The short answer
The question is not what you can measure. It is who will look at the result, how quickly, and what they will do. Answer that first, because collecting data you cannot act on is worse than not collecting it.
Once information is in your system, not reviewing it is a position you have to defend.
Design the response, then the collection
- Decide which readings trigger a response and which are recorded only.
- Name who responds, in hours, including out of hours.
- Define what the response is at each level.
- Decide what happens when a patient stops sending readings.
- Only then choose devices and thresholds.
Point four is regularly forgotten. Silence from a monitored patient is itself information, and a system that only reacts to readings will never notice.
Alert fatigue is the failure mode
| Design | Outcome |
|---|---|
| Alert on every out-of-range reading | Muted within weeks |
| Alert on sustained change | Fewer, more meaningful |
| Personalised thresholds | Better, needs baseline data |
| Trend alerts rather than single readings | Catches deterioration, less noise |
A single out-of-range reading is frequently measurement error or a temporary state. Requiring persistence before alerting cuts volume substantially without missing much.
Adherence decides everything
A monitoring programme with poor adherence produces sparse data that is hard to interpret and easy to over-read. Making the routine simple matters more than the sophistication of the analysis.
- One device, one action, at a predictable time
- Automatic transmission rather than manual entry
- Feedback to the patient so it feels worthwhile
- A gentle prompt when readings stop, from a person where possible
- A realistic view of who can manage the technology
Be clear what it is not
Remote monitoring supports care between appointments. It is not continuous observation, and patients should understand that a reading is not being watched in real time unless it genuinely is.
Setting that expectation clearly at enrolment protects both the patient and the service, and it is usually a governance requirement as well as good practice.