A watch displays a heart-rate trace that rises during a walk. What does that number tell you, and what should happen next? A sensor can collect a signal, but information becomes clinically useful only when it answers a defined question. Devices may monitor, support treatment, or assist care in other ways; those roles are not interchangeable. A smart device still needs a reason to measure, a way to interpret the result, and a plan for what the result can responsibly change.
A sensor measures a signal, not a diagnosis
Wearables use sensors to measure signals such as heart rate, blood pressure, or glucose. Some transmit data to a phone, a record, or health professionals [1]. The sensor may measure a physical or chemical quantity, but a measurement alone does not establish the cause of a change. A heart-rate value, for instance, may vary with activity, rest, and other circumstances.
A clinical question gives the signal context: Is the goal to track a known condition over time? To detect a change that should prompt review? To guide a treatment already designed to respond to measurements? Without that question, a stream of numbers can be difficult to interpret. Digital health tools have potential to support prevention, treatment, and wellness, but the role depends on how the device and its information are used [1].
Monitoring and treatment are different jobs
A continuous glucose monitor illustrates a device linked to a defined treatment pathway. Its wearable sensor detects glucose levels, and in an automated insulin delivery system it can communicate with a pump that adjusts insulin delivery [2]. Here, sensing is connected to a system designed to act on the measurement. That does not mean every wearable has this role or can make treatment decisions.
Other devices mainly monitor. A wearable lung patch paired with a computer model has been studied for detecting wheezing from tiny vibrations; researchers are exploring whether the approach could help classify respiratory disease [1]. Detecting a sound pattern is a step in an evidence chain. It does not automatically prove a diagnosis, and a research device’s potential is distinct from routine clinical use.
The signal needs a reliable route
A useful device must capture the intended signal under real conditions. Airborne noise can interfere with digital stethoscopes listening for lung sounds, which is one reason researchers developed a patch sensor designed to detect vibrations with less distortion [2]. The example shows that engineering is partly a problem of separating the signal of interest from distractions in the environment.
Then the result must reach someone or something able to interpret it. Digital health technologies can send data to clinicians or caregivers, and smartphone applications may support monitoring and follow-up [1]. But transmission is not interpretation. A measurement needs defined thresholds or context, and the people receiving it need to know what action—if any—it supports.
Ask what decision the device supports
Before calling a device “smart,” ask what it measures, what question the measurement answers, and who is expected to respond. For a monitor, the purpose might be to observe trends in a known condition. For a treatment system, sensing may be linked to a specific intervention. Medical devices can monitor, modulate, or supplement bodily functions, but these are distinct functions [2].
This framing also helps separate promise from proof. A small study or pilot may show that a sensor can collect a signal or detect a pattern; that is not the same as showing improved health outcomes across routine care. The most useful device is not necessarily the one that measures the most. It is the one whose measurement is reliable, interpretable, and tied to a clear clinical question and appropriate response.
Sources & further reading
- Digital Health NIBIB
- Medical Device Technologies NIBIB
Prepared for the independent PSI Center reading library. About edition dates and source checking.
This explainer describes scientific principles. It does not assess individual treatment choices.
