Around the world, less than 50% of the medicine are taken as prescribed (Xie, 2008). Out of these, only about 2 in 10 prescriptions are actually refilled at the pharmacy.
The anual costs of people not taking their medicine properly is estimated to be 290 billion dollars (Cutler, 2018), without mentioning the physical and emotional repercussions.
Existing interventions that focus on reminders often lack detailed insights into people’s daily intake routines. Ubiquitous sensor systems can facilitate the observation of everyday behaviors and provide detailed insights into medication routines.
We explore the routines and behavior of the user through a combination of quantitative sensor data and qualitative interview sessions. To this end, a sensor-enriched pillbox attachment was designed and deployed, to gather data on the movement and opening times of the participant’s pillbox. The researchers’ findings were discussed with the user, uncovering novel insights into medical adherence.
The study uses an augmented pill box to gather data from the users and avoid biases and "white coat" effects. The device is designed to fit in the everyday life of people who use medicine chronically. The custom made enclosure seamlessly integrates with the Anabox Daily pillbox, one of the most popular models in the Netherlands.
The research tool uses a set of sensors that store data locally, to avoid possible privacy issues. An accelerometer/gyroscope combo tracks the movement and orientation of the pillbox. A micro switch detects whether the box is open or closed and a clock keeps track of the time stamps. All of the sensors are controlled via an ESP32 microcontroller.
To evaluate the user experience and the technical implementation of using this sensor-enriched pillbox, a small feasibility study was set up. A middle-aged participant was recruited who had multiple medications on a daily basis, with an intake schedule of three times a day. In an intake session, the participant was explained how to use the Anabox. The sensor-enriched add-on did not need to be charged or removed. The participant was encouraged to follow their regular intake routines.
The pilot lasted for seven days during which all sensors collected data for six days because the battery ran out before the last day. This was most likely caused by a flawed implementation of the “deep sleep” mode of the ESP. The sensors successfully collected data about the movement of the box, the orientation, average rotation speed and the state of the box; open or closed. After this week, the sd-card with the data was received from the participant and visualized by creating figures to help discuss behavior patterns with the participant.
Despite its limitations, certain insights were revealed, even highlighting behaviors the participant was not aware of. Moreover, the study shows that a Data Enabled Design methodology can serve as a way forward in designing interventions for increasing the medicine adherence, while using data as a creative material.
The Data-Enabled Design method can offer researchers and designers a novel approach to build towards a better understanding of daily routines and which can eventually contribute to design interventions to increase medication adherence. The data can provide a better understanding about the daily routines and also help to test the effects of different concepts in the future.