Background: Much has been written about the Internet's potential to revolutionize health care delivery. As younger populations increasingly utilize Internet-based health care information, it will be essential to ensure that the elderly become adept at using this medium for health care purposes, especially those from minority, low income, and limited educational backgrounds. Objective: This paper presents the results of a program designed to teach elderly adults to use the Internet to access health care information. The objective was to examine whether the training led to changes in participant's perceptions of their health, perceptions of their interactions with health care providers, health information-seeking behaviors, and self-care activities. Methods: Participants attended a 5-week training course held in public libraries and senior community centers within the greater Pittsburgh and Allegheny County region. Classes within each seminar lasted 2 hours and consisted of lecture and hands-on training. Baseline surveys were administered prior to the course, 5-week follow-up surveys were administered immediately after the course, and final surveys were mailed 1 year later. Instruments included the Multidimensional Health Locus of Control (MHLC) Scale, which measures three domains of locus of control (internal, external, and chance); the Krantz Health Opinion Survey (HOS); and the Lau, Hartman, and Ware Health Value Survey. Two additional questionnaires included multiple choice and qualitative questions designed to measure participants' Internet utilization and levels of health care participation. The Health Participation Survey was administered with the baseline survey. The Internet Use Survey was administered at the 1-year mark and contained several items from the Health Participation Survey, which allowed comparison between baseline and 1-year responses. Results: Of the 60 elderly adults who began the training course, 42 (mean age 72) completed the entire 5-week training program and the 5-week follow-up questionnaire administered immediately after the program, and 27 completed the 1-year follow-up survey. Statistically significant differences were found between baseline and 5-week follow-up results for MHLC chance subscores in males (P = .02) and females (P = .05), as well as total HOS information seeking scores (P = .05). However, these statistically significant findings disappeared when all 60 original participants were included using a "last observation carried forward" imputation. No statistically significant changes were found between baseline and 5-week follow-up surveys for MHLC external (P = .44) and internal (P = .97) locus of control scores in both genders, or for the HOS behavioral involvement subscale (P = .65). Conclusions: We failed to show robust before-after effects for most of the outcomes measured. Elderly adults may be willing to use the Internet as a source for general health information; however, when making decisions about their health care, our participants seemed to adhere to a physician-centered model of care. Demographic and situational variables may play a large role in determining which seniors will use the Internet for making behavioral decisions about their health care and in which scenarios they will do so.
