1. Identification of Acute Respiratory Distress Syndrome subphenotypes denovo using routine clinical data: a retrospective analysis of ARDS clinical trials
- Author
-
Osborn J, Israel Silva Maia, Lucas Bulgarelli, Ary Serpa Neto, Kast R, Abhijit Duggal, Rey D, Alexandre Biasi Cavalcanti, Fernando G. Zampieri, Laranjeira Ln, Octávio Deliberato R, Matthew Siuba, Paisani Dm, and Van Ark E
- Subjects
medicine.medical_specialty ,education.field_of_study ,ARDS ,business.industry ,Population ,medicine.disease ,law.invention ,Clinical trial ,Randomized controlled trial ,law ,Internal medicine ,Cohort ,medicine ,Biomarker (medicine) ,education ,business ,Prospective cohort study ,Cohort study - Abstract
RationaleThe acute respiratory distress syndrome (ARDS) is a heterogenous condition, and identification of subphenotypes may help in better risk stratification.ObjectivesIdentify ARDS subphenotypes using new simpler methodology and readily available clinical variables.DesignRetrospective Cohort Study of ARDS trials.SettingData from the U.S. ARDSNet trials and from the international ART trial.Participants3763 patients from ARDSNet datasets and 1010 patients from the ART dataset.Primary and secondary outcome measuresThe primary outcome was 60-day or 28-day mortality, depending on what was reported in the original trial. K-means cluster analysis was performed to identify subgroups. For feature selection, sets. Sets of candidate variables were tested to assess their ability to produce different probabilities for mortality in each cluster. Clusters were compared to biomarker data, allowing identification of subphenotypes.ResultsData from 4,773 patients was analyzed. Two subphenotypes (A and B) resulted in optimal separation in the final model, which included nine routinely collected clinical variables, namely: heart rate, mean arterial pressure, respiratory rate, bilirubin, bicarbonate, creatinine, PaO2, arterial pH, and FiO2. Participants in subphenotype B showed increased levels of pro-inflammatory markers, had consistently higher mortality, lower number of ventilator-free days at day 28, and longer duration of ventilation compared to patients in the subphenotype A.ConclusionsRoutinely available clinical data can successfully identify two distinct subphenotypes in adult ARDS patients. This work may facilitate implementation of precision therapy in ARDS clinical trials.ARTICLE SUMMARYStrengths and limitations of this studyLargest cohort of patients used to identify subphenotypes of ARDS patients.Subphenotypes were validated in the population of a large international ARDS randomized controlled trial.Subphenotypes were identified by using only routinely collected clinical data.Our use of data exclusively from randomized controlled trials does not prove generalizability to unselected ARDS populations.The clinical utility of the subphenotypes have to be validated in a prospective study.
- Published
- 2021
- Full Text
- View/download PDF