The following is a summary of “Clustering COVID-19 ARDS patients through the first days of ICU admission. An analysis of the CIBERESUCICOVID Cohort,” published in the March 2024 issue of Critical Care by Ceccato et al.

Acute respiratory distress syndrome (ARDS) can be categorized based on inflammatory response, with prognostic value even if these categories change over time.

Researchers conducted a retrospective study investigating when patients with ARDS can be categorized into multiple groups based on their data, how these groupings change over time, and how the change impacts patient outcomes.

They conducted a multicenter study in Spain involving patients with ARDS linked to COVID-19. Patients were categorized using a clustering algorithm (k-prototypes) based on a mix of continuous and categorical variables available at baseline and day three.

The results showed 6205 patients, 3743 (60%) were included. Silhouette analysis revealed two clusters. Initially, 1402 (37%) patients were in cluster 1, and 2341 (63%) were in cluster 2. By day 3, 1557 (42%) patients were in cluster 1, and 2086 (57%) were in cluster 2. Patients in cluster 2 tended to be older, more hypertensive and had higher rates of shock, organ dysfunction, inflammatory biomarkers, and worse respiratory indexes at both time points. 90-day mortality was notably higher in cluster 2, with 43.8% (n = 1025) versus 27.3% (n = 383) at baseline and 49% (n = 1023) versus 20.6% (n = 321) on day 3. 458 (33%) patients from the first group shifted to the second group by day 3, while 638 (27%) patients from the second group shifted to the first group.

Investigators concluded that while most patients with ARDS remained in the same inflammatory group during the first few days, up to a third switched groups, highlighting the dynamic nature of ARDS and the potential impact on prognosis.

Source: ccforum.biomedcentral.com/articles/10.1186/s13054-024-04876-5