Noninvasive Surrogate for Physiologic Dead Space Using the Carbon Dioxide Ventilatory Equivalent: Testing in a Single-Center Cohort, 2017–2023*

Abstract

OBJECTIVES: We sought to evaluate the association between the carbon dioxide (co 2) ventilatory equivalent (VEqco 2 = minute ventilation/volume of co 2 produced per min), a marker of dead space that does not require a blood gas measurement, and mortality risk. We compared the strength of this association to that of physiologic dead space fraction (VD/Vt = [Paco 2–mixed-expired Pco 2]/Paco 2) as well as to other commonly used markers of dead space (i.e., the end-tidal alveolar dead space fraction [AVDSf = (Paco 2–end-tidal Pco 2)/Paco 2], and ventilatory ratio [VR = (minute ventilation × Paco 2)/(age-adjusted predicted minute ventilation × 37.5)]). DESIGN: Retrospective cohort data, 2017–2023. SETTING: Quaternary PICU. PATIENTS: One hundred thirty-one children with acute respiratory distress syndrome. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: All dead space markers were calculated at the same 1-minute timepoint for each patient within the first 72 hours of using invasive mechanical ventilation. The 131 children had a median (interquartile range, IQR) age of 5.8 (IQR 1.4, 12.6) years, oxygenation index (OI) of 7.5 (IQR 4.6, 14.3), VD/Vt of 0.47 (IQR 0.38, 0.61), and mortality was 17.6% (23/131). Higher VEqco 2 (p = 0.003), VD/Vt (p = 0.002), and VR (p = 0.013) were all associated with greater odds of mortality in multivariable models adjusting for OI, immunosuppressive comorbidity, and overall severity of illness. We failed to identify an association between AVDSf and mortality in the multivariable modeling. Similarly, we also failed to identify an association between OI and mortality after controlling for any dead space marker in the modeling. For the 28-day ventilator-free days outcome, we failed to identify an association between VD/Vt and the dead space markers in multivariable modeling, although OI was significant. CONCLUSIONS: VEqco 2 performs similarly to VD/Vt and other surrogate dead space markers, is independently associated with mortality risk, and may be a reasonable noninvasive surrogate for VD/Vt.