Databases
MIMIC
The MIMIC database, which stands for Medical Information Mart for Intensive Care, is a North American database containing data from over 50,000 patients admitted to intensive care. It is one of the most widely used intensive care databases, due to its free access.
Despite having data of imperfect quality, it serves as a good foundation for learning to manipulate data from health data warehouses (HDWs).
It exists in several versions, with the most recent being MIMIC-IV.
Do you need programming knowledge?
No, it is not necessarily required to have programming knowledge to manipulate the data in this database.
Handling the database itself will require programming knowledge, especially in SQL. It is a database with a particular schema, which is difficult to grasp.
However, it is possible to manipulate this data entirely with a graphical interface: this is one of the reasons why LinkR was created.
How to access the data?
The MIMIC database includes test databases for versions III and IV, which contain anonymized data from 100 patients and are publicly accessible.
You can download the data here:
- MIMIC-III test: data with the MIMIC data schema
- MIMIC-IV OMOP test: data with the OMOP data schema
To access the complete databases, you need to complete a few steps.
Visit the MIMIC-III database page.
You will see this box at the bottom of the page:
- be a credentialed user
- complete required training:
- CITI Data or Specimens Only Research You may submit your training here.
- sign the data use agreement for the project
You must start by registering on the site physionet.org.
You will need to submit an access request to Physionet, providing some information and the contact details of a supervisor or colleague, who will receive an email.
You will then need to complete the CITI Course, a necessary training to access data hosted on the Physionet site. The different steps are detailed here.
You can then download the certificate once the CITI Course is completed, and you can submit it here for validation by the Physionet team.
Finally, you will need to sign the data use agreement.
Use cases
ICU dashboard
Overview
This dashboard provides a visualization of quality indicators in intensive care units.
One of its advantages is that it can be configured directly by clinicians without the need for prior programming knowledge.
It is organized into several tabs:
-
Demographics: This section shows the number of patients admitted during the selected period, as well as their demographic information (age, gender), mortality rate, number of admissions, readmission rate, and the main ICD-10 diagnoses of the stays.
-
Ventilation: This section provides data on mechanical ventilation, including the number of patients on ventilators, the duration of ventilation, the extubation failure rate, the self-extubation rate, and the selected ventilation parameters.
-
Sedation: Here you will find the medications used for sedation, the total duration of sedation, neuroleptic consumption, etc.
-
Dialysis: This section provides information on the number of patients undergoing dialysis and the type of dialysis used.
Installation
To install this project, you can follow the “Quick start” tutorial in the documentation.
You can also follow these tutorials if you wish to import your own data:
- Install LinkR
- Import data
- Install a project: Select the DOMASIA laboratory in the city of Rennes (France), and you will see the project “Intensive care units quality indicators dashboard” appear under the projects.
Contributions
LinkR is a project under development.
If you encounter any issues during installation or use, or if you have suggestions to improve the application, please contact us at linkr-app@pm.me.
Next steps
For now, only the “Demographics” section is available, with other sections currently under development.
This dashboard is configured to visualize intensive care data. The next step will be to apply this dashboard model to other hospital departments and private medical practices.