PP41 COVID-19 Modeling To Support Decision Making In Brazil: A Scoping Review

Abstract

Introduction

In the context of the COVID-19 pandemic, which required urgent responses from health systems, and ongoing decision making in a context of limited and evolving evidence, modeling played a significant role in supporting public policy making. Nonetheless, particularly in low and middle-income countries, modeling groups are scarce, and usually not routinely involved in supporting public health policy making. We aimed to appraise COVID-19 modeling work in Brazil during the pandemic.

Methods

We performed a scoping review following PRISMA guidelines to identify groups conducting COVID-19 modeling to support health decision-making in Brazil. Search strategies were applied to MEDLINE, LILACS, Embase, ArXiv, and also included National data repositories and gray literature. We excluded reports of models without modeling results. Titles, abstracts, data repository descriptions and full-text articles identified were read and selected by two reviewers. Data extracted included modeling questions, model characteristics (structure, type, and programming), epidemiologic data sources, main outcomes reported, and parameters. To further identify modeling groups that might have not yet published results, snowball sampling was performed, and a short survey was sent electronically. Investigators and policymakers were invited to an online interview, to obtain further information on how they interacted, communicated, and used modeling results.

Results

We retrieved 1,061 references. After removing duplicates (127), 1,016 abstracts and titles were screened. From an initial selection of 142 abstracts, 133 research groups were identified, of which 67 didn’t meet the eligibility criteria. Of these, 66 groups were invited for an interview, of which 24 were available, including 18 modeling groups from academic institutions, and four groups from State Health departments. Most models assessed the impact of mitigation measures in cases/hospitalization/deaths and healthcare service demand. Interaction and communication with decision-makers were not well established in most groups.

Conclusions

Despite a large number of modeling groups in Brazil, we observed a significant gap in modeling demand and communicating its results to support the decision-making process during the COVID-19 pandemic.