Toye 2016.

Methods	Parallel group, single‐blind, randomised controlled trial (recruitment April 2015‐November 2015)
Participants	Caregivers of older people discharged from the medical assessment unit within a metropolitan tertiary hospital in Western Australia with over 600 beds. Caregivers were recruited in the hospital at the time of patient discharge. A family caregiver was defined as a family member or friend who provides unpaid personal care, support, and assistance. Inclusion criteria for dyads were that they comprised a patient aged 70 years or older being discharged to their home or the home of their family caregiver, plus a family caregiver who could speak and read English. Caregiver mean age in years for the intervention group was 63.1 (12.6 SD) and the control group 61.3 (13.4 SD). Females comprised 74% (n = 104) and males 26% (n = 37). Relationship to care‐recipient was husband (n = 13), wife (n = 29), son (n = 62), daughter (n = 14) and other (n = 104).
Interventions	Intervention: Further Enabling Care at Home (FECH) program and usual discharge care (n = 86). Aim: The FECH intervention is intended to identify family caregivers of older patients during the hospital admission, help ensure their understanding of discharge information that has implications for the caregiving role, prompt reflection by the caregiver on this role and what is needed to help sustain this, and guide the caregiver as they identify and address prioritised needs for support (information provided by author via email). Interventionist(s): A nurse with acute care knowledge relevant to the care of older people in poor health, knowledge of how to access local services, understanding of the family caregiver role, the capacity to work flexible hours to fit in with caregivers’ needs, and the skills to support the caregiver during the process of reflection and self‐assessment. Mode of delivery: Telephone Duration: Up to 40 days (planned calls were delayed). Planned calls were weekly to biweekly (call 1: within a week post‐discharge, call 2: 7‐10 days post‐discharge, call 3: 14 days after discharge). Actual call delivery: call 1 within and up to 9 (instead of 7), contact 2 within 24 (instead of 10) and contact 3 within 40 (instead of 14), days post‐discharge. Mean and standard deviation call contact time in minutes was contact 1: 15.4 (9.6), contact 2: 59.7 (24.1) and contact 3: 28.3 (17.7). Content: The Further Enabling Care at Home program involved the implementation of a strict telephone protocol by the specially trained nurse, using the Carer Support Needs Assessment Tool, which has fourteen items covering: (a) support that enables the caregiver to care for the patient at home, and (b) support for the caregiver in their caring role. There were three, sequential, telephone contacts. Contact 1 was planned to take place within a week post‐discharge. Contact 2 was designated to occur from 7 to 10 days post‐discharge. Contact 3 was planned to follow within 14 days of the discharge. Standardisation: The delivery of the intervention was monitored at regular meetings with investigators responsible for this issue. Field notes were taken by the FECH nurse during intervention contacts to provide a record that allowed the discussion of cases during these meetings so that fidelity could be assured. Using this process, intervention delivery was as planned and consistent, except with respect to the planned time of the contacts, which were delayed in some instances because of the busy schedules of the caregivers. The selection and preparation of the FECH nurse, plus the use of a pre‐prepared resource manual, also helped to ensure the standardised quality of the intervention (information provided by author via email). Comparison group: Usual discharge care (n = 89).
Outcomes	1. Burden: Caregiver strain subscale of the Family Appraisal of Caregiving Questionnaire – Palliative Care; higher scores indicated greater burden. 2. Skill acquisition (preparedness to care): Preparedness for Caregiving Scale from the Family Care Inventory; higher scores indicated better perceived preparedness to care. 2. Family functioning: Family Well‐Being subscale of the Family Appraisal of Caregiving Questionnaire – Palliative Care; higher scores indicated better family functioning. 3. Health status and well‐being (physical health): SF‐12 v2 Health Survey used for assessing physical health and not as a QoL outcome (caregiver ratings of their own health and well‐being). Higher scores indicated better physical health. Cost: No specific instrument (intervention costs recorded include (i) nurse time for the duration of each contact; (ii) nurse time to implement and organise resources; (iii) nurse time to write notes following each contact for each patient‐carer dyad; (iv) cost of training the FECH nurse; (v) telephone charges; and (vi) stationary and postage costs. Costs in the control group were estimates of nurse time for usual discharge procedures). Higher scores indicated greater cost. Outcome data were collected at Time 1 (within 4 days of discharge), Time 2 (15–21 days after discharge) and Time 3 (end of intervention time point, six weeks after discharge).
Notes	Unpublished mean and standard deviations along with details of the cost data collected were obtained from authors via email. Author confirmed via email that care‐recipients may have included those with an exacerbation of a chronic condition or an additional acute illness or both. Funding source: A Department of Health Western Australia, SHRAC Research Translation Project grant (p.40).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated list of random allocations prepared prior to the study commencing, using a permuted random blocks strategy (p.35)
Allocation concealment (selection bias)	Low risk	Allocation schedule held by researcher not involved in recruitment (p.35)
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Unable to blind due to nature of the intervention
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All researchers involved in quantitative data collection were blinded to the allocation schedule and actual group assignment (p.35).
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition Intervention group 19.5%; control group 8.1% (p.37). “From the 12 dyads withdrawing after randomisation without providing data, most failed to provide a reason but three caregivers had concerns about, or difficulties with, the planned telephone data collection and one experienced a bereavement (p.37)”.
Selective reporting (reporting bias)	High risk	Physical and mental health outcomes not reported
Other bias	Low risk	No significant difference in caregiver characteristics between the groups (p.37) No significant differences in baseline measures for outcomes (Table 2, p.38; table 3 page 39)