Willingness to Share Internet Use Data for Research on Early Disease Detection: Cross-Sectional Survey.

Willingness to share internet use data for medical research was high across all three health conditions studied.

• Overall willingness to share ranged from 74%-77% (95% CI 70.5%-80.3%) across cancer, heart disease, and depression conditions.
• Health app data showed particularly high willingness to share at 73%-76% (95% CI 69.8%-79.1%).
• The survey included 2390 UK adults with and without a history of cancer, heart disease, and depression using quota sampling.
• Participants were randomly assigned to one of three health conditions and provided with a pictorial example of internet use data.

Perceived benefits were strongly and consistently associated with higher willingness to share internet use data across all three conditions.

• Perceived benefits were associated with willingness to share with odds ratios ranging from 5.692 to 8.850 across all three conditions.
• All associations were statistically significant (all P<.001).
• This factor was identified through logistic regression analysis with α=.05 for each condition separately.

Concerns about data sharing were consistently associated with lower willingness to share internet use data across all three conditions.

• Concerns were associated with reduced willingness to share with odds ratios ranging from 0.343 to 0.432 across all conditions.
• All associations were statistically significant (all P<.001).
• Key concerns identified included data privacy, potential for misuse, and lack of relevance.

The pictorial example of browsing history did not affect willingness to share internet use data.

• Participants were provided with a pictorial example of internet use data as part of the survey design.
• Despite this concrete visualization, willingness to share was not altered by the example.
• This finding suggests that visual demonstration of the type of data involved does not significantly shift participants' data-sharing intentions.

Familiarity with internet use data was associated with lower willingness to share data for heart disease detection.

• Familiarity with internet use data was associated with lower willingness to share for heart disease detection (OR 0.740, 95% CI 0.561-0.976).
• This association was specific to the heart disease condition and was not consistently observed across all three conditions.

Asian ethnicity was associated with lower willingness to share internet use data for cancer detection.

• Asian ethnicity was associated with lower willingness to share for cancer detection (OR 0.234, 95% CI 0.076-0.723).
• This sociodemographic finding highlights a potential source of bias and inequality in datasets if data sharing is selective.

Younger age and male gender were associated with higher willingness to share internet use data for depression detection.

• Younger age was associated with higher willingness to share for depression detection (OR 0.975, 95% CI 0.951-0.999).
• Male gender was associated with higher willingness to share for depression detection (OR 2.615, 95% CI 1.511-4.526).
• These associations were specific to the depression condition.

Thematic analysis of open-ended comments identified key facilitators and concerns related to sharing internet use data.

• Key concerns identified were data privacy, potential for misuse, and lack of relevance.
• Suggestions to increase willingness to share included contributing to society and research, data security assurances, clarification of research purposes, and monetary incentives.
• Open-ended comments were analyzed thematically alongside the quantitative logistic regression findings.

Public understanding of internet use data was reported to be limited despite high overall willingness to share.

• The paper states that 'understanding of internet data is limited yet willingness to share for research is high.'
• This disconnect between limited familiarity and high willingness suggests participants may not fully understand what sharing their internet use data entails.
• The findings were described as the 'first large-scale assessment of public willingness to share internet use data for early disease detection.'