We helped implement the so-called ‘just-in-time adaptive interventions (JITAI)’ feature for the DynaM-INT study. We implemented an intervention algorithm (developed by researchers from DynaMORE) using the RADAR-base’s real-time analytics pipeline to provide personalized mental health interventions.
Background
The DynaMORE (DYnamic MOdelling of REsilience) project consists of a multidisciplinary team of researchers that aim to improve the prevention of − or quick recovery from − stress-related mental health problems by developing a prognostic tool for people to monitor their mental stability and to intervene effectively before the personal “tipping point.” The researchers also want to better understand how different people handle difficult life situations and stress and to find out what protects them from developing mental disorders in the face of adversity.
DynaM-OBS is an observational study to collect multi-modal longitudinal real-world data sets from subjects exposed to everyday stressors. The DynaMORE consortium leverages the RADAR-base platform and app to collect a real-world data set, mainly using the active Remote Monitoring Technologies (aRMT) app to collect Ecological Momentary Assessment (EMA) data and Ecological Physiological Assessment (EPA) data extracted from IMEC Chill+ devices (feature data extracted and sent by IMEC).
The consortium uses this dataset to build an in silico model of stress resilience, which can serve as a base for predicting resilience and mental health disorders in humans and for the development of an algorithm to provide personalized, resilience-promoting interventions. These novel technologies are validated and tested in the DynaM-INT study, an interventional study with real-world monitoring.
The Challenge
To successfully apply and validate the developed intervention models for mental health, it is crucial that these models can identify in real time when an individual needs an intervention. It is equally important that the identified interventions are delivered to the individual within a given time to be able to test their effectiveness. Hence, a fully integrated solution is required that can collect data from subjects, detect time points for interventions based on the model, deliver an intervention, and capture feedback. To implement this solution in practice and integrate it with the RADAR-base platform, the DynaMORE researchers requested assistance from The Hyve’s RADAR-base team.
How we solved it
The Hyve’s team helped implement the so-called ‘just-in-time adaptive interventions’ (JITAI) into the RADAR-base platform for DynaMORE. The Hyve implemented the intervention algorithm that had been developed by the DynaMORE researchers using RADAR-base’s real-time analysis pipeline. It allows for instant detection of a stressed state that calls for an intervention. The interventions were then converted to push notifications and Ecological Momentary Interventions (EMI) in RADAR-base’s aRMT app (or questionnaire app).
The aRMT application was extended to receive these customized push notifications, deliver the intervention as a questionnaire on the app, and capture subjects' responses.
Another important requirement of the researchers was that they wanted to be able to selectively customize interventions for one participant or a group of participants. By customizing the model parameters per subject based on data acquired during a calibration week, i.e., two weeks prior to the intervention period, the scientists can provide personalized interventions.
The Hyve’s colleagues integrated the new application-configuration (app-config) tool and integrated it with RADAR-base’s real-time analytics pipeline to support the desired personalization.
The outcomes
We implemented a fully integrated solution to provide interactive, resilience-promoting mHealth interventions for the DynaM-INT study. The solution was initially validated by DynaMORE researchers. We then helped to integrate the newly developed components into the existing RADAR-base environment hosted by Donders Institute in Nijmegen, the Netherlands. The solution is currently used in the DynaM-INT study.
The DynaM-INT study is a proof of principle that validates and tests the effectiveness of model-based mobile health interventions. It is also a proof of concept and one of the excellent use cases of utilizing the real-time feedback framework of RADAR-base. The Hyve will continue to support and maintain the platform for the DynaMORE consortium throughout the DynaM-INT study.
Testimonial
"The DynaMORE consortium has chosen to use RADAR-base as its platform for building a real-time infrastructure for just-in-time adaptive interventions. We have had a very pleasurable and transparent collaboration with the Hyve. We could not have set up this infrastructure without their help."
- Prof. dr. Erno Hermans, Work Package leader and (co)PI in the DynaMORE consortium, Donders Institute
Acknowledgment: We acknowledge and thank our RADAR-base colleagues at KCL (PhiDataLab) for supporting us with parts of the implementation and validation of this real-time feedback framework of RADAR-base.