Sulmasy, D.P., Hughes, M.T., Yenokyan, G., Kub, J., Terry, P.B., Astrow, A.B., . . . Nolan, M.T. (2017). The trial of ascertaining individual preferences for loved ones' role in end-of-life decisions (TAILORED) study: A randomized controlled trial to improve surrogate decision making. Journal of Pain and Symptom Management, 54, 455–465.

To test the effectiveness of a nurse-facilitated discussion intervention between patients and their caregivers (surrogates) about the role patients would prefer their surrogates play when they are unable to make decisions

A nurse-led intervention administered by a nurse to dyads of patients with gastrointestinal (GI) cancer and their caregivers (CGs). The nurse compared a patient and a CG responses on a decisional control preferences (DCP) tool and facilitated a discussion between the two about the role a patient would like the CG to play when a patient becomes unable to make decisions. Patients and their surrogates received a Patient-Loved One Decision Making Handout that reflected a patient’s DCP; they also received the advanced directive information. A follow-up call was made at four weeks in which the nurse encouraged them to discuss the patients’ preferences for involving the surrogate in healthcare decisions. In addition, they were reminded to review the handout. Patients and CGs were surveyed at baseline and at eight (± 2) weeks postintervention. The trial is known as The Trial of Ascertaining Individual Preferences for Loved Ones’ Role in End-of-Life Decisions (TAILORED)

• N = 137 dyads, 46 of which were patients with GI cancer and their surrogates   
• AGE: 61.2 for all the dyads.
• MALES: 57%  
• FEMALES: 43%
• CURRENT TREATMENT: Chemotherapy, other
• KEY DISEASE CHARACTERISTICS: Patients with advanced GI cancers excluding those with isolated hepatic metastasis. The study also included dyads of surrogates and patients who had ALS.
• OTHER KEY SAMPLE CHARACTERISTICS: Spoke English

• SITE: Multi-site. Johns Hopkins GI oncology clinic, GI surgical clinic, and ALS clinic; the GI oncology clinic at the University of Chicago 
• SETTING TYPE: Multiple settings    
• LOCATION: United States

• PHASE OF CARE: End-of-life care
• APPLICATIONS:  Palliative care

Randomized controlled trial

• Decision Control Preference Scale-Modified to assess surrogates’ perception of their loved one’s preference for surrogates’ involvement.
• Horowitz Impact of Events Scale to assess surrogates’ stress
• Family Decision Making Self-Efficacy scale to assess self-efficacy
• Zarit-Short Form to assess caregiver burden
• A single-item question to assess satisfaction

• Improved patient-surrogate concordance on the DCP scale after intervention, but not statistically significant. Significant increase in likelihood of surrogates endorsing shared decision making in the intervention group from baseline to follow-up (OR_adj = 2.05, p = 0.03) as compared to control (OR_adj = 1.44, p = 0.28). Increased endorsement of shared decision making from baseline to follow-up in the treatment group (51% from 38%) as compared to control (48% from 40%).
• Surrogate stress declined significantly from baseline (23.1, SD = 15.8) to follow-up (20.8, SD = 15.3) in the treatment group (p = 0.046). At the eight-week follow-up, the treatment group stress score was 20% lower than that of control (adj ratio of means = 0.8, p = 0.04).
• No significant change in surrogates self-efficacy or significant difference between treatment and control group
• Caregiver burden increased from baseline to follow-up in the treatment group (12.5, SD = 6.5, to 14.7, SD = 8.1; p < 0.01).
• A significantly higher percent of the treatment group (71%) were satisfied than in the control (52%, p = 0.02).

Interventions focusing on improving shared decision-making process (decision control preferences) between patients and their surrogates have favorable effects on surrogate stress and satisfaction. Patients prefer shared decision making over pure substitutive judgment. Although perceived burden increased among the treatment group, they were less stressed and more satisfied. This may suggest that increased burden may not always indicate general deterioration.

• Risk of bias (no blinding)
• Key sample group differences that could influence results
• Questionable protocol fidelity
• Subject withdrawals ≥ 10%
• The intervention was not described well, how long and whether administered by nurses who were trained in a standardized way or to follow a protocol during the discussion. Patients with ALS and GI cancer were combined so it is not possible to know results just for patients with cancer.

Facilitating discussions between patients and caregivers/surrogates about decision-making preferences promotes improvement in certain patients’ and surrogates’ outcomes. These discussions supported with handout are feasible interventions that can be administered by nurses.