Back

Development of a tool to prioritize the monitoring of COVID-19 patients by public health teams

Vecino-Ortiz, A. I.; Guzman-Tordecilla, N.; Guzman Ruiz, Y. F.; Penaloza-Quintero, R. E.; Fernandez-Nino, J. A.; Ruiz Gomez, F.; Trujillo, A. J.

2021-04-11 health policy
10.1101/2021.04.08.21254922 medRxiv
Show abstract

BackgroundIn the context of the COVID-19 pandemic, public health teams have struggled to conduct monitoring for confirmed or suspicious COVID-19 patients. However, monitoring these patients is critical to improving the chances of survival, and therefore, a prioritization strategy for these patients is warranted. This study developed a monitoring algorithm for COVID-19 patients for the Colombian Ministry of Health and Social Protection (MOH). MethodsThis work included 1) a literature review, 2) consultations with MOH and National Institute of Health officials, and 3) data analysis of all positive COVID-19 cases and their outcomes. We used clinical and socioeconomic variables to develop a set of risk categories to identify severe cases of COVID-19. ResultsThis tool provided four different risk categories for COVID-19 patients. As soon as the time of diagnosis, this tool can identify 91% of all severe and fatal COVID-19 cases within the first two risk categories. ConclusionThis tool is a low-cost strategy to prioritize patients at higher risk of experiencing severe COVID-19. This tool was developed so public health teams can focus their scarce monitoring resources on individuals at higher mortality risk. This tool can be easily adapted to the context of other lower and middle-income countries. Policymakers would benefit from this low-cost strategy to reduce COVID-19 mortality, particularly during outbreaks.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.