Veteran-Centered Digital Health Protocol

Recruitment for the semistructured interviews was conducted using purposive and snowball sampling at a large Veterans Affairs hospital in the Southeast United States. Veterans and their care partners were identified from the following sources: (1) on-site in services conducted with primary care teams, (2) web-based in services, (3) direct staff referrals, and (4) internal tracking of the report of current registrants. Potential participants were approached via phone. A subsample of the interview participants were asked to participate in the subsequent user testing phase. A subsample of three agreed to participate in user testing. One health care staff became aware of the project and volunteered to be part of the user testing for a staff perspective.
Inclusion criteria included veterans who were aged ≥18 years, who were registered My HealtheVet users, who had no cognitive impairment that prevented the use of a PC or the ability to engage in project activities, and who reported having a caregiver who assisted them with health care management. Inclusion criteria for care partners included those aged ≥18 years, who had no cognitive impairment that prevented the use of a PC or the ability to engage in project activities, and who reported providing caregiving assistance.
On the basis of qualitative sampling methods, saturation was anticipated to occur between 12 and 15 interviews for each VDT user type (ie, veteran and care partner) [22 ]. An overrecruitment strategy was used to allow for attrition. Up to 25 individuals representing each user type were recruited to ensure saturation across domains.