External Factors and e-Health Adoption Among Older Adults Using the Internet: Structural Equation Modeling

Engagement in self-management of health and effective communication with health professionals are important. Moreover, the Health Information Technology for Economic and Clinical Health Act of 2009 encourages every American to review his/her personal health information (PHI) through the computerized U.S. healthcare systems.  Now, health consumers including older adults are expected to measure, store, and manage their own health data in the computerized systems. However, older adults lag far behind younger adults both in Internet use and in e-health adoption (Adler, 2006; Losh, 2010; Morris, 2007). Few studies examined older adults’ e-health adoption behaviors in relation specifically (1) to e-communication with healthcare practitioners and (2) to keep tracking of their PHI via online.

 To fill the gap, this study hypothesized, guided by the Technology Acceptance Model ([TAM], e.g., Davis, 1989), that external variables (user attributes—self-efficacy and individuals’ values/perceptions) indirectly determine older adults’ acceptance of e-health by influencing perceived benefits and perceived safety in using e-health.  

 Methods

Data on adults (≥55) from the Health Information Trends Survey collected in 2008 were selected. Of the unweighted 3,290 adults, the final sample of 1,800 (51%) respondents using the Internet was analyzed. The proposed hypothesis was tested using Structural Equation Modeling (SEM). Guided by the parsimony rule for SEM (Hoyle, 1995), the best-fit model is reported.

 The main measures include:

• DVs—e-health (binary):
  • (1) e-communication with healthcare practitioners; (2) track of PHI via online (for care received, test results, upcoming medical appoints, etc.).
• IVs—External Factors(3-5 Likert-type):
  • Path 1:
    • Personal Values: (1) Doctors should be able to share patients’ medical information with other health professionals; (2) patients should be able to get their medical information electronically; and (3) researchers/scientists can review patients’ medical information when no personal identifications are linked.
    • Self-efficacy: Confident in getting health-related advice/information; Confident in taking good care of health
  • Path 2:
    • Personal Perception/Trust: (1) Doctors are maintaining patients’ medical information in a portable, electronic format; (2) Doctors are guarding patients’ medical information safely.

Results

The SEM analyses partially support the proposed hypothesis: the path from self-efficacy to personal perception/trust does not support the model. Hence, this path and some covariates (age, income, race/ethnicity, nativity, health status) not being significant are removed for a parsimonious best-fit model.

The parsimonious model exhibited an excellent fit to the data (Chi-square=196.48, p=0.000, RMSEA=0.045, RMR=0.023, GFI=0.984, AGFI=0.964). The results indicates that older adults’ values (β=0.786, p<0.001) indirectly determine their adoption of e-health by influencing perceived benefits and perceived safety in using e-health (β=0.293, p<0.001).  The exogenous variables in the model include education (β=-0.039, p<0.001), sex (β=0.034, p=0.031), marital status (β=0.027, p=0.093), urban/rural (β=0.041, p=0.013), number of doctor visits (β=-0.010, p=0.011), and health information search (β=0.082, p<0.001).

Implications

Given valuing the benefits of e-health and trusting doctors safeguarding patients’ medical information affect older adults’ adoption of e-health, intervention efforts may be effective when collective efforts by policy makers, e-health system designers, and healthcare professional focus on maximizing its benefits and addressing health consumers’ concerns by promoting e-health systems.