Causes and attributable fraction of death from ARDS in inflammatory phenotypes of sepsis

Bruno Evrard,Pratik Sinha,Kevin Delucchi,Carolyn M. Hendrickson,Kirsten N. Kangelaris,Kathleen D. Liu,Andrew Willmore,Nelson Wu,Lucile Neyton,Emma Schmiege,Antonio Gomez,V. Eric Kerchberger,Ann Zalucky,Michael A. Matthay,Lorraine B. Ware,Carolyn S. Calfee
DOI: https://doi.org/10.1186/s13054-024-04943-x
IF: 15.1
2024-05-15
Critical Care
Abstract:Hypoinflammatory and hyperinflammatory phenotypes have been identified in both Acute Respiratory Distress Syndrome (ARDS) and sepsis. Attributable mortality of ARDS in each phenotype of sepsis is yet to be determined. We aimed to estimate the population attributable fraction of death from ARDS (PAF ARDS ) in hypoinflammatory and hyperinflammatory sepsis, and to determine the primary cause of death within each phenotype.
critical care medicine
What problem does this paper attempt to address?
The problem this paper attempts to address is: The proportion of mortality attributable to acute respiratory distress syndrome (ARDS) in different inflammatory phenotypes of sepsis (PAFARDS) is not yet clear. Specifically, the study aims to estimate the population attributable fraction of mortality due to ARDS (PAFARDS) in sepsis patients with low and high inflammatory phenotypes and to identify the primary causes of death in each phenotype. ### Main Research Objectives: 1. **Estimate PAFARDS within 60 days of hospitalization**: Estimate the population attributable fraction of mortality due to ARDS in sepsis patients with low and high inflammatory phenotypes, respectively. 2. **Identify causes of death**: Identify the primary causes of death in sepsis patients with low and high inflammatory phenotypes to further explain the results of the PAFARDS analysis. ### Research Background: - **ARDS**: Acute respiratory distress syndrome, with a mortality rate of 35-45%, presents significant heterogeneity, posing challenges for effective treatment. - **Inflammatory phenotypes**: Based on plasma biomarker levels, ARDS can be divided into low and high inflammatory phenotypes. The high inflammatory phenotype accounts for about one-third of ARDS cases and is associated with high levels of inflammatory biomarkers, higher use of vasopressors, and higher mortality rates. The low inflammatory phenotype, on the other hand, accounts for about two-thirds of ARDS cases and is associated with lower levels of inflammatory biomarkers and lower mortality rates. - **PAFARDS**: Refers to the proportion of deaths that could be prevented if ARDS were eliminated. This metric is significant for designing clinical trials, especially when PAFARDS is low, requiring larger sample sizes to detect changes in mortality rates. ### Research Methods: - **Participants**: 1737 sepsis patients from two prospective cohort studies. - **Definition**: ARDS defined according to the Berlin criteria, with patients classified into low and high inflammatory phenotypes based on latent class analysis (LCA). - **Statistical Analysis**: PAFARDS estimated using indirect standardization methods, with multiple sensitivity analyses including propensity score matching. ### Main Results: - **PAFARDS**: 19% (95% CI 10, 28%) in low inflammatory sepsis, and 14% (95% CI 6, 20%) in high inflammatory sepsis. - **Differences in causes of death**: The primary cause of death in low inflammatory sepsis was respiratory failure (59%), while in high inflammatory sepsis it was circulatory failure (63%). - **Severe underlying conditions**: 81% of deaths in low inflammatory sepsis had severe comorbidities, compared to 67% in high inflammatory sepsis. ### Conclusion: - **PAFARDS** is relatively low in both phenotypes, but there are significant differences in the primary causes of death between the phenotypes. - **Design of future clinical trials**: These findings suggest that larger sample sizes or more stringent selection of trial populations may be needed to detect changes in mortality rates in ARDS clinical trials. Additionally, these results provide important insights into the role of ARDS in different phenotypes and its impact on mortality rates.