Patient Initiation of End-of-Life Discussions, a Randomized Control Trial

Abstract

Rationale -For the aging and comorbid veteran population, COVID-19 has made “ventilator” a household word. With the excessive cost and low quality of life associated with aggressive end of life (EOL) care, the need for effective goals of care (GOC) conversations prior to development of acute illness has never been higher. Physicians are often reluctant to initiate these conversations, but patients could be prompted to broach the topic using standardized media delivered in the waiting room. Methods -We conducted a randomized controlled trial evaluating educational media in the outpatient setting. Veterans in the waiting room who were over 65 were randomized to one of two interventions or control. The interventions were a VHA produced brochure on GOC or a 7-minute video on GOC featuring a mock code. Participants were given a survey, and had a follow up phone interview to assess if they had brought up EOL care at their office visit. At 30 days, chart review assessed documentation of GOC. Primary endpoint was whether the patient initiated an EOL discussion at their office visit. Secondary endpoints included code status, GOC documentation, and evaluation of emotional response. Results -Despite hundreds of eligible patients, <10% opted to discuss enrollment, and <5% enrolled in this study. Needed sample size was 153, with only 30 enrolled at study conclusion. There was low rates of all endpoints. Only one participant initiated EOL discussions, but this discussion was not documented. Three filed new GOC documents (including one who died while CMO). Two found the material upsetting (including one in the control). None indicated that they would not trust their physician to make EOL decisions for them, though several were unsure. All participants thought that their material should be shown to other veterans. Due to underpowering, there was no statistical difference in any outcome (Table 1). Conclusion -EOL discussions remain an important job of the outpatient physician, though many patients do not discuss EOL care until they are acutely ill. A standardized patient centered format delivered in the clinic waiting room remains a promising option to facilitate these discussions, though there are still physician level barriers in documenting these conversations. Larger studies are required to demonstrate that this type of intervention is effective. Our study shows that patients have low rates of negative emotional responses to this type of material, and would universally recommend this material to other veterans.