Bioreactance versus PICCOTD/PC in critically ill septic shock patients

Introduction

We designed this study to compare the performance in cardiac output (CO) monitoring capabilities of two devices in refractory septic shock patients (RSS Pts): non-invasive transthoracic bioreactance (NICOM) and pulse contour analysis (PICCO PC) coupled to transpulmonary thermodilution (PICCOTD).

Methods

We included RSS Pts in vasopressor/inotrope need monitored with both devices. Triplicate measurements of CO by PICCOTD were used to measure CO at baseline and to calibrate PulseCO. The CO values recorded simultaneously minute-by-minute by the two systems were compared at baseline (nonperturbated system), in response to a passive leg-raising maneuver (PLR = leg elevation to 45° for 2 minutes starting from a supine position) and PEEP test (10 and 15 cmH₂O for 10 minutes each) (perturbated system). We used PICCOTD/PC as the reference technology evaluating the accuracy and estimating the precision of both devices.

Results

Continuous CO recording with both devices was performed on 12 consecutive RSS Pts (Nep + Epi = 0.66 ± 0.15 μg/kg/minute, all ventilated with TV <8 ml/kg). At baseline (nonperturbated system), correlation analysis of NICOM vs PiCCOTD CO showed r² of 0.78 (P < 0.001). Bland-Altman analysis evidenced a mean bias of 0.08 l/min (LOA -1.31 to 1.49). The mean CO was 6.01 ± 1.48 l/min. In a perturbated system the bias of NICOM vs PICCO PC was respectively-0.05 l/min (LOA -1.52 to 1.42) and 0.3 l/min (LOA -2 to 2.6) during PLR and PEEP test. The percentage error was <30% in 92% of patients at baseline (nonperturbated system), in 92% of patients during PLR and in 74% during PEEP test. In a nonperturbated system the CO precision (calculated as 2 SD/mean over 10 consecutive measurements) was 6.5 ± 6% and 6.7 ± 9% for NICOM and PiCCOTD, respectively (NS). Precision for NICOM and PICCOPC was respectively 6.8 ± 13% and 4.7 ± 10% during PLR and 7 ± 15% and 7.6 ± 15% during PEEP test.

Conclusions

Although limited to a small number of patients, NICOM and PICCO PC, calibrated by TD, seem to have comparable accuracy and precision in CO monitoring in RSS Pts even in a perturbated system.