Abstract
We evaluated the feasibility and acceptability of adding more detailed choices for race, ethnicity, sex, gender, and socio-economic status for a demographic survey used by families both within and outside a large learning health network, the Autism Care Network (ACNet). We updated our demographic survey using an iterative approach, incorporating qualitative and quantitative feedback from interested parties across the US and Canada. Pilot testing of the revised survey was conducted with families with and without autism served by two large academic pediatric tertiary care centers. Through purposive sampling, recruitment was enriched for families from ethnic, racial, or gender minority backgrounds. The updated demographic survey increased the number of response options for race and ethnicity, sex, gender, and language. 85 families within the ACNet and 242 families outside the ACNet provided feasibility and acceptability data. 41% of respondents were from nonWhite or multiple race groups. 99% of respondents rated the updated form same or better than the original. 91% of respondents rated the updated form as acceptable, while 97% rated the survey as feasible. Despite concerns about the burden on respondents, we found high rates of feasibility and acceptability of more granular response options in demographic surveys. Researchers can adapt this approach to make their own more granular demographic forms focused on the specific variables relevant to their study and local contexts. More granular demographic data can identify strengths and gaps in representation that could impact a study's generalizability.