Limitations of stroke volume estimation by non-invasive blood pressure monitoring in hypergravity

Abstract

Background: Altitude and gravity changes during aeromedical evacuations induce exacerbated cardiovascular responses in unstable patients. Non-invasive cardiac output monitoring is difficult to perform in this environment with limited access to the patient. We evaluated the feasibility and accuracy of stroke volume estimation by finger photoplethysmography (SVp) in hypergravity.

Methods: Finger arterial blood pressure (ABP) waveforms were recorded continuously in ten healthy subjects before, during and after exposure to +Gz accelerations in a human centrifuge. The protocol consisted of a 2-min and 8-min exposure up to +4 Gz. SVp was computed from ABP using Liljestrand, systolic area, and Windkessel algorithms, and compared with reference values measured by echocardiography (SVe) before and after the centrifuge runs.

Results: The ABP signal could be used in 83.3% of cases. After calibration with echocardiography, SVp changes did not differ from SVe and values were linearly correlated (p<0.001). The three algorithms gave comparable SVp. Reproducibility between SVp and SVe was the best with the systolic area algorithm (limits of agreement -20.5 and +38.3 ml).

Conclusions: Non-invasive ABP photoplethysmographic monitoring is an interesting technique to estimate relative stroke volume changes in moderate and sustained hypergravity. This method may aid physicians for aeronautic patient monitoring.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Schematic representation of the protocol.

The subject, equipped with anti-G trousers, was exposed to +Gz accelerations during two centrifuge runs (total exposure time: 10 min). Arterial blood pressure was recorded by finger photoplethysmography continuously. Echocardiographic measurements were performed before the runs (E1, supine left lateral, E2, sitting with a 32° backrest angle in the centrifuge gondola), immediately after (E3, sitting in the gondola, E4, supine left lateral) and after 60 min of recovery (E5, supine left lateral). N = 10.

Fig 2. Time course of cardiovascular parameters measured by or computed from finger arterial blood pressure (ABP).

Parameters measured: SAP, systolic arterial pressure; DAP, diastolic arterial pressure; Parameters computed: HR, heart rate; SV, stroke volume; CO, cardiac output.

Fig 3. Stroke volume (SV) computed from echocardiographic and finger arterial blood pressure recordings (SVe and SVp respectively), at the time points described in Fig. 1.

Three algorithms were used for SVp computation: Liljestrand, the systolic area and Windkessel.

Fig 4. Bland-Altman plots for determination of agreement between SVe and SVp computed with the Liljestrand algorithm.

----: limits of agreement (maximum, minimum, mean).

Fig 5. Bland-Altman plots for determination of agreement between SVe and SVp computed with the systolic area method.

----: limits of agreement (maximum, minimum, mean).

Fig 6. Bland-Altman plots for determination of agreement between SVe and SVp computed with the Windkessel algorithm.

----: limits of agreement (maximum, minimum, mean).

Fig 7. Correlation plots between SVe and SVp computed with the Liljestrand algorithm.

----: equality line.

Fig 8. Correlation plots between SVe and SVp computed with the systolic area method.

----: equality line.

Fig 9. Correlation plots between SVe and SVp computed with the Windkessel algorithm.

----: equality line.