[Validation of the transcutaneous carbon dioxide tension measurements in critical paediatric patients]

Abstract

Objective: To estimate the accuracy of the transcutaneous carbon dioxide tension measurement (PtCO(2)) compared to the measurement of the arterial carbon dioxide tension (PaCO(2)).

Material and methods: An analytical, longitudinal, prospective and observational study, of a dynamic cohort taken from the in-patients of a Paediatric Intensive Care Unit (PICU). The PtCO(2) was measured with the SenTec AG analyzer, and the sensor was applied with the specific Multi-Site Attachment Ring. PtCO(2) and PaCO(2) were recorded at the same time. The statistical significance of the association between paired measurements was evaluated with the Snedecor's F test, the Pearson's r(2) correlation coefficient and the Interclass Correlation Coefficient (ICC). The degree of agreement was evaluated with the Bland & Altman method. The consistency of the results was evaluated with the ANalysis Of the VAriance (ANOVA).

Results: One hundred and six paired measurements, PtCO(2) and PaCO(2), from twelve patients, were compared. The means of the PaCO(2) and PtCO(2) were 51.0+/-13mmHg and 50.1+/-14mmHg, respectively; r(2)=0.87 (p<0.001), ICC=0.96, (95% CI: 0.94-0.97). The Bland-Altman analysis showed a mean difference of-0.9mmHg (95% CI:-2.0 to 0.2mmHg). The correlation was better in cases with no respiratory disease, with low respiratory assistance, with PaCO(2)>50mmHg and with the sensor applied on the forehead. The results were consistent. No side effects derived from the use of the ring were observed.

Conclusion: The correlation between PtCO(2) and PaCO(2) is excellent and stable. The ring sensor was safe and easy to use.