A model to predict SARS-CoV-2 infection based on the first three-month surveillance data in Brazil.

Fredi A Diaz-Quijano; Jose Mario Nunes da Silva; Fabiana Ganem; Silvano Oliveira; Andrea Liliana Vesga-Varela; Julio Croda

The pandemic caused by the novel coronavirus, Sars-Cov-2, challenges the capabilities of health care services, especially in low-and middle-income countries. 1 A major issue is to meet the diagnostic requirements of the suspected cases reported to the surveillance system. 2 The proportion of suspected cases being tested in each country is not systematically presented in most of the epidemiological reports. [3] [4] [5] However, with the increasing number of new suspected cases of the disease (COVID- 19) worldwide, diagnosis has clearly become a growing problem, mainly due to the lack or delay in the test results. 6, 7 Clinical manifestations of COVID-19 are unspecific and include respiratory symptoms, fever, cough, dyspnea, and viral pneumonia. 8, 9 Polymerase chain reaction by real-time reverse transcriptase (RT-PCR) is considered the gold standard for the diagnosis of SARS-CoV-2 infection. However, its limited availability and the strict laboratory requirements delay diagnosis, which represents an unprecedented challenge to control transmission and provide timely health care. 10, 11 The incorporation of predictive diagnostic models based on surveillance data could help identify patients who could need specific treatment and early isolation. Consequently, we aimed to describe the profile of COVID-19 patients and to obtain a multiple model to predict the diagnosis among suspected cases reported in Brazil based on data routinely collected by the surveillance system.