Longitudinal Randomized Controlled Trial of Advance Care Planning for Teens: Religiousness and End of Life Treatment Preferences

Abstract

Purpose: One pilot study among HIV positive adolescents indicated spirituality, as measured by the FACIT-SP, was not associated with the decision to use intensive life-prolonging treatments. Yet adult studies indicate the opposite (Phelps, 2009), particularly for African-Americans (Johnston, et al., 2005). The purpose of this study is to examine the relationship between religiousness in adolescents living with HIV and their end of life treatment preferences in a longitudinal randomized clinical trial of an advance care planning intervention. Hypothesis: high religiousness at baseline will interact with an advance care planning intervention to increase the likelihood of choosing to continue all treatments across end of life situations at 3 month post-intervention follow-up. Methods: Adolescents ( 14, <21 years of age, N1⁄4107) were randomized to either the FAmily-CEntered Advance Care Planning intervention or to the Healthy Living Control condition. Religiousness was assessed at baseline through the Brief Multidimensional Measurement of Religiousness/Spirituality (BMMRS; Fetzer Institute/NIA 1999). The Statement of Treatment Preferences measured treatment preferences in three HIV related bad outcome situations (SOTP; Hammes & Briggs, 2008) at 3 month post-intervention. Analysis included 85 adolescents who completed 3 month follow-up, using latent profile analysis (LPA) of models of religiousness. The logistic procedure for analysis of maximum likelihood estimates was then used to test for direct intervention effects and intervention interactionwith latent profiles of religiousness on preference outcomes. Fisher’s exact test for binary relationship and logistic model were then used. Results: Adolescents were: 51% female, 72% perinatally-acquired HIV; mean age 18 years old; 93% African-American or black; 27% advanced HIV. LPA yielded four profiles: Profile 1 and 3 were religious while Profile 2 and 4 were not. There was one significant relationship of continue treatment in Situation 1 and religious profile. Patients in Religious Profile 1 and Profile 3 (higher religiousness) were more likely to choose to continue all treatments in Situation 1 (prolonged hospital stay with multiple interventions) than patients in Religious Profile 2 and Profile 4 (low religiousness) (b1⁄43.00-0.731⁄42.27, p-value1⁄40.0137). The main effect of the advance care planning intervention was to significantly decrease the choice to continue all treatments in Situation 1 (Maximum likelihood estimate -2.3286, p1⁄40.0276). In the control group there was no interaction of religiousness with treatment preferences. There were no significant effects for age, gender, race or mode of transmission. Conclusions: The main effect of the intervention, to decrease the choice to continue all treatments in Situation 1, disappeared if the adolescent had strong religiousness. This is similar to adult studies that show religiousness as a mediator of end of life decisions. Religious HIV positive adolescents who were randomized to the intervention were willing to endure long hospitalizations compared to their nonreligious peers. Sources of Support: