An AI care workforce for post-discharge follow-up.
OneDose runs post-discharge follow-up automatically. It reads the discharge summary, schedules the next 30 days of contact, and calls or messages each patient on Day 1, Day 7, Day 14 and Day 30 — checking medication, side-effects and recovery, and escalating to a clinician only when something genuinely needs one.
It runs inside the EMR and pharmacy systems a hospital already uses, so the follow-up plan lives against the patient record rather than in a spreadsheet beside it.
Why does the period after discharge go wrong?
The patient leaves with a prescription, a discharge summary and a plan — and then becomes invisible. Whether they collected the medication, understood the dose, or stopped taking it on day four because it made them dizzy is unknown until something goes wrong badly enough to bring them back.
The work that closes that gap is well understood: call the patient, ask structured questions, notice the answer that matters, escalate it. The reason it doesn’t happen is not clinical disagreement — it is that it is a lot of phone calls, and the people qualified to make them are the people you least want spending their day dialling.
So it gets rationed. Follow-up goes to the highest-risk cohort, and everyone else gets a leaflet.
How does OneDose run the 30 days after discharge?
Every patient gets a programmatic follow-up plan driven by their condition, prescription and risk profile. The care team sees what matters, when it matters.
- Day 0
Discharge
OneDose ingests the discharge summary, parses the prescription, and schedules the next 30 days of follow-up. Nobody builds the plan by hand.
- Day 1–3
Activation
The voice agent calls to confirm the patient collected the medication, understands the dose and schedule, and knows which side-effects matter — in their preferred language.
- Day 7
Adherence check
A structured check-in by call or SMS. Patient-reported outcomes captured, adherence recorded, and drift on any medication flagged.
- Day 14
Recovery signal
Symptoms triaged and vitals reviewed where a device is connected. The care manager is paged on signal — not on schedule, which is what causes alert fatigue.
- Day 30
Risk & refills
The refill window is anticipated before the patient runs out. Risk is scored, and high-risk patients reach a human with the full history attached.
Who decides when a patient needs a clinician?
A clinician does. The agent collects and escalates; it does not decide. Clinical questions, low-confidence cases and anything suggesting deterioration route to a human pharmacist or care manager with the full conversation attached — never silently dropped, and never answered by the agent on its own authority.
This is also why the escalation is on signal rather than on schedule. A system that pages the care manager for every check-in trains them to ignore it, which is worse than not paging at all.
Where does it run?
Inside the systems already in place. OneDose reads and writes against the EMR and pharmacy management system a hospital already runs, so the follow-up record lives against the patient rather than in a parallel tool someone has to reconcile later.
Every agent action is logged. Which specific systems are live in your region is a question worth asking us directly rather than reading off a page — see the integrations we can confirm for your estate.
Frequently asked
- What is post-discharge follow-up automation?
- Post-discharge follow-up automation uses software to run the structured check-ins a patient needs after leaving hospital — confirming they collected and understood their medication, checking for side-effects, and flagging deterioration to a clinician — without a nurse placing each call by hand.
- How does OneDose handle a patient after discharge?
- OneDose reads the discharge summary, parses the prescription, and schedules the next 30 days of contact automatically. The patient is called or messaged on Day 1, Day 7, Day 14 and Day 30. Each contact is structured, recorded against the patient, and escalated to a care manager only when something needs a human.
- Does an AI agent make clinical decisions?
- No. OneDose agents run with explicit guardrails. Clinical questions, low-confidence cases and anything indicating deterioration are routed to a human pharmacist or care manager with the full context of the conversation. The agent collects and escalates; the clinician decides.
- What languages can the follow-up agent work in?
- The voice agent contacts patients in their preferred language. Language coverage depends on deployment — ask us which languages are live for your region.
- Does OneDose replace our care management team?
- No. OneDose removes the manual calling and charting so the care team spends its time on the patients who actually need attention. The intended effect is fewer routine calls placed by clinicians, not fewer clinicians.