Abstract
Background
Stroke volume measurement should provide estimates of acute treatment responses. The current pulse contour method estimates left ventricle stroke volume. Heart-lung interactions change right ventricular stroke volume acutely. We investigated the accuracy, precision, and trending abilities of four calibrated stroke volume estimates based on pulmonary artery pulse contour analysis.
Results
Stroke volume was measured in 9 pigs with a pulmonary artery ultrasound flow probe at 5 and 10 cmH2O of PEEP and three volume states (baseline, bleeding, and retransfusion) and compared against stroke volume estimates of four calibrated pulmonary pulse contour algorithms based on pulse pressure or pressure integration. Bland-Altman comparison with correction for multiple measurements and trend analysis were performed. Heart rate and stroke volumes were 104 ± 24 bpm and 30 ± 12 mL, respectively. The stroke volume estimates had a minimal bias: − 0.11 mL (95% CI − 0.55 to 0.33) to 0.32 mL (95% CI − 0.06 to 0.70). The limits of agreement were − 8.0 to 7.8 mL for calibrated pulse pressure to − 10.4 to 11.5 mL for time corrected pressure integration, resulting in a percentage error of 36 to 37%. The calibrated pulse pressure method performed best. Changes in stroke volume were trended very well (concordance rates 73–100%, r2 0.26 to 0.987, for pulse pressure methods and 71–100%, r2 0.236 to 0.977, for integration methods).
Conclusions
Pulmonary artery pulse contour methods reliably detect acute changes in stroke volume with good accuracy and moderate precision and accurately trend short-term changes in cardiac output over time.
Comparison of non-invasive (ClearSight®) and invasive (Pulsioflex®) pulse contour analysis to measure stroke volume during major surgery
