In this blog, we shortly present the reflections which lead to the selection of a suitable sensor system for the MedGUIDE solution. The first selection criteria is, which data is required in order to identify medication side-effects. The table below shows a list of common side-effect in dementia drugs as well as behavior based measurable changes or side-effects.
| Medication side effect | Measurable behavior change of side effect |
| Nausea, vomiting, diarrhoea | Increased number and duration of toilet usage |
| Loss of appetite | Reduced fridge usage, reduced time spent in kitchen at meal times |
| Muscle pain | Reduced mobility indoors, longer time spent in bed, less outgoing behaviour |
| Sleep problems | Increased bed exits during night, interrupted sleep, reduced global sleep quantity |
| Feeling tired | Increased daytime sleeping in bed, decreased mobility indoor |
| Headaches | Increased daytime sleeping in bed, decreased mobility indoor |
| Dizziness and feeling faint | Increased time staying in bed, decreased mobility indoor |
Based on these considerations, the following broad activities of daily living (ADL) were chosen to be able to detect the above changes, and corresponding medication side-effects.
| Selected activity | Explanation |
| Sleeping | Time spent asleep, including day-time naps |
| Total hygiene | Time spent using the bathroom and toilet |
| Community mobility | Time spent outside the home, for instance visiting a day care center |
| Feeding | Events of food intake |
| Functional mobility | Indoor mobility and activity |
This list of activities of daily living lead to the selection of a sensor network composed only of ambient sensors, requiring no interaction or wearing. The chosen system is provided by the Swiss company DomoSafety SA (www.domo-safety.com), which is recognized as a leader in the in-home behavior analysis, focusing on elderly care. The MedGUIDE consortium technical partners integrated this sensor system and test-drove it with success in three pilots.