We recently shared a first draft of our data access policy within our community for collecting feedback. We got a lot of useful feedback so far and are in close contact with several experts.
On Sunday April 12th 2020 after implementing a data cleaning and data anonymization process we started publishing the public data set of the LEOSS cohort study.
We will provide a comprehensive Jupyter Notebook to explore the data set in more detail very soon.
LEOSS public data set
The LEOSS public data set constitutes patient data from the LEOSS cohort after a data cleaning process, e.g. verifications for a complete documentation of course of disease and plausibility checks. Variables included are:
- age
- gender
- month of diagnosis
- classification of the present clinical phases
- outcome
- presence of administered vasopressor agents
- performed invasive ventilation
- superinfections (proven/suspected/probable as one category)
- remaining symptoms in the recovery phase
The public data set is anonymized using our data protection concept. Several parameters refer to clinical phases. Values where the respective phase was not present have been defined as Not applicable (N/a)
🚨 Please be aware that this is only a part of the complete LEOSS dataset (due to anonymization). This could cause bias. With an increasing number of patients, the LEOSS public dataset will include more patients with rare patient characteristics.
