Systematic Review and Weight of Evidence Are Integral to Ecological and Human Health Assessments: They Need an Integrated Framework

Abstract

Scientific assessments synthesize the various results of scientific research for policy and decision making. Synthesizing evidence in environmental assessments can involve either or both of 2 systems: systematic review (SR) and weight of evidence (WoE). Systematic review was developed to systematically assemble results of clinical trials to be combined by meta-analysis. Weight-of-evidence approaches have evolved from jurisprudence to make inferences from diverse bodies of evidence in various fields. Our objectives are to describe the similarities and differences between SR and WoE and to suggest how their best practices can be combined into a general framework that is applicable to human health and ecological assessments. Integrating SR and WoE is based on the recognition that 2 processes are required, assembling evidence and making an inference. Systematic review is characterized by methodical literature searching, screening, and data extraction, originally for meta-analysis but now for various inferential methods. Weight of evidence is characterized by systematically relating heterogeneous evidence to considerations appropriate to the inference and making the inference by weighing the evidence. Systematic review enables the unbiased assembly of evidence from literature, but methods for assembling other information must be considered as well. If only 1 type of quantitative study estimates the assessment endpoint, meta-analysis is appropriate for inference. Otherwise, the heterogeneous evidence must be weighed. A framework is presented that integrates best practices into a methodical assembly and weighing of evidence. A glossary of terms for the combined practice and a history of the origins of SR and WoE are provided in Supplemental Data. Integr Environ Assess Manag 2020;16:718-728. Published 2020. This article is a US Government work and is in the public domain in the USA.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 35 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.