Remote patient monitoring

Remote patient monitoring (RPM) means measuring a patient's body outside the clinic, at home or at work, with devices that send the readings over the internet to software. The software alerts a doctor or a nurse when a reading needs attention. The devices include glucose sensors, blood pressure cuffs, ECG patches and pulse oximeters, which measure the oxygen in the blood.

Here is how it works in general. A device on the body or at home takes a reading. Some take one reading a day, such as a blood pressure cuff. Others take a reading every few minutes, such as a glucose sensor. The device sends each reading to a phone app or a small hub, which sends it on over the internet. The software stores every reading with the time it was taken. This kind of data is called time series data. The software then compares the readings with limits that the doctor has set for this patient. More and more often, an AI model also looks for patterns that a fixed limit would miss, which is a form of anomaly detection. When a reading crosses a limit, the care team receives an alert. A nurse then calls the patient, or the doctor changes a dose.

Why do it? A patient with a long-lasting illness sees the doctor for a few minutes every few months. Readings from home show what happens in between. A patient sent home early from hospital can still be watched.

Take, for example, a person with diabetes who wears a continuous glucose monitor on the arm. It reads blood sugar every few minutes, and the app warns both the person and the doctor when the sugar falls too low.

Consider a patient who goes home three days after heart surgery. A patch from remote patient monitoring wearables sends pulse and breathing to the hospital, so a nurse can call if the readings get worse.

Say a drug company runs a trial in which patients stay at home. In a decentralised clinical trial, the readings from home devices go straight into the trial's records.

How is remote patient monitoring different from a fitness band? A fitness band shows the readings to the wearer only, and no doctor acts on them. In remote monitoring, a care team receives the readings and is responsible for acting on them. A Holter monitor is different too: it records the heart for a day or two, and a doctor reads the whole recording afterwards. Factories use the same idea for machines, in predictive maintenance.