End-Tidal Alveolar Dead Space Fraction Calculated With Venous PCO2 Is Associated With Mortality Risk

Monday, February 24, 2025

1:30pm – 2:30pm Eastern Time

Location: Connections Central Theater 1

Debra J. Rosenbaum, MD (she/her/hers)

Childrens Hospital Los Angeles

Disclosure(s): No relevant financial relationship(s) to disclose.
CN

Christopher J. L. Newth, MD, ChB

Director of Research, Professor of Pediatrics
Children's Hospital Los Angeles, USC

Disclosure information not submitted.
RK

Robinder Khemani, MD, MS

Associate Director of Research, Professor of Pediatrics
Children's Hospital Los Angeles

Disclosure information not submitted.
JH

Justin Hotz, BSRT, RRT-NPS

Software Engineer
Children's Hospital Los Angeles

Disclosure information not submitted.
AB

Anoopindar Bhalla, MD

Associate Professor of Clinical Pediatrics
Children's Hospital Los Angeles, USC

Disclosure information not submitted.

Disclosure(s):

Debra Rosenbaum, MD: No relevant financial relationship(s) to disclose.

Introduction: The end-tidal alveolar dead space fraction (AVDSf = [PaCO2-PETCO2]/PaCO2) is strongly associated with mortality risk and fewer 28-day ventilator-free days (VFD) in critically ill children. As arterial blood gases are being used less frequently, we sought to determine if there is a similar association with mortality when a venous PCO2 drawn from a central venous catheter (VBG) is used to estimate AVDSf.

Methods: Retrospective, single-center observational study of children admitted to the pediatric ICU (non-cardiac unit) between 2013-2021 who received invasive mechanical ventilation (IMV) and had a VBG within 72 hours of IMV start. The first VBG was used in the analysis. Patients without required data and those on repeat ICU admission were excluded. Oxygenation Index (OI = [Mean Airway Pressure x FiO2 x 100]/PaO2) was calculated using the Sauthier et al. estimate for PaO2 from SpO2.

Results: Of 687 patients included, the median age was 3.9 years (IQR 0.9, 12.3) and 17.9% had immunosuppression. The median SpO2 was 99% (IQR 97, 100), OI was 4.4 (IQR 2.6, 9.3), and 20.7% died.
In univariate analyses, AVDSf and OI were associated with mortality risk (MR): AVDSf (OR 1.96 [95% CI 1.62-2.36], p< 0.001); OI (OR 1.50 [95% CI 1.27-1.78], p< 0.001). AVDSf continued to be associated with MR in multivariable modeling (OR 1.47 [95% CI 1.18-1.84], p=0.001) when adjusting for OI and overall severity of illness (defined by PRISM III); OI no longer had an association (OR 1.13 [95% CI 0.92-1.39], p=0.23).
An association with fewer VFD was seen for both AVDSf (HR 0.72 [95% CI 0.66-0.79], p< 0.001) and OI (HR 0.74, [95% CI 0.66-0.82], p< 0.001) in univariate analyses. This association continued in multivariable modeling for AVDSf (HR 0.85 [95% CI 0.77-0.93], p=0.001) and OI (HR 0.86 [95% CI 0.78-0.95], p=0.002) when adjusting for AVDSf, OI and PRISM III.
Among survivors, AVDSf and OI were associated with a longer length of IMV (AVDSf: HR 0.90 [95% CI 0.82-0.99], p=0.031; OI: HR 0.82 [95% CI 0.74-0.91], p< 0.001) in univariate analyses. However, OI alone was associated with longer length of IMV in multivariable analysis when adjusting for AVDSf and PRISM III (HR 0.84 [95% CI 0.75-0.94], p=0.002).

Conclusions: AVDSf calculated with a VBG outperforms OI for risk stratification of mortality in critically ill children.