SESSION TYPE: Respiratory Support Posters
PRESENTED ON: Wednesday, October 24, 2012 at 01:30 PM - 02:30 PM
PURPOSE: Non-Invasive Ventilation (NIV) is utilized in patients with acute respiratory failure (ARF) due to a COPD exacerbation (high compliance lungs) and CHF exacerbation (low compliance). Positive end-tidal carbon dioxide (PetCO2) determinations are not routinely utilized in NIV as the mask leak may prevent accurate measurements, especially at low tidal volumes with high leak rates. We conducted a bench study with an artificial lung to determine the accuracy of a compensation algorithm on PetCO2 measurements.
METHODS: We connected a V60 ventilator to a NIV mask placed on a mannequin head that was connected to a test lung via a simulated trachea (ET tube) and infused CO2 into the test lung. CO2 and flow were measured at a reference site (trachea) and a test site between the mask and exhalation port. Test data was collected across a range of airway pressures, respiratory rates, resistances, compliances, leak rates and CO2 infusion rates. A compensation algorithm (custom software) was used to predict the reference PetCO2 from the test flow and CO2 signals.
RESULTS: The average error across all settings between compensated and reference PetCO2 was 0.3 ± 5.0 mmHg or 3.0 ± 12.4 percent (mean ± SD). Linear regression yielded a slope and offset of 0.97 and 1.58 respectively with r2 = 0.97. For high compliance settings the error was 0.03 ± 4.6 mmHg or 2.7 ± 15.2%. Linear regression slope and offset were 0.93 and 2.0 with r2 = 0.95. For low compliance settings error was -0.34 ± 6.5 mmHg or 0.6 ± 9.4%. Linear regression slope and offset were 0.97 and 1.6 with r2 = 0.96.
CONCLUSIONS: This study indicates the mask leak compensation algorithm improves the accuracy of PetCO2 measurement during NIV, regardless of lung compliance, airway resistance, mask leak and respiratory rate.
CLINICAL IMPLICATIONS: Our algorithm may be applied to COPD and CHF patients in ARF managed with NIV regardless of mask fit and leak to predict PetCO2 values, which may be used in place of blood gases.
DISCLOSURE: Joseph Orr: Grant monies (from industry related sources): Phillips - Research Support
Lara Brewer: Grant monies (from industry related sources): Phillips - Research Support (Salary and Equipment)
The following authors have nothing to disclose: Branden Rosenhan, Boaz Markewitz, Simon Rodriguez
No Product/Research Disclosure InformationUniversity of Utah, Salt Lake City, UT