Patients With Advanced Cancer Choose Less Aggressive Medical Treatment on Vignettes After Using a Computer-Based Decision Aid

• Published in: American journal of hospice & palliative medicine 2020-07, Vol.37 (7), p.537-541
• Main Authors: Green, Michael J., Van Scoy, Lauren J., Foy, Andrew J., Dimmock, Anne E.F., Lehman, Erik, Levi, Benjamin H.
• Format: Article
• Language: English
• Subjects: Advance Care Planning - organization & administration; Advance Directives - psychology; Aged; Aged, 80 and over; Cardiopulmonary Resuscitation - psychology; Cross-Over Studies; Humans; Middle Aged; Patient Acceptance of Health Care - psychology; Severity of Illness Index; Single-Blind Method; Withholding Treatment - statistics & numerical data
• ISSN: 1049-9091; 1938-2715
• DOI: 10.1177/1049909119892596

Background: Although patients often prefer less rather than more treatment at the end of life, in the absence of contrary instructions, the medical profession’s de facto position is to treat aggressively. It is unknown whether a computer-based decision aid can affect treatment choices. Methods: Secondary analysis of a single-center, single-blind randomized controlled trial of an advance care planning (ACP) intervention among 200 patients with stage IV cancer. Participants were randomized to intervention (Making Your Wishes Known, a values-neutral, educational, computer-based decision aid) or control (standard living will + brochure). After reading a hypothetical clinical vignette, participants were asked whether they would want 11 medical/surgical treatments in that situation (dialysis, cardiopulmonary resuscitation [CPR], ventilator, feeding tube, etc). The median number of treatments wanted by participants was compared between groups, and logistic regression was used to compare between-group likelihood of not wanting each specific treatment. Results: The median number of treatments wanted was 1 in the intervention group versus 5 in the control (P < .001). For 6 of 11 treatments, the intervention group was significantly less likely than control to want aggressive treatment. Most notably, compared to control, intervention participants were less likely to want CPR (odds ratio [OR] = 0.31), short-term mechanical ventilation (OR = 0.34), short-term dialysis (OR = 0.38), surgery (OR = 0.37), and transfusion (OR = 0.21). Conclusions: Individuals using an educational ACP decision aid were less likely to want aggressive medical treatment than those completing standard living wills. These findings have implications not only for how to respect patient’s wishes but also potentially for reducing costs at the end of life.