Xie 2019.

Study characteristics
Methods	Study design: RCT  Duration of study: March 2014 to March 2015
Participants	Country: China   Income classification: upper‐middle income   Geographical scope: Yankou Town is a representative town of Lengshuijiang City of Hunan Province, where many young couples have moved to the cities to work; their parents are typically left behind without care  Healthcare setting: local community health centre  Mental health condition: depression Population: ‘left behind’ elders with GDS score 11 to 25 in Hunan Province  1. Age: ≥ 65 years  2. Gender: both  3. Socioeconomic background: economic support < 500 RMB/month   3. Inclusion criteria    a. GDS score 11 to 25 AND  b. Over 65 years of age   c. Only 1 participant from each family   d. Left behind for longer than 6 months   5. Exclusion criteria   a. Psychiatric and medical comorbidities that are potentially life threatening or expected to severely limit client participation or adherence  b. Currently seeing a cognitive‐behavioural therapist, psychotherapist, or counsellor  c. People unable to understand and fill out the questionnaires
Interventions	Stated purpose: to evaluate the effectiveness of a modified behavioural activation treatment (MBAT) intervention on reducing depressive symptoms in rural left‐behind elderly  INTERVENTION (n = 37) Name: Modified Behavioral Activation Treatment (MBAT) and regular care Delivered by: LHW and PHP Title/name of PW and number: facilitators ‐ 4, volunteers ‐ 10   1. Selection: 4 postgraduate nursing students served as facilitators. In addition, 10 local nurses familiar with the dialect were recruited as volunteers and were trained over a period of 2 days on test procedures and questionnaire items. These 10 volunteers were primarily responsible for the survey and for regional dialect co‐ordination 2. Educational background: postgraduate nursing students, local nurses  3. Training: no details of training provided. Facilitators were trained by the investigator to administer treatments. Volunteers were trained over a period of 2 days on test procedures and questionnaire items 4. Supervision: each facilitator worked with 1 of the 4 groups of participants in the intervention period in close collaboration with the investigator. To ensure competent provision of intervention, all facilitators met for weekly individual supervision sessions with the investigator  5. Incentives/remuneration: not mentioned  Intervention details  1. Duration/frequency: 8 weekly sessions, each lasting 2 hours  2. Content of intervention: before the first session, participants in the intervention group were divided into 4 groups. Then, a treatment rationale and outline for BA for depression were provided to participants by the Co‐PI. The daily monitoring form (a blank calendar of the hours of each day of a week) was typically assigned as homework but was started in session to ensure that participants understood the concepts and expectations. Strategies such as enlisting a supportive family member to help with written assignments were encouraged. After submitting the assignment, local volunteers translated the log and confirmed it with participants  CONTROL: usual care ("regular care") (n = 36) Included regular physical examinations, such as checking temperature, blood pressure, heart, and breathing rate, and reviewing current health symptoms depending on health needs, and receiving education (knowledge on prevention of and treatments for chronic diseases, infectious diseases, etc.), delivered by village doctors weekly during the 8‐week intervention period  CO‐INTERVENTIONS: both groups received regular care
Outcomes	(Participants were given 30 to 45 minutes to finish the questionnaires; reading assistance was provided when necessary by nurse volunteers) Patients   1. Beck Anxiety Inventory (BAI) (score ≥ 45 was defined as positive for anxiety)  2. Oxford Happiness Questionnaire (OHQ)  3. Geriatric depression scale (long form) (remission of depression symptoms was defined as score < 11 on the GDS)  Carers  Nil  Process/health workers Nil Economic outcomes  Nil  Time points: baseline, 3 months post‐intervention
Notes	Source of funding: China Family Foundation Health Fellowship Program of Yale‐China Association and the National Natural Science Foundation of China (NO.81502701)  Notes on validation of instruments (screening and outcomes): all locally validated  Additional information: declarations of interest ‐ none Handling the data: as per footnotes in data and analysis  Prospective trial registration numbers: ChiCTR‐IOR‐17011289 and NCT02785211
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A random number table was used to allocate the 80 participants randomly to 2 groups of 40 each
Allocation concealment (selection bias)	Unclear risk	Judgement comment: both reviewers rated unclear; no mention of any effort made to conceal allocation of participants. Participants were allocated to their groups at the start of the study
Blinding of participants and personnel (performance bias) All outcomes	High risk	Outcomes were self‐assessed. Facilitators read out questions to participants while participants filled out their responses. Participants are aware of their treatment assignment and may choose to answer favourably if they are in the intervention group because of pressure from attending the intervention or from facilitators who are translating and reading out questions on the questionnaires
Blinding of outcome assessment (detection bias) all outcomes	High risk	Outcomes were self‐assessed. Facilitators read out questions to participants while participants filled out their responses. Participants are aware of their treatment assignment and may choose to answer favourably if they are in the intervention group because of pressure from attending the intervention or from facilitators who are translating and reading out questions on the questionnaires
Baseline outcome measurements similar	Low risk	Baseline GDS, BAI, and OHQ score were similar in both groups
Baseline characteristics similar?	Low risk	Partipants in the 2 groups were similar in age, gender, education, economic means, visit frequency by children, previous exposure to psychoeducation or counselling, and need for psychoeducation or counselling
Incomplete outcome data (attrition bias) Efficacy data	Low risk	Yen ‐ unclear: dropout numbers were small (3/80 immediately post intervention and 7/80 3 months post intervention), but because effect sizes are small, it is possible for participants who had dropped out to have an influence on study results if they had stayed on Rakesh ‐ low
Protection against contamination	Unclear risk	Participants in both groups lived in the same area and may have shared information with each other
Selective reporting (reporting bias)	Unclear risk	No clinical trial registry data available. Outcomes planned in methods section were reported in results section
Other bias	Low risk	No comment