A randomized trial of a cardiopulmonary resuscitation (CPR) video (V) in advance care planning (ACP) for progressive hepatopancreaticobiliary cancer patients.

Abstract

9105 Background: CPR is an important advance directive topic in patients with progressive cancer; however such discussions are challenging. Previous work by Volandes et al suggests that V images help, however whether V educational information about CPR engenders broader ACP discourse is unknown. Methods: A randomized trial of an educational V or narrative (N) in patients with progressive hepatopancreaticobiliary cancers was conducted. The V and N were 3 minutes long and contained an identical description of CPR. Primary endpoint: ACP documentation 1-month post-test (either a formal advance directive or a documented ACP discussion), 80% powered with 0.01 type-I error rate for estimated advance directive completion rates of 70% V arm vs. 25% N arm with a sample of 56 subjects. Secondary endpoints (exploratory): impressions of the study information; pre and post knowledge of and preferences for CPR and mechanical ventilation; and follow-up longitudinally (to death or 6 months post-test). Results: 56 subjects consented and analyzed. 1-month post-test ACP documentation was not significantly higher in the V arm (12/30, 40.0%) vs. the N arm (4/26, 15.4%); Fisher’s exact test, OR = 3.583 [95% CI: 0.886 – 17.937], p = 0.07. Knowledge increased in both arms after the intervention. Preferences for CPR changed in V arm but not in N arm; for mechanical ventilation, there was no change in either arm. The majority of subjects in both arms reported the information as helpful, comfortable to discuss, and recommended to others. Longitudinally, outcomes were similar between V & N arms, respectively: % advance directives completed (63% & 58%); median days from test to advance directive (60 & 72) and advance directive to death (21 & 20); median ACP documentation after 1 month (0 & 1); median hospital admissions (1 & 1), days of stay (5 & 7), ICU stays (0 & 3) and CPR/mechanical ventilation (1 & 3). 52% of all subjects died by study end, the majority in hospice care. Conclusions: This first randomized trial addressing downstream ACP effects of V vs. N decision tools demonstrated a trend towards more ACP documentation in V patients. Further research into these types of educational media is needed.