Sidestream microstream end tidal carbon dioxide measurements and blood gas correlations in neonatal intensive care unit

Objective The study was designed to assess the use of newer sidestream microstream end tidal carbon dioxide (ETCO 2 ) device in predicting blood carbon dioxide (PCO 2 ) measurements in very low birth weight (VLBW = birth weight <1,500 g) and non‐VLBW NICU neonates. Study Design Sidestream microstream ETCO 2 detectors were allowed time to calibrate and reach steady state prior to blood gas measurements. Blood CO 2 (PCO 2 ) and simultaneous ETCO 2 were recorded. Ratio of dead space to tidal volume (VD/VT) was calculated using modified Bohr's equation. Correlation coefficient, estimates of difference, standard deviation, and 95% limits of agreement between ETCO 2 and PCO 2 concentrations were calculated. Results Two hundred eighty‐six paired samples were collected from 48 ventilated NICU patients. Average PCO 2 and ETCO 2 were 58.4 and 50.6 with a correlation of 0.76. Subgroup analysis showed a correlation of 0.73 in 204 paired blood from 34 VLBW infants and 0.82 in 82 paired samples from non‐VLBW infants. Estimates of difference ± standard deviation between PCO 2 and ETCO 2 concentrations in these three groups, respectively (ALL, VLBW, and non‐VLBW) were 7.84 ± 9.96, 8.2 ± 10.16, and 6.95 ± 9.45. The correlation coefficient significantly improved in the VLBW group to 0.86 with dead space to tidal volume ratio (VD/VT) <30% (0.86 vs. 0.42; P  < 0.001). Conclusion ETCO 2 measurements using sidestream microstream technology in VLBW demonstrated that the correlation of ETCO 2 and PCO 2 was moderate, but the agreement was less than adequate (bias > 5 mmHg in all groups). The results improved with lower VD/VT, suggesting that sidestream capnography is more reliable in conditions of less severe lung disease. Pediatr Pulmonol. 2013; 48:250–256. © 2012 Wiley Periodicals, Inc.