Reinhardt 2014.

Study characteristics
Methods	Aim: to assess the effects of a face‐to‐face, structured conversation about EoL intervention, compared with social contact via telephone only Study design: RCT, 2 arms: intervention, usual care Consumer involvement: none reported Funding source: Alzheimer’s Association provided funding support. No declarations of interest are provided from authors
Participants	Participants: family members of residents with advanced dementia of a large skilled nursing facility Setting and geographic location: residential nursing facility; New York, USA Methods of recruitment: few details reported; surrogates were chosen from primary contacts of patients Surrogates provided informed consent (for themselves), and surrogate informed consent for patient participation Selection criteria for participation in study: Inclusion: patients: advanced dementia (cognitive performance scale score = 4,5,6), English or Spanish speaking, not currently receiving hospice care (where EoL care would have been discussed) surrogates were primary contact for the patient, selection criteria not reported Exclusion: no additional exclusions reported Diagnosis of person approaching EoL: advanced dementia Target of intervention: surrogates: primary family member or friend contact, including healthcare agent Age: intervention 59.6 years (SD 12.3), UC 58.9 (11.9) Gender: intervention 37/47 (78.7%) female, UC 32/40 (80%) female Ethnicity/culture/language: 107/110 English speaking (3 Spanish) Black non‐Hispanic (intervention 42.5%, control 40%); White non‐Hispanic (intervention 31.9%, control 30%), Hispanic (intervention 23.4%, control 23%), other (intervention 2.1%, control 7%) Other PROGRESS aspects: study occurred in urban centre, in relatively highly educated group of surrogates (over 50% educated to college (university) level). Findings may not be applicable to rural or more remote populations, lower‐income countries and settings, or to those with lower levels of education and/or health literacy Authors also note that the intervention could only be conducted in this facility because of the employment of full‐time physicians by the care home, including palliative medicine physicians, which is not typical of most nursing homes Numbers of participants: see Additional Table 1
Interventions	Intervention: face‐to‐face structured conversation about EoL care, with telephone follow‐up Aim of intervention: to provide information and support to surrogates of patients with advanced dementia, including about the pros and cons of treatment decisions that may arise when the patient’s dementia severity worsens Comparison: social contact by telephone (to control for additional attention and interaction between staff and surrogates in intervention group) plus routine care. Delivered by trained research assistant Delivered by: intervention delivered by 1 of the PCT physicians and the social worker The elements listed in the structured meeting description were reviewed in a training session with the clinicians delivering the intervention Setting: at the care facility, no further details reported. Follow‐up calls via telephone Materials, procedures, content/When and how much: Structured meeting was one‐off; mean duration 47 minutes (range 20 to 75 minutes) Meeting was one‐off; PCT was available for further information or assistance with decision making but only 3 surrogates requested additional information Social worker contacted surrogates every 2 months via telephone to provide support and assess surrogate’s level of emotional comfort. This was an opportunity for surrogates to have concerns addressed, and designed to continue discussions about any issues raised in the initial meetings Each call lasted mean 10 minutes UC: baseline and 2‐month intervals telephone calls. Discussed whatever the surrogate raised on the call Mean 11 minutes at baseline, 9 minutes for follow‐up calls Tailoring: meetings and follow‐up phone calls aimed to cover those issues that surrogates wished to discuss Modified during study: not applicable Co‐intervention(s): not reported Fidelity assessed: quality of intervention not assessed explicitly. No further measures of fidelity reported Theoretical base: no theoretical basis cited but prior research mentioned
Outcomes	Primary outcomes None reported Primary outcomes ‐ adverse events None reported Secondary outcomes Surrogate ratings of patient's symptom management (EOLD SM) [Quality of EoL care] Method: interview, questionnaire Timing: baseline, 3 and 6 months Scale and scoring: Symptom Management at the End of Life in Dementia Scale. Frequency of 9 symptoms rated on 6 point scale (0 = never, 6 = daily); range 0 to 45. Higher score = better symptom control Surrogate care satisfaction (EOLD SWC) [Quality of EoL care] Method: interview, questionnaire Timing: baseline, 3 and 6 months Scale and scoring: Satisfaction with Care at EoL in Dementia Scale. Frequency of 14 items, rated on 4‐point scale (strongly agree to strongly disagree; possible range 0 to 42). Higher score = greater satisfaction Surrogates' satisfaction with care [Quality of EoL care] Method: interview, questionnaire Timing: baseline, 3 and 6 months Scale and scoring: single item, 0 to 10 rating (0 worst possible care to 10 best possible care)
Notes	Trial also reported surrogate depressive symptoms, satisfaction with life, and patient medical data (medical record review) at 3 and 6 months, which are not reported in this review as they were judged to be primarily clinical outcomes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Participants randomly assigned but no further details provided
Allocation concealment (selection bias)	Unclear risk	No details reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	No measures described to blind providers of intervention (and unlikely to be possible). Effect on outcomes unclear No measures described to blind surrogates to intervention, although comparison group received some telephone contact. May introduce bias if reporting on satisfaction with care etc. if participants knew they were part of the intervention group
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Interviewers assessing outcomes via surrogate interviews were blinded to study group allocation. Medical records were sourced for patient outcomes; not clear whether assessors were blinded to group allocation of patient but data are objective so risk of bias seems low
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Missing data were reported 110 were randomly assigned and completed baseline interviews; 96 (87%) completed 3‐month outcomes; 90 (82%) completed 6‐month outcomes. Losses were fairly comparable across groups and no major differences between those who completed the study and those who dropped out (on key demographic features) were noted by the authors However, numbers were lower for some outcomes (such as ratings of care management) where n = 65 (numbers fairly comparable between the 2 groups). For some outcomes, such as care satisfaction (intervention n = 45 and control n = 36) missing data may affect the results and there were differences between the groups Not clear what the impact might be on the result, or reasons for these missing data
Selective reporting (reporting bias)	Unclear risk	No protocol identified. Outcomes seem complete based on those sought in methods
Other bias	Low risk	Baseline imbalances: groups were comparable on key features at baseline