Pediatric acute respiratory distress syndrome (PARDS) is a severe and diffuse lung injury that is a common cause of admission and mortality in the pediatric intensive care unit (PICU). PARDS can be secondary to many different causes, and there are few therapies that have been shown beneficial in PARDS. This study seeks to identify important PARDS subtypes using gene expression profiling of bronchial epithelial cells from control and PARDS subjects.

Enrolled subjects will have nasal brushings collected at days 1, 3, 7, and 14 of intubation with collection of serum at these same time points. Brushing RNA will be processed by mRNA-Seq for gene expression analysis and compared to previously published serum biomarkers (interleukin-8, advanced glycosylation end-product specific receptor, and angiopoietin-2) to assess correlation and ability to discriminate PARDS endotypes. Changes in gene expression over time will be assessed to define a PARDS recovery gene expression signature, and correlation between bronchial and nasal gene expression will be determined.

Inclusion Criteria:

All potential participants must:

1. Be aged zero to 18 years (both control and ARDS, not age matched)
2. Be admitted to the PICU with expected duration of hospitalization 7 days or greater.

ARDS patients must:

1. Have acute changes in chest x-ray (CXR)

2. Have a known or suspected insult within the prior 7 days that is consistent with ARDS

3. Have an oxygenation index (OI) of 4 or greater or and oxygen-sat index (OSI) of 5 or greater

   1. OI = mean airway pressure X fraction inspired oxygen (FiO2) / arterial oxygen partial pressure (PaO2)
   2. OSI = mean airway pressure X FiO2 / oxyhemoglobin saturation (SpO2) with sat <= 97%.

Exclusion Criteria:

1. Have a baseline oxygen requirement of 2 liters of oxygen or greater at home
2. Have disruption of the nasal passages
3. Have a history of excessive bleeding or known bleeding disorders
4. Be at high risk of bleeding
5. Have a do not resuscitate (DNR) or Limited Resuscitation Order

Control subjects will be patients admitted to the PICU for non-lung injury related conditions.

PARDS subjects will be intubated patients with PARDS in the PICU.

Study info will be shared with interested investigators on a case by case basis. Microarray data will be posted on an archive such as GeoDatasets but we are unclear how correlative clinical and serum biomarker data would be shared.