How might we think about mechanisms of action in digital health? When taking traditional medication, you take a particular dose that provides reliable symptom relief. Randomized control trial data inform the therapeutic dose. In digital health, how do we know an app works? What data drives outcomes? These are some questions emerging from this developing field.
A paper in Health Psychology by Megan McVay from Duke University explores dose-response in digital health interventions. Whether you are designing interventions for smoking cessation, weight loss, or diabetes management, there are evidence-based strategies that guide “what works” to support someone in their behavior change journey. If you were teaching this via an in-person class, you can pace the learning, read the room, and answer questions in real-time. How does this work in a digital intervention?
Designing to deliver a dose of intervention that leads to behavior change is critical while ensuring there is no harm resulting in that experience. The authors speak to three categories of behavioral prescriptions that support behavior change. This relates to what the intervention is asking the participant to do. Intervention action prescriptions are where the participants receive content to support their behavior change; for example, they may be asked to read an app notification. Participant action prescriptions are where the intervention asks the participant to provide content to the intervention, for example, their latest weight or physical activity levels, or respond to a text with information. The third is a behavioral targeted prescription where participants are asked to engage in behaviors outside of the intervention; these may include being asked to meditate twice a day or eat more fruit and vegetables.
The table above outlines the differences between the intended dose and the enacted dose. The intervention developer can design the intended dose: the duration, frequency, and amount of intervention a participant receives. In traditional research, this can shed light on what supports behavior change.
The enacted dose is in the participant’s hands. A parallel is seen in the pharmacology literature, the difference between looking at the ingested medication versus the drug’s specific blood concentrations. In the digital space, the same intended dose of intervention may result in differences in the enacted dose due to individual differences. Being able to parse out the differences will be important in determining which aspects of the intervention, and the subsequent participant actions add value to the health outcome.
Many of the digital therapeutic solutions on the market today offer mixed modalities; this may include sensor derived data, chatbots, coaching, and content delivered to support behavior change. While traditional literature can point us toward what works, digital can provide even deeper context given the continuous data collection opportunities digital affords. This space will evolve rapidly given the biggger players have maturity in leveraging continuous data collection.
Thanks for reading – Trina
(Opinions are my own)
McVay MA, Bennett GG, Steinberg D, Voils CI. Dose-response research in digital health interventions: Concepts, considerations, and challenges. Health Psychol. 2019 Dec;38(12):1168-1174.
doi: 10.1037/hea0000805. Epub 2019 Oct 3. PMID: 31580127; PMCID: PMC6861635.