COVID-19 diagnosis prediction in emergency care patients: a machine learning approach

Andre Filipe de Moraes Batista; Joao Luiz Miraglia; Thiago Henrique Rizzi Donato; Alexandre Dias Porto Chiavegatto Filho

During the last month, coronavirus disease 2019 (COVID-19) has spread rapidly from China to European countries, and is now experiencing exponential growth in the United States and Canada. 1 Recently, there has also been a sharp increase in cases in developing countries, but due to a limited number of COVID-19 tests available for these countries, the real extent of the spread is still unknown, which may lead to dire consequences in terms of adequate clinical care and the epidemiological knowledge needed to guide public containment measures. 2, 3 The most common symptoms of COVID-19, such as fever and cough, are similar to a range of other infectious diseases, making prompt diagnosis a challenge for health professionals. 4 Results from the transcription polymerase chain reaction test (RT-PCR), currently the most reliable diagnostic test, are frequently taking more than a week to become available according to reports in the Brazilian media, while in the meantime there is a need for immediate decisions about clinical care and preventive measures. On the other hand, the recent increase in the usage of new rapid diagnostic tests, which are prone to some accuracy issues, may increase the risk of inefficient allocations of health resources.