-in cadence

the cadence that captures the outcomes this card surfaces

Prevent / Care

Titration-readiness clinician handoff

where the labs and outcomes feed a clinician dose decision

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo