Dalarna University's logo and link to the university's website

OBJECTIVES: This umbrella review aimed to evaluate whether certain interventions can mitigate the negative health consequences of caregiving, which interventions are more effective than others depending on the circumstances, and how these interventions are experienced by caregivers themselves.

DESIGN: An umbrella review of systematic reviews was conducted.

DATA SOURCES: Quantitative (with or without meta-analyses), qualitative and mixed-methods systematic reviews were included.

ELIGIBILITY CRITERIA: Reviews were considered eligible if they met the following criteria: included primary studies targeting informal (ie, unpaid) caregivers of older people or persons presenting with ageing-related diseases; focused on support interventions and assessed their effectiveness (quantitative reviews) or their implementation and/or lived experience of the target population (qualitative reviews); included physical or mental health-related outcomes of informal caregivers.

DATA EXTRACTION AND SYNTHESIS: A total of 47 reviews were included, covering 619 distinct primary studies. Each potentially eligible review underwent critical appraisal and citation overlap assessment. Data were extracted independently by two reviewers and cross-checked. Quantitative review results were synthesised narratively and presented in tabular format, while qualitative findings were compiled using the mega-aggregation framework synthesis method.

RESULTS: The evidence regarding the effectiveness of interventions on physical and mental health outcomes was inconclusive. Quantitative reviews were highly discordant, whereas qualitative reviews only reported practical, emotional and relational benefits. Multicomponent and person-centred interventions seemed to yield highest effectiveness and acceptability. Heterogeneity among caregivers, care receivers and care contexts was often overlooked. Important issues related to the low quality of evidence and futile overproduction of similar reviews were identified.

CONCLUSIONS: Lack of robust evidence calls for better intervention research and evaluation practices. It may be warranted to avoid one-size-fits-all approaches to intervention design. Primary care and other existing resources should be leveraged to support interventions, possibly with increasing contributions from the non-profit sector.

PROSPERO REGISTRATION NUMBER: CRD42021252841; BMJ Open: doi:10.1136/bmjopen-2021-053117.