Case Description

From SICDB Doc
Jump to navigation Jump to search

Case Selection

The SICdb dataset contains all admissions to metavision, that were treated on an intensive care unit. This includes patients that are admitted post-surgery as well as internal medicine cases. Note that non-surgical reasons for admission are highly underrepresentated in this dataset because most internal medicine patients are treated on a unit not using an electric PDMS.

All admissions to participating intensiv care units are included, as long as following conditions do not apply:

  • Less the 10 minutes of Stay
  • Missing mandatory admission data
  • No monitor signals
  • No medication data
  • hospital number mismatch

Intensive care or intermediate care?

To provide as much data as possible all patients admitted to one of our intensive or intermediate care units are included. While the associated beds have a designated care level, the patients treatment level is not defined and are treated according to their needs, they may be transferred multiply between the strongly cooperating units. In general, patients that are uniquely associated (see unitlog) to INIC or INID did not receive high level intensive care treatment. Please note that the field cases.HospitalUnit gives no information about the level of intensive care a patient received. We may provide a preprocessed field in future versions of the SICdb dataset concerning the level of treatment.

Unit Description

The participating metavision enabled wards are CWIN, INBD, INIC, and INIC. INCV and CWCV are used for cases associated with SARS-COV-2, this includes SARS-COV-2 ARDS and patients with coincidental positive test and other cause of intensive care. CWIN and INBD, INCV and CWCVC are high level intensive care units. INIC and INID are lower level of care and do not provide invasive mechanical ventilation (NIV possible on INID, HFNC on all wards), CRRT or ECMO.


All cases are cross-matched with a second database. In case of unsuccessful matching, cases were excluded, as these refer to entries for testing and demo.

The dataset is in constant development. Feel free to contact us when you find a significant amount of invalid data.


All data is pseudo anonymized as defined by the European General Data Protection Regulation, Article 4(5). (The European Parliament, 2016) The deidentification strategy additionally complies with the US regulations for health data, HIPAA (“Guidance Regarding Methods for De-identification of Protected Health Information in Accordance with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule”). Please handle this data with care and note that any efforts to reidentify cases are illegal.