Background The experience of costly, unplanned care and hospitalisations is not equal across our society; populations living in socioeconomically disadvantaged circumstances where ill-health is more prevalent, experience higher rates. There is a policy, financial and moral imperative to act to address these inequalities and reduce mounting hospital pressures more broadly. Previous systematic reviews have only examined the average effect of healthcare interventions on avoidable hospitalisations, they have not looked at the differential impact of interventions across populations, nor taken a health systems approach and considered the impact of interventions beyond healthcare services on hospitalisations. We aimed to establish which interventions reduce, maintain or increase socioeconomic inequalities in hospitalisations and readmissions.

Methods An electronic search of MEDLINE, Embase, CINAHL, Cochrane CENTRAL and Web of Knowledge was conducted January 1, 1999 to Feb 23, 2022, supplemented with full citation searches of included studies, website searches and expert consultation. Empirical studies involving individuals of any age, in Organisations for Economic Co-operation and Development countries, published in any language were included. Quantitative studies of interventions across any domain of activity (wider public-policy and population health, health and care service, and integrative interventions) were included provided they reported the differential impact of the intervention on either hospitalisations or readmissions across socioeconomic groups (individual or area-based measure of socio-economic status included). Risk of bias was assessed using the Effective Public Health Practice Project tool and narrative synthesis conducted with direction of effect plots.

Results After removing duplicates, 23,496 titles and abstracts were screened, followed by 502 full texts. 37 studies met the inclusion criteria. Studies used a range of quantitative study designs and over half were from the US. Socioeconomic status was variously measured and reported, study quality was variable, and equity impacts mixed. Effective interventions for reducing socioeconomic inequalities were found across all domains of activity and were characterised as either requiring low personal agency to benefit (being legislative in nature and applied at a population or health system level) or for interventions implemented at an individual level, a vaccine programme or intervention focused on holistic integrative care.

Conclusion Evaluating the equity impact of interventions should become more routine; interventions implemented to reduce hospitalisation rates may maintain or increase inequalities, but there are also promising examples of interventions which may reduce inequalities. Whole systems action and investment in non-healthcare activity alongside healthcare intervention is required to improve equity in health and care outcomes.