Deliverable D3.3 of the H2020 ALTERNATIVE project (environmentAL Toxicity chEmical mixtuRes through aN innovative platform based on aged cardiac tissue model, Grant agreement ID: 101037090). Protocol for integration .

Background: The assessment of human health risks due to the environmental exposure to chemicals is crucial to support evidence-informed public health decision-making and to establish safe environmental regulations. Data integration of epidemiological and toxicological evidence is a well-established requirement in chemicals toxicity assessment. Nonetheless, rigorous and systematic methods are not exhaustively described and developed to accomplish it. The aim of this deliverable is to disclose the methodologies used in the evidence integration phase of the ALTERNATIVE project to provide a systematic, transparent and integrated approach which combines multiple strands of evidence. Methods: We conducted a review of the existing frameworks for the integration of epidemiological and toxicological data to develop a state of the art on procedures and methods proposed by other authors and to identify unresolved limitations and shortcomings. Results: The phase of evidence integration in the ALTERNATIVE project incorporates current best practices and adds on some innovative steps suggesting the use of more appropriate evaluation tools. In particular, improvements concern: 1) transparency and systematicity of literature search by reporting on methodological steps in adherence to appropriate international guidance; 2) adoption of tailored quality assessment tools for observational studies to upgrade the contribution of critical human evidence in risk assessment; 3) improvement of the level of certainty in the knowledge of the causal links by combining the "system of evidence", both human and non-human, by means of an "a priori" identification of the earlier health outcomes ("markers of effect"); 4) support causality links observed in epidemiological and toxicological studies by taking into account the adverse outcome pathway (AOP) framework. Conclusions: The integration of different strands of evidence can be improved by taking procedural and organizational precautions aiming to: (i) include regulators in the collaborative team of experts to define a relevant research question to support conclusions; (ii) before starting the systematic reviews of evidence on humans and non-humans, define the search strategies subordinated to the criteria of maximizing the confidence on the causal link and the relevance of the effect for human populations; and (iii) develop an AOP framework based on the epidemiological and toxicological conclusions achieved.