Corrected end-tidal P(CO(2)) accurately estimates Pa(CO(2)) at rest and during exercise in morbidly obese adults

Abstract

Background: Obesity affects lung function and gas exchange and imposes mechanical ventilatory limitations during exercise that could disrupt the predictability of Pa(CO(2)) from end-tidal P(CO(2)) (P(ETCO(2))), an important clinical tool for assessing gas exchange efficiency during exercise testing. Pa(CO(2)) has been estimated during exercise with good accuracy in normal-weight individuals by using a correction equation developed by Jones and colleagues (P(JCO(2)) = 5.5 + 0.9 x P(ETCO(2)) – 2.1 x tidal volume). The purpose of this project was to determine the accuracy of Pa(CO(2)) estimations from P(ETCO(2)) and P(JCO(2)) values at rest and at submaximal and peak exercise in morbidly obese adults.

Methods: Pa(CO(2)) and P(ETCO(2)) values from 37 obese adults (22 women, 15 men; age, 39 ± 9 y; BMI, 49 ± 7; [mean ± SD]) were evaluated. Subjects underwent ramped cardiopulmonary exercise testing to volitional exhaustion. P(ETCO(2)) was determined from expired gases simultaneously with temperature-corrected arterial blood gases (radial arterial catheter) at rest, every minute during exercise, and at peak exercise. Data were analyzed using paired t tests.

Results: P(ETCO(2)) was not significantly different from Pa(CO(2)) at rest (P(ETCO(2)) = 37 ± 3 mm Hg vs Pa(CO(2)) = 38 ± 3 mm Hg, P = .14). However, during exercise, P(ETCO(2)) was significantly higher than Pa(CO(2)) (submaximal: 42 ± 4 vs 40 ± 3, P < .001; peak: 40 ± 4 vs 37 ± 4, P < .001, respectively). Jones’ equation successfully corrected P(ETCO(2)), such that P(JCO(2)) was not significantly different from Pa(CO(2)) (submax: P(JCO(2)) = 40 ± 3, P = .650; peak: 37 ± 4, P = .065).

Conclusion: P(JCO(2)) provides a better estimate of Pa(CO(2)) than P(ETCO(2)) during submaximal exercise and at peak exercise, whereas at rest both yield reasonable estimates in morbidly obese individuals. Clinicians and physiologists can obtain accurate estimations of Pa(CO(2)) in morbidly obese individuals by using P(JCO(2)).

Figure 1.

Linear regression (solid line) and 95% prediction intervals (dotted lines) between 320 pairs of PaCO₂ and P_JCO₂ at rest and during exercise. P_JCO₂ = Jones-corrected end-tidal PCO₂; SEE = SE of estimates.

Figure 2.

Bland-Altman plot of the average of PaCO₂ and P_JCO₂ vs their difference. N = 311, SEE = 1.7, bias = 0.2 mm Hg (solid line), 95% prediction limits (dashed lines). P(_J-a)CO₂ = difference between Jones-corrected and PaCO₂. See Figure 1 legend for expansion of abbreviations.

Figure 3.

Mean ± SD values of PaCO₂, P_ETCO₂, and P_JCO₂. P_ETCO₂ significantly overestimated PaCO₂ during exercise. Jones’ equation successfully corrected P_ETCO₂ such that P_JCO₂ was not significantly different from PaCO₂. *P < .001 between P_ETCO₂ and PaCO₂ as well as between P_ETCO₂ and P_JCO₂. P_ETCO₂ = partial pressure of end-tidal CO₂; $\dot{\mathrm{V}}$_E = minute ventilation. See Figure 1 legend for expansion of other abbreviation.

Figure 4.

Differences (mean ± SD) between PaCO₂ and P_ETCO₂ (P[_ET − a]CO₂), as well as PaCO₂ and PJCO₂ (P[_ET − a]CO₂) values. *P < .001 significant difference between P(_ET − a)CO₂ and zero as well as between P(_ET − a)CO₂ and P(_J − a)CO₂. See Figure 3 legend for expansion of abbreviations.

Figure 5.

Mean arterial, end-tidal, and Jones-corrected V_D/V_T ratios at rest, submaximal exercise, and peak exercise. *P < .001 significant difference between V_D/V_T(ET) vs V_D/V_T(art) and V_D/V_T(J). V_D/V_T(art) = dead space to tidal volume ratio using PaCO₂; V_D/V_T(ET) = dead space to tidal volume ratio using end-tidal P_etCO₂; V_D/V_T(J) = dead space to tidal volume ratio using P_JCO₂. See Figure 1 and 3 legends for expansion of other abbreviations.