Hypotheses 1a and 1b: Compared to Supportive Conversation arm, the EMPOWER intervention will significantly decrease surrogate decision makers' symptoms of grief and Post Traumatic Stress Disorder (PTSD) (primary outcomes); and H1b. experiential avoidance, depression, regrets, and increases in patients' value-concordant care (secondary outcomes) at T1-T4.

Hypothesis 2. Qualitative data will provide insights not captured by quantitative data.

Hypothesis 3. Reductions in experiential avoidance will mediate reductions in grief and Post Traumatic Stress Disorder (PTSD) symptoms, highlighting it as important to target in future implementation.

Intensive Care Units (ICUs) are stressful places fraught with grief for family members who witness dying loved ones, often in pain, struggling to breathe and/or maintain consciousness. Compounding their distress, family members are often thrust into the position of patient "surrogate," needing to make life-and-death decisions on the patient's behalf. Researchers have shown that end-of-life (EoL) decision-making is undermined by grief, which interferes with acceptance of the patient's impending death and leads to care choices that adversely affect patients' quality of care and death.1-3 These circumstances heighten surrogates' risk of meeting criteria for Prolonged Grief Disorder (PGD), Posttraumatic Stress Disorder (PTSD), and decisional regret about the EoL care that the patient received, each associated with poor bereavement outcomes.4-7 Nearly 60% of ICU surrogates report moderate to extreme grief; 34% report extreme levels of peritraumatic stress symptoms.1

The coronavirus (COVID-19) pandemic has made an already bad situation worse. At the start of the pandemic, social distancing policies forced millions of families to confront obstacles to communication, medical decision-making, and care.8-10 Surrogates were left struggling with severe pre-loss grief and peritraumatic stress -- intensely longing to be near to the patient, confused about their roles, lonely, horrified, angry, disoriented and emotionally numb.10,11 Now, as the Delta variant creates a new "wave" of mortality and infection, bereaved family members may have remorse about vaccine refusal,12 feel guilty for transmitting the virus to the patient, or regret decisions about EoL care. With over 35 million cases and 600,000 deaths in the United States from COVID-19,13 the need for psychosocial interventions to support surrogates in the ICU is clear.

Prior efforts to address the plight of family surrogates of critically ill patients have proved disappointing14-20 - with one ICU intervention significantly increasing the surrogate's severity of PTSD symptoms.14 A key limitation of these interventions is that while they targeted psychological outcomes, they were not psychological interventions. To address this, the investigators developed a brief, flexibly administered cognitive-behavioral, acceptance-based psychological intervention called EMPOWER (Enhancing & Mobilizing the POtential for Wellness & Emotional Resilience).21,22 Our pilot NIH-R21 (N=39) showed that EMPOWER had superior efficacy to enhanced usual care for reducing symptoms of PGD (d=1.20) and PTSD (d=.99). Consistent with mediation, EMPOWER reduced experiential avoidance (d=1.20); these reductions were correlated with PGD and PTSD change scores (p<0.01). Large reductions in decisional regret (d=1.57) were observed, with no notable differences by surrogate race or delivery format (telehealth vs. in-person).

Investigators propose to conduct a Phase II mixed methods randomized controlled trial (RCT) to further evaluate the efficacy of EMPOWER for reducing surrogate symptoms of PTSD and PGD. Surrogates (N=172) will be randomized to EMPOWER (n=86) or a standardized supportive conversation (SC; n=86). Effects of the intervention will be assessed via measures administered pre-intervention (T1), immediately post-intervention (T2), and at 3 months (T3), and 12 months (T4) following the T2 assessment. Investigators will also conduct semi-structured interviews with surrogates (n≈48) to probe intervention effects on mental health and explore contextual factors (e.g., medical mistrust, visitation restrictions) likely to affect surrogates during the pandemic.

Inclusion Criteria:

• Patients who, during their hospital stay, have been admitted to the ICU and are near EoL, as indicated by a modified "surprise question" whereby physicians are asked to identify patients whom they do not expect to survive the next 12 months.
• Surrogate decision-makers of patients who were admitted to the ICU or step-down unit during their current admission/stay, or within 1 month of discharge from their last admission/stay.
• Surrogate decision-makers are 18 years or older.
• Surrogate decision-makers whom physicians or advance practice providers (i.e. physician assistants, nurse practitioners) indicate as the designated health care proxy or decision-making patient surrogates, or who are listed as such in the patient's medical charts or by self-report of the surrogate.
• Surrogate decision-makers must speak English.
• Surrogate decision-makers must report "syndromal" levels of pre-loss grief (PG-12 score ≥ 25) or peritraumatic distress (PDI ≥ 23).
• Surrogate decision makers will need to reside in a state in which an interventionist is licensed or otherwise be able to comply with current telehealth regulations.
• Surrogate decision-makers will need to be willing to utilize a device (computer, tablet, phone) with internet.
• Surrogate decision-makers who are able and willing to provide an emergency contact.

Exclusion Criteria:

• Patients and surrogate decision-makers who do not meet the eligibility criteria.
• Surrogate decision-makers who indicate the presence of cognitive impairment based on responses to the Ultra-Brief Confusion Assessment Method and/or significant psychiatric or cognitive disturbance sufficient, in the investigator/study staff's judgment, to preclude completion of the assessment measures, interview or informed consent.
• Surrogate-decision makers who endorse suicidal ideation in the past month based on responses to the Columbia Suicide Severity Rating Scale.
• Surrogate-decision makers who are unable to access a functional device for videoconferencing and decline the offer to use a study loner device.

All participant data, including self-report measures and qualitative interviews, will be shared to qualified investigators under a Data Use Agreement in de-identifiable form. Data will be made available upon approval of PIs and will be used solely for research purposes.

Data will become available once the study's specific aims will be published and will be destroyed once analyses are completed.

Accessing data requires making a formal request at the appropriate time and the PIs and investigative team will review the request. If approved, the data analyst from our Cornell Center for Research on End-of-Life Care will make a de-identifiable dataset to address the analyses in the request.