Reproduce a published study: Ranzani 2021, COVID-19 hospital admissions in Brazil
Tutorial · CohortDAG 0.2.9
Find the study, get its data and load it
Find the study in Pipelines
Click Pipelines. Under Studies, find the 2021 study by Ranzani and colleagues. It describes the first 250,000 COVID-19 hospital admissions in Brazil.

Get the data
The study needs an extract of Brazil's SIVEP-Gripe hospital surveillance records. Click Get data. CohortDAG downloads it and checks it against its recorded checksum before marking the data ready.

Load the pipeline and fit it to the window
Click Load. The pipeline opens on the canvas, laid out by its authors, with the results of an earlier run already restored. Close the results panel to give the graph room, then click Fit Graph.

Read the method, run it and read the result
Read the method in the node guide
The mortality comes from a Proportions & CIs node. Open the Nodes tab and click the info button next to Proportions & CIs. Its guide explains the proportion and its confidence intervals.

Run the pipeline
Press Escape to close the guide, then click Run All Outputs. CohortDAG compiles the pipeline and runs it on Apache Spark. When the run finishes, the results panel opens; close the results panel to keep the whole graph in view.

Read the result
Select the overall mortality node and open the Workspace tab. Of 232,036 patients with a known outcome, 87,515 died in hospital. That is 37.7%, which the paper reports as 38%.
