This cluster-randomized waitlist-control trial evaluated Multifamily Healing Spaces (MFHS), a culturally adapted behavioral family intervention for genocide-affected communities in Rwanda. Community members across five districts were screened through a staged triage process. Individuals with acute or severe clinical concerns were routed to clinical referral, those with individual mental-health needs were prioritized for Resilience-Oriented Therapy, and families were invited to MFHS when at least one screened family member met criteria for family communication difficulties. Twenty eligible clusters were randomly selected and then allocated to MFHS or waitlist control. The baseline cohort included 319 family members aged 12 years or older. The trial tested whether MFHS improved family resilience, parenting, couple and youth family-system outcomes, and secondary mental health, psychosocial, community, and livelihood outcomes from baseline to immediate post-intervention endline.

MFHS is a group-based, trauma-informed, family-systems intervention developed for post-genocide Rwanda. It combines psychoeducation, structured dialogue, family communication and problem-solving exercises, parenting and couple-relational work, and youth voice and identity activities. The intervention was delivered in intergenerational, whole-family, parent-only, and youth-only formats. MFHS comprised 18 core 3-hour meeting occasions, with parent-only and youth-only sessions delivered in parallel, plus four planned follow-up sessions where relevant.

Participants were identified through a broader Rwanda societal-healing screening process. Screening was conducted at the individual level, but MFHS invitation occurred at the family level: if any screened individual met the MFHS family-communication criterion, that person's family was invited. The audited screening workbook contained 4,214 adult/community screening records. After referral/triage for severe clinical flags and other staged interventions, 839 screened individuals were staged to MFHS. At the cell level, 56 eligible cells had enough staged MFHS-screened individuals to form a group; those cells contained 794 staged MFHS-screened individuals. Twenty trial clusters were randomly selected from the eligible cells and then allocated to MFHS or waitlist control. The final baseline cohort included 319 family members in 20 clusters: 167 in MFHS clusters and 152 in waitlist-control clusters.

The primary manuscript analysis is a baseline-cohort intention-to-treat analysis. All baseline-assessed participants were retained in their final fielded cluster allocation. Missing endline outcomes for 36 participants were handled through multiple imputation. Effects were estimated with baseline-adjusted cluster-level Analysis of Covariance (ANCOVA) weighted by baseline cluster size, with complete-case and unweighted analyses retained as sensitivity checks.

Inclusion Criteria:

• Member of a family living in or connected to a selected MFHS-eligible cluster/cell in one of the five study districts in Rwanda.
• Aged 12 years or older.
• Family invited because at least one screened family member met the staged MFHS criterion for family communication difficulties after higher-priority clinical and individual mental-health triage.
• Able and willing to participate in baseline assessment and, if allocated to MFHS, family/group sessions.
• Written informed consent, assent, and/or caregiver permission obtained according to age and local ethics requirements.

Exclusion Criteria:

• Individuals with suicidality, psychosis, or severe substance-use flags were routed for clinical interview or hospital referral rather than MFHS as first-line allocation.
• Individuals meeting higher-priority individual mental-health criteria were routed to Resilience-Oriented Therapy before the MFHS screening stage.
• Individuals or families unable to provide consent/assent or safely participate in group family sessions were not enrolled.
• Remaining individuals with social-conflict indicators after earlier stages were routed to sociotherapy rather than MFHS.

De-identified participant-level data underlying the manuscript analyses may be shared with qualified researchers on reasonable request to the first author, subject to ethics, data-protection, and partner-organization approvals.

Available after publication, with no predetermined end date.

Researchers must submit a methodologically sound proposal to the first author. Access will be subject to approval by the study team and any applicable ethics, data-protection, and partner-organization requirements. Shared data will be de-identified.

Baseline enumerators and participants were intended to be unaware of condition assignment during baseline outcome measurement. After allocation, masking could not be assured because local implementation required allocation lists before groups could be convened, and endline masking was difficult to maintain because participants could disclose their allocation status.