Abstract
BACKGROUND: A 'Health in All Policies' (HiAP) approach has been widely advocated as a way to involve multiple government sectors in addressing health inequalities, but implementation attempts have not always produced the expected results. Explaining how HiAP-style collaborations have been governed may offer insights into how to improve population health and reduce health inequalities. METHODS: Theoretically focused systematic review. Synthesis of evidence from evaluative studies into a causal logic model. RESULTS: Thirty-one publications based on 40 case studies from nine high-income countries were included. Intersectoral collaborations for population health and equity were multi-component and multi-dimensional with collaborative activity spanning policy, strategy, service design and service delivery. Governance of intersectoral collaboration included structural and relational components. Both internal and external legitimacy and credibility delivered collaborative power, which in turn enabled intersectoral collaboration. Internal legitimacy was driven by multiple structural elements and processes. Many of these were instrumental in developing (often-fragile) relational trust. Internal credibility was supported by multi-level collaborations that were adequately resourced and shared power. External legitimacy and credibility was created through meaningful community engagement, leadership that championed collaborations and the identification of 'win-win' strategies. External factors such as economic shocks and short political cycles reduced collaborative power. CONCLUSION: This novel review, using systems thinking and causal loop representations, offers insights into how collaborations can generate internal and external legitimacy and credibility. This offers promise for future collaborative activity for population health and equity; it presents a clearer picture of what structural and relational components and dynamics collaborative partners can focus on when planning and implementing HiAP initiatives. The limits of the literature base, however, does not make it possible to identify if or how this might deliver improved population health or health equity.