Purpose: The purpose of this study was to determine if the Integrated Behavioral Model (IBM) was useful in explaining and predicting American adults’ behavioral intentions to complete end-of-life care planning. Of specific interest was whether the IBM could help to explain racial/ethnic differences in end-of-life care planning.
Methods: The study featured a theory-based, non-experimental, cross-sectional, survey research design. Participants were American adults between the ages of 40 and 80 years old. A geographically proportional sample based on the US census regions was randomly selected. An a priori sample size estimate indicated that a minimum of 384 completed surveys was needed to generalize the results to the US population of adults ages 40-80. Oversampling of racial/ethnic minorities was done to ensure that Blacks and Hispanics were adequately represented in the sample. To compensate for non-responses, a total of 1,650 American adults were included in the final sample. The IBM and the Precaution Adoption Process Model (PAPM) were used to design a valid and reliable survey. Prior to data collection, an elicitation phase featuring interviews and focus groups was conducted to ensure that the survey items were based on attitudes, beliefs, and perceptions that were highly salient to the priority population. After the survey was designed, it was pilot tested to ensure that the survey was readable, culturally acceptable, valid, and reliable. Data were collected via a four-wave postal mailing method that used best practices in survey research to ensure the optimum return rate.
Results: A total of 386 surveys were completed and returned (25% response rate). The respondents were non-Hispanic White (65%), married (61%), male (52%), with a post-secondary degree (59%), and with a total household income of $70,000 and above (32%). The majority of participants (75%) had not completed end-of-life care planning. Approximately 1 in 3 respondents (36%) had completed a living will, 32% had completed a durable power of attorney for health care (DPAHC), 60% had discussed end-of-life care wishes with at least one family member, and 18% had discussed end-of-life care wishes with their physician.
Statistically significant differences in the completion rates of the three key planning behaviors were found by race/ethnicity (X2(2) = 18.90, p < .001). One in three Whites (33%) had completed end-of-life care planning versus 18% of Hispanics and 8% of Blacks. Specifically, more Whites (43%) than Hispanics (26%) or Blacks (16%) were likely to have completed a living will. Similarly, more Whites (39%) than Hispanics (25%) or Blacks (14%) were likely to have completed a Durable Power of Attorney for Health Care (DPAHC). Likewise, more Whites (68%) than Hispanics (59%) or Blacks (38%) were likely to have had discussions with family members. Finally, more Whites (21%) than Hispanics (15%) or Blacks (9%) were likely to have had discussions with a health care provider.
The IBM constructs of direct attitudes and indirect attitudes were statistically significant predictors of behavioral intention. Attitude scores differed significantly by race/ethnicity. Whites were more likely to have positive direct attitudes toward advance directives (M = 23.85, SD = 5.37) than Blacks (M = 22.36, SD = 5.20) and Hispanics (M = 21.84, SD = 5.50) (F (2, 338) = 4.15, p < .05). As direct attitudes toward advance directives increased/improved, respondents’ behavioral intentions to complete end-of-life care planning also increased [r = .446, n = 197, p < .001]. Likewise, direct attitudes were significantly and positively associated with discussing wishes with family members and intention to complete end-of-life care planning [r = .323, n = 198, p < .001]. Indirect attitudes were also significantly and positively associated with end-of-life care planning and intention to complete end-of-life care planning [r = .457, n = 200, p < .001].
The IBM construct of perceived norms was also influential. Respondents reported that they would be more likely to complete end-of-life planning if their physician, loved ones, and best friend wanted them to complete it. Statistically significant differences were noted by race/ethnicity for direct perceived norms (F (2, 364) = 4.18, p < .05) and indirect perceived norms (F (2, 359) = 15.39, p < .001). Whites were more likely to have higher levels of direct perceived norms (M = 7.57, SD = 1.94) than Blacks (M = 7.00, SD = 2.79) and Hispanics (M = 6.76, SD = 2.38). In addition, Whites were more likely to have higher levels of indirect perceived norms for end-of-life care planning (M = 16.99, SD = 3.39) than Blacks (M = 14.90, SD = 4.20) and Hispanics (M = 14.48, SD = 4.39).
The IBM construct of personal agency may also help to explain the differences noted by race. Respondents’ level of self-efficacy impacted their PAPM stage of readiness to take action to complete end-of-life care planning. Those with the lowest level of self-efficacy were more likely to decide they did not want to complete end-of-life care planning, while those with the highest self-efficacy level were more likely to decide they did want to complete end-of-life care planning (F (4,254) = 15.32, p < .001). Self-efficacy was also positively and significantly associated with behavioral intentions [r = .312, n = 202, p < .001]. Whites had higher self-efficacy (M = 24.17, SD = 5.12) to complete end-of-life care planning than Blacks (M = 23.48, SD = 5.46) and Hispanics (M = 21.96, SD = 5.81) (F (2, 262) = 3.17, p < .05).
Other factors outside of the IBM (e.g., racial disparities in education and income) likely played a role in the differences in end-of-life care planning and behavioral intention by race. Respondents with higher education and household income levels were more likely to have completed end-of-life care planning. Racial/ethnic minorities had lower education levels and lower household incomes than Whites. Sixty-six percent of Whites in the current study had a post-secondary degree compared to 51% of Hispanics and 38% of Blacks. Similarly, nearly 50% of Blacks, 34% of Hispanics, and only 10% of Whites had total household incomes less than $30,000 per year.
Conclusion: Health care providers, patient educators, hospice organizations, and health educators should use the Integrated Behavioral Model and the Precaution Adoption Process Model as a framework for the design, implementation, and evaluation of custom tailored messages, social marketing campaigns, and other educational initiatives to increase awareness, knowledge, and completion of end-of-life care planning among American adults, particularly among racial/ethnic minorities.