Content-Type: text/html R-LAIR: Riverside Lab for Artificial Intelligence Research

UCR

Computer Science and Engineering



R-LAIR: Riverside Lab for Artificial Intelligence Research

Algorithms to estimate PaCO2 and pH using non-invasive parameters for children with Hypoxemic Respiratory Failure (2014)

by Robinder G. Khemani, E. Busra Celikkaya, Christian R. Shelton, Dave Kale, Patrick A. Ross, Randall C. Wetzel, and Christopher J. L. Newth

Abstract:

Background: Ventilator management for children with hypoxemic respiratory failure may benefit from ventilator protocols, which rely on blood gases. Accurate non-invasive estimates for pH or PaCO2 could allow frequent ventilator changes to optimize lung protective ventilation strategies. If these models are highly accurate, they can facilitate the development of closed-loop ventilator systems. We sought to develop and test algorithms for estimating pH and PaCO2 from measures of ventilator support, pulse oximetry, and end tidal carbon dioxide (ETCO2). We also sought to determine whether surrogates for changes in dead space can improve prediction.

Methods: Algorithms were developed and tested using 2 datasets from previous published investigations. A baseline model estimated pH and PaCO2 from ETCO2 using the previously observed relationship between ETCO2 and PaCO2 or pH (using Henderson-Hasselbalch equation). We developed a multivariate Gaussian processes (MGP) model incorporating other available non-invasive measurements.

Results: The training dataset had 2,386 observations from 274 children, and the testing dataset 658 observations from 83 children. The baseline model predicted PaCO2 within +/- 7 mmHg of observed PaCO2 80% of the time. The MGP model improved this to +/- 6 mmHg. When the MGP model predicted PaCO2 between 35-60 mmHg, the 80% prediction interval narrowed to +/- 5 mmHg. For pH, the baseline model predicted pH within +/- 0.07 of observed pH 80% of the time. The MGP model improved this to +/- 0.05.

Conclusions: We have demonstrated a conceptual first step for predictive models that estimate pH and PaCO2 to facilitate clinical decision making for children with lung injury. These models may have some applicability when incorporated in ventilator protocols to encourage practitioners to maintain permissive hypercapnia when using high ventilator support. Refinement with additional data may improve model accuracy.

Download Information

Robinder G. Khemani, E. Busra Celikkaya, Christian R. Shelton, Dave Kale, Patrick A. Ross, Randall C. Wetzel, and Christopher J. L. Newth (2014). "Algorithms to estimate PaCO2 and pH using non-invasive parameters for children with Hypoxemic Respiratory Failure." Respiratory Care, 59(8), 1248-1257.       ext

Bibtex citation

@article{Kheetal14,
     author = "Robinder G. Khemani and E. Busra Celikkaya and Christian R. Shelton and Dave Kale and Patrick A. Ross and Randall C. Wetzel and Christopher J. L. Newth",
     title = "Algorithms to estimate PaCO2 and pH using non-invasive parameters for children with Hypoxemic Respiratory Failure",
     journal = "Respiratory Care",
     year = 2014,
   month = aug,
     volume = 59,
     number = 8,
     pages = "1248--1257",
}

full list

More Information

Address

University of California, Riverside
Chung Hall, room 368
Riverside, CA 92521

 


External Links