Figure - PMC

Skip to main content

An official website of the United States government

Here's how you know

Here's how you know

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

View full-text article in PMC

. 2018 Apr 26;35(6):588–594. doi: 10.1177/0885066618772498

Search in PMC
Search in PubMed
View in NLM Catalog
Add to search

© The Author(s) 2018

This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).

PMC Copyright notice

Figure 3. — Relative variable importance for sequential random forest models performed from inclusion in the study and up to 14 days; panels (B) and (C) present the results for the non-acute respiratory distress syndrome (non-ARDS) and ARDS subgroups. Patients included in each model included those still alive in the intensive care unit (ICU) until the model reference day and which did not have ventilator-associated lower respiratory tract infection (VA-LRTI) until that moment. The relative importance was calculated as the mean Gini decrease for each variable divided over the total sum of Gini values for the model. Notice how the relative importance of Simplified Acute Physiology Score 2 (SAPS 2) decreased in patients without ARDS while the importance of Barthel index and VA-LRTI increased. For patients with ARDS, no clear trend was seen, with a questionable decrease in the importance of VA-LRTI over time.