Dionne‐Odom 2015.

Methods	Randomised controlled trial
Participants	Inclusion criteria: family caregiver ('a person who knows you well and is involved in your medical care') of a patient who: was aged > 18 years; had new diagnosis, recurrence, or progression of an advanced‐stage cancer within about 30–60 days of the date the patient was informed of the diagnosis by his or her oncology clinician; and had an oncologist‐determined prognosis of 6–24 months; was English speaking; was able to complete baseline questionnaires. Exclusion criteria: scored < 4 on the Callahan Cognitive Screen; had an untreated axis 1 psychiatric condition or an active substance‐use disorder; or had uncorrectable hearing disorder or unreliable telephone service. No additional exclusion criteria for caregivers. Number randomised: 124 dyads (63 early intervention; 61 delayed intervention) Follow‐up: by telephone, once every 6 weeks until 24 weeks; then every 3 months until end of study or patient death. Not all were followed up after 24 weeks. Setting: participants were recruited from a cancer centre (and affiliated outreach clinics) and a medical centre in the US.
Interventions	Intervention group: 3 structured 1‐to‐1 telephone sessions (with guidebook; once a week) between an advanced‐practice palliative care nurse coach and a caregiver. Session 1 addressed taking on the caregiver role, defined palliative and supportive care, and introduced problem‐solving using the framework of the COPE attitude. Session 2 covered caregiver self‐care and effective partnering in patient symptoms assessment and management. Session 3 addressed the building of a support team, decision making, decision support, and advance care planning. Sessions lasted on average 23 minutes, and the same nurse coaches followed up with participants monthly until end of study or patient death. Control group: no treatment but could take part in the intervention after 3 months.
Outcomes	Caregiver outcomes QoL (Caregiver QoL Scale – Cancer) Depression (Center for Epidemiologic Study – Depression Scale) Caregiver burden (Montgomery‐Borgatta Caregiver Burden Scale) Outcomes reported elsewhere (Dionne‐Odom 2016): Complicated grief (Prigerson Inventory of Complicated Grief – Short Form) Outcomes not reported on (confirmed via correspondence): Personality traits (NEO Personality Inventory‐3) Patient outcomes (Bakitas 2015): QoL (Functional Assessment of Chronic Illness Therapy – Palliative Care; Treatment Outcome Index) Symptom impact (QoL at End of Life Symptom Impact sub scale) Mood (Center for Epidemiologic Studies – Depression Scale) 1‐year and overall survival Resource use and location of death (hospital and intensive care unit days, emergency department visits, period between last assessment and death, chemotherapy use in last 14 days, location of death)
Notes	Published and unpublished data. 22% neuro‐oncology (124 participants, 63 to the intervention group and 61 to the control group. In total, 27 caregivers of patients with brain tumour were included (3 primary brain tumour; 24 secondary brain tumour; confirmed via correspondence). ENABLE III (Educate, Nurture, Advise, Before Life Ends III) trial
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Not described in the report. Patients were randomised 1:1 using a computer‐generated scheme. Caregivers were assigned to the same group. Confirmed via correspondence.
Allocation concealment (selection bias)	Low risk	Not described in the report. Computer‐generated randomisation after enrolment, confirmed by authors.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding of participants was not possible due to the nature of the intervention.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Analyses were not performed blind (confirmed via correspondence).
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Dropout: about 32% of caregivers did not complete all follow‐up assessments; analyses revealed no significant associations between attrition and measured caregiver characteristics or outcome. Maximum likelihood methods were used to estimate missing outcome data.
Selective reporting (reporting bias)	Unclear risk	There was no published study protocol. All outcomes specified in the manuscript were reported on; 1 additional outcome (complicated grief) was reported in 2016 publication; personality was assessed but not reported on. Authors confirmed no other outcomes were collected.