[Effect of time elapsed on continuous pulse contour cardiac output measurement in septic shock patients].

OBJECTIVE: To evaluate the effect of time elapsed on continuous pulse contour cardiac output (PCCO) measurement in septic shock patients.

METHODS: Data during February 2011 to February 2013 from 25 septic shock patients equipped with a pulse indicator continuous cardiac output (PiCCO) device in Department of Critical Care Medicine of Guangzhou Medical University were retrospectively analyzed. PCCO was recorded immediately before transpulmonary thermodilution (COTPTD) calibration. After divided by ideal body surface area, cardiac index (CI) was calculated, and PCCI/CITPTD pairs were analyzed. Four subsets of CI pairs were defined according to intervals of time elapsed from the previous calibration [within the first hour (including 1 hour), between 1 and 8 hours (including 8 hours), between 8 and 16 hours (including 16 hours), and more than 16 hours]. Linear regression, the threshold value of concordance (as indicated by bias ± 2SD) and percentage error (2SD/the mean of CITPTD) were used to compare agreement between PCCI and CITPTD.

RESULTS: A total of 162 data pairs from 25 patients were analyzed. For all data pairs, PCCI correlated significantly with CITPTD (r² = 0.494, P<0.001), the bias±2SD was -(0.06 ± 1.41) L × min⁻¹ × m⁻² and the percentage error was 37%. Among the four time-interval subsets, the percentage error was <30% only in subset between 1 and 8 hours, and the percentage error in other subsets was over 30%. Linear regression analysis between ΔPCCI and ΔCITPTD showed a r² of 0.217 (P<0.001) for the whole 162 data pairs. A r² of 0.327 (P<0.001) and a r² of 0.303 (P<0.001) were calculated for the subset of between 1 and 8 hours and between 8 and 16 hours respectively.

CONCLUSIONS: Our study in septic shock patients suggests that the accuracy of PCCO will be decreased as the increase of the time interval for calibration. Transpulmonary thermodilution calibration should be performed again if hemodynamic changes or was inconsistent with the clinical presentation. It is suggested that re-calibration should be done within 8 hours.