Martindale‐Adams 2013.

Study characteristics
Methods	Design: Parallel randomised study. No. of participating centres: One, Veterans Affairs Medical Center, Memphis, Tennessee, USA. Study dates: October 2004 to September 2007.
Participants	Inclusion criteria: Caregivers reporting stress or difficulty with care, living with care recipient, and providing 4 or more hours/day of supervision or care for at least 6 months. Care recipients had a dementia diagnosis or MMSE score of 23 or less and at least 1 ADL or 2 IL limitations. One member of the dyad had to be a veteran receiving services at VAMC Memphis. Exclusion criteria: Planned nursing home admission within 6 months. No. of participants randomised to interventions: ‐ Telephone Support for Dementia Caregivers (n = 77). ‐ information alone (n = 77). Baseline characteristics: Caregivers: Mean age for participants was 65.6 years (SD = 12.4). Most caregivers were female (83.7%). The mean length of time caregiving was 48.9 months (SD = 38.5). The mean caregiving hours on duty was 16.4 (SD = 8.6), and the mean caregiving time was 6.5 hours/day (SD = 4.9). The mean score for burden was 17.05 (SD = 8.7). Care recipients: Mean age for patents was 77.4 years (SD = 7.4), and most were male (87.6%). The mean MMSE score was 15.4 (SD = 7.5).
Interventions	Experimental: Telephone Support for Dementia Caregivers included 12 individual in‐home and telephone sessions and 5 telephone support group sessions. The support groups met bi‐weekly for 2 months and monthly thereafter for 1 year, for a total of 14 hour‐long sessions. The sessions were semi‐structured telephone calls with education, skills‐building, and support. Control: Control caregivers received pamphlets on dementia and safety and information for local resources. Length of intervention: 12 months.
Outcomes	Primary outcomes: ‐ General health, assessed with one question from the SF‐36. ‐ Caregiver burden, assessed with the ZBI scale (12‐item), (score range 0 to 48, higher scores denote greater burden). ‐ Depression, assessed by the CES‐D scale, (score range 0 to 30, higher scores denote more depressive symptoms). ‐ Overall mental health, assessed by the General Well‐Being Scale 22‐item (scoring is 1 = definitely true, all of the time to 5 = definitely false, not at all; higher scores indicate greater well‐being). ‐ Bother, assessed with the RMBPC scale (score range 0 to 96, higher scores represent higher perceived burden). ‐ Care recipient health, assessed with one question from the SF‐36 (1 = poor to 5 = excellent). Secondary outcomes: None stated. Outcomes assessment: At 6 (intermediate assessment), and 12 months (final assessment).
Notes	Clinical trial registration code: NCT00119561. Funding: Veterans Health Administration, Health Services Research and Development Service (IIR‐03‐287‐1), US Department of Veteran Affairs; and Memphis Veterans Affairs Medical Center. Conflicts of interest: None declared. Fidelity / adherence: 61% of participants took part in at least 75% of the support group sessions of the program.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: Study described as randomised but there is no information on the randomization procedure.
Allocation concealment (selection bias)	Unclear risk	Comment: There is no information on allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: Interventions cannot be blinded. Study described as open trial in the protocol.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: Intervention components and procedures were not comparable in an unmasked study with participants not blind to interventions and subjective participant‐reported outcomes.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: There is a low attrition rate and similar between groups: 8 participants of 77 (10.4%) were lost to follow up in the intervention group, 7 participants of 77 (9.1%) were lost to follow up in the control group.
Selective reporting (reporting bias)	Low risk	Comment: Clinical Trials Identifier: NCT00119561.
Other bias	Low risk	Comment: None detected.