Figure 2.

Patient characteristics associated with use, both among all included studies (entire bar) and within the subgroup of studies with more robust methodology (black). Studies that found no association (P ≥ .05) between use and patient characteristics were labeled “nonsignificant.” There were no tools that favored men or those with lower educational attainment, limited digital literacy, lower income, limited health literacy or numeracy, or English proficiency. Robust methodology was defined as meeting 2 of the following 3 criteria: (1) clearly included and reported characteristics of nonusers of the digital health solution, (2) included ≥50 participants in analyses, and (3) presented multivariable relationships to assess whether a characteristic was predictive of use holding all other characteristics constant. EP: English proficiency; LEP: limited English proficiency.