The Living Lab at Home: Feasibility and Acceptability of Multimodal In-Home Data Collection Among Youth Across the Developmental Spectrum

Abstract

Objective: Dynamic, real-time, in-home methods of data collection are increasingly common in child health research. However, these methods are rarely cocreated or used with families of youth with developmental disabilities. We aimed to determine the feasibility of codesigned methods for in-home data collection for youth across the developmental spectrum.

Methods: Sixteen youth (14-18 years) with autism spectrum disorder, cerebral palsy, and/or chronic pain completed 14 days of data collection, wearing an accelerometer, answering Ecological Momentary Assessment (EMA) questionnaires, and collecting salivary cortisol samples. Participants completed a poststudy interview regarding their experiences. Data were analyzed for feasibility, quantity, and quality.

Results: At least 1 EMA response was provided on 73% of days, with 54% of the total number of administered prompts answered before the next prompt arrived. In total, 77% of participants wore the accelerometer ≥10 hours for at least 7 days. Adherence to 8-day saliva sampling after accounting for protocol violations and dry samples was 28%. No significant adverse events were reported aside from mild emotional distress (25%). Families reported generally high satisfaction, willingness to participate again, and acceptability, with moderate burden and interference. Qualitative interviews described: (1) the research question's value to the family as a motivator of engagement; (2) in-home data collection is not a passive or neutral experience; (3) personalized approaches and context are important to families; and (4) a clear need for continued iteration and engagement.

Conclusion: In-home multimodal data collection is potentially feasible for families across the developmental spectrum but requires iteration based on family feedback to increase adherence.

Figure 1.

LLAH daily data collection overview. Saliva samples were collected for the first 8 days on awakening, 20 minutes after awakening, and in the evening. EMA questionnaires were delivered to participants' smartphones 3 times a day for 14 days, at the time selected by the participant. Accelerometer was worn throughout wakeful time for 14 days. Images include graphics from flaticon.com. EMA, ecological momentary assessment.

Figure 2.

Adherence with each element of LLAH computed as the proportion of the expected amount (days) of data. A, Adherence by day of study, across participants, and does not include those who opted out of a particular element (1 for accelerometer, 3 for saliva). The black Xs indicate where no data was collected as the protocol dictated saliva collection for only the first 8 days. B, Adherence by participant, across days of study, and grouped by diagnosis. The black X (n = 4) indicates where no data was collected as the participant opted out of the modality before starting data collection. The gray Xs (n = 3) indicate where no data was available as the device was lost or no data could be extracted. C, Number of modalities of data collection with partial/complete data (at least 1 sample/data point/hour of accelerometer wear) by each participant on each day of study. Note that as saliva was only collected for the first 8 days, the maximum number of modalities present from days 9 to 14 was 2. LLAH, living laboratory at home.