Remote patient monitoring (RPM) only helps patients if clinicians actually see the readings. If blood pressure, glucose or weight data sits in a separate portal, or is buried in a results tab clinicians rarely open, the integration can be technically complete and clinically useless.
So the real job is getting device readings into the EHR, in the right place, with the right time, and without noise. Here is how that works for Epic and Oracle Health (formerly Cerner).
Epic: three places RPM data can show up
Epic offers three main integration points, each for a different user:
- Flowsheet rows store time-series clinical data such as vitals. RPM readings go into designated "remote" rows, separate from in-office vitals, so clinicians can see both side by side. One catch: flowsheet row IDs are specific to each organization, so your team must work with the hospital's Epic analysts to get the right IDs for each site.
- MyChart. When readings are written correctly to flowsheet rows, they appear in the patient's MyChart health summary alongside office readings.
- Storyboard, the sidebar a clinician sees on opening a chart, can show the latest device readings and any active alerts through a custom activity.
Oracle Health: HL7 v2 results or FHIR
Oracle Health supports two main methods for external device data:
- HL7 v2 result messages. Many installations still prefer HL7 v2 for high-volume data, because their interface engines are already set up for it. Two details matter: the observation time must be the moment the device took the reading, not when the message was sent, and abnormal flags drive alerts in PowerChart.
- FHIR R4 Observation writes through Oracle's developer program.
In PowerChart, the source's recommendation is a dedicated "Remote Monitoring" results category that gathers all RPM vitals in one view.
One integration for both: FHIR-first
Rather than building separate integrations for Epic flowsheets and Oracle Health result messages, a FHIR-first approach writes standard FHIR R4 Observations. Both EHRs support these writes, which makes a single codebase possible. Each reading carries a standard LOINC code plus a reference to the device that took it, so its source is traceable. The original playbook shows a full example, including how to map each vital sign.
Four problems that break RPM integrations
1. Timestamps. A patient in the Eastern time zone takes a reading at 8:30 AM. The device records it in UTC, and the EHR server may run in another time zone. Handled badly, one reading can show three different times. The fix: store everything in UTC on arrival, keep the patient's time zone, and convert to the time zone each EHR installation expects when writing.
2. Duplicate readings. Patients often take two readings in quick succession, or trigger a scale twice. The fix: a deduplication window per vital type, for example 2 minutes for blood pressure and 1 minute for weight, keeping the last reading.
3. Measurement errors. Home readings are less reliable than office readings: a cuff over clothing, glucose measured right after eating, weight with shoes on. The fix: check readings for plausibility before writing them back, and never auto-file flagged readings. Send them to a clinical review queue instead.
4. Chart placement. Show RPM vitals on the same trend graph as office readings, with a marker showing the source. Put the latest reading and a 7-day trend where clinicians see it on opening the chart. Show how consistently the patient is sending readings, so clinicians can judge engagement.
How long it takes
The source's estimated total timelines:
- Epic: 6 to 9 months
- Oracle Health: 4 to 7 months
- FHIR-only EHRs: 2 to 4 months
Each includes registration, sandbox development, testing, vendor review and go-live, with go-live repeated at every customer site. Epic takes longest mainly because of the customer review, which covers security, clinical workflow and validation.
Before launch, check time zones, duplicate handling and the mapping for every vital at each site. The full playbook includes a pre-launch checklist covering technical items, staff training, alert thresholds and patient consent.
Nirmitee.io builds remote patient monitoring software and connects device data to Epic, Oracle Health and other EHRs. For mapping tables, example messages and the full checklist, read the original: Integrating RPM Data into Epic and Cerner on Nirmitee.io.
Dinesh Thorat is Healthtech Growth Partner at Nirmitee.io.
Originally published at nirmitee.io.
Top comments (0)