Introduction: Terminally ill patients are often faced with the decision to forgo potentially life-prolonging treatment or to accept hospice care leading to a peaceful death. The decision process in such situations is heavily affected by emotions, chief among them is regret. Modern cognitive science increasingly accepts a dual processing approach to human cognition which takes into account both emotion-based (type 1) and analytical-based (type 2) cognitive processing. Because regret is a human emotion (type 1), which involves counterfactual deliberations (type 2), we have previously proposed that it can activate both cognitive domains by serving as a link between type 1 and type 2 processes and therefore help with end of life decisions more precisely than other decision making methodologies. Here, we report the application of a regret-based model built to facilitate referral to hospice while helping patients clarify their preferences related to how they wish to spend the remaining days of their lives.

Methods: Between March 2013 to December 2015, we conducted a prospective cohort study at the Tampa General Hospital and the Moffitt Cancer Center that enrolled 178 consecutive adult patients aware of the terminal nature of their disease. Eligible patients were those who were at the point in their care where they had to decide between continuing potentially "curative/life-prolonging" treatment (Rx) or accepting hospice care. The study was broken down into 4 steps. First, we computed the patient's probability of survival at 6 months using a validated Palliative Performance Score (PPS-based) predictive model. This probability was communicated to patients as i. percentage, ii. pictorial, and iii. life expectancy in days. Then, we used the Dual Visual Analog Scale technique (DVAS) to elicit patient preferences towards continuing current treatment vs. accepting hospice care. The first scale in DVAS measured the levels of regret of omission (RGO) (e.g. failure to reap hospice benefits and incurring treatment harms) while the second scale measured regret of commission (RGC) (e.g. incurring harms from hospice and failing to provide potential benefits of treatment). The ratio RGO/RGC was used to compute the threshold probability at which a patient is indifferent between accepting hospice care or continuing current treatment. Each patient's threshold was contrasted against the previously estimated survival probability to suggest a patient specific management plan, which was later compared with the patient's actual choice. The final step of the study involved asking each patient a series of qualitative questions to evaluate the usefulness of the regret model in the hospice referral process.

Results: 96% (171/178) of the patients found the information provided by the model helpful; 90% (160/178) stated that it will influence their care decision. 85% (151/178) of the patients agreed with the model's recommendations to either accept hospice care or continue with current treatment [p<0.000001]. The regret model predicted the actual choices for 72% (128/178) of patients [p <0.00001]. Logistic regression analysis showed that people who were initially inclined to be referred to hospice and were predicted to choose hospice over disease-directed treatment by the regret model had close to 98% probability of choosing hospice care at the end of their lives. No other factors (age, gender, race, educational status and pain level) affected the patient actual choice.

Conclusions: To our knowledge, this is the first formal study in which helping patient clarify their preferences enabled them to make actual choices with high level of satisfaction. The regret model was well received by patients and its recommendations were largely accepted. We found that people suffering from a terminal disease who are initially inclined to choose hospice and do not regret such a choice will select hospice care with high level of certainty. We conclude that using the regret model to elicit patient choices is both descriptively and prescriptively valid and can be easily implemented in the actual practice.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution