Background and Goal of Study

Obtaining accurate continuous blood pressure monitoring is mandatory in critically-ill, anesthetized patients or during long and challenging surgical procedures. The gold standard for continuous BP measurement- invasive arterial catheter- is complex and associated with morbidity. 
This study presents a simple way to measure non-invasive BP continuously without additional equipment in the operating room. To allow non-invasive beat-to-beat BP measurement, we used a commercially available pulse oximeter probe and analyzed its signals in order to assess performance of optically acquired signals transformed into continuous BP values with an algorithm approach.
The aim of the study is to prove the accuracy of the optical non-invasive blood pressure device compared to the invasive arterial measurement. 

Materials and Methods

We included and obtained informed consent from 40 patients aged more than 18 years and scheduled for an elective surgery necessitating general anesthesia and invasive BP monitoring at CHUV- University Hospital of Lausanne. An arterial catheter was inserted in the radial artery.
The fingertip device (Philips Healthcare) was placed on the contralateral side. Both signals were recorded at induction of general anesthesia and compared for 10 minutes.

The signals recorded by the fingertip device were analyzed based on published pulse wave analysis algorithms. So-called Optical Blood Pressure Monitoring (OBPM) technology, these algorithms allow to continuously estimate BP from the raw optical signals acquired by pulse oximeters. 
 
The OBPM algorithms were tested under two calibration scenarios. The first one (CAL 1), simulates the clinical setting where the OBPM algorithms are used to continuously estimate BP in between two standard sphygmomanometer measurements. For this study a 3 minutes interval between two sphygmomanometer calibrations is simulated. 
 
The second scenario (CAL 2) simulates the clinical setting where the OBPM algorithms are used to continuously estimate BP of a patient after a unique initial calibration of seconds. 

Results

We provide here results from the 40 patients for which OBPM continuously estimated mean arterial BP (MAP) and diastolic arterial BP (DIA). Data analysis on the systolic BP is underway. 
 
The measured mean error for DIA running CAL 1 scenario is 0.44. mmHg, and the measured standard deviation of the error for DIA is 7.6 mmHg. For the CAL 2 scenario, the mean error is -3.1 mmHg, and the measured standard deviation is 12.4 mmHg.
For MAP, the mean error is 0 mmHg and a standard deviation error of 10 mmHg with the CAL1 scenario.
The measured mean error for MAP using CAL 2 scenario is 2.3 mmHg, and the measured standard deviation of 14.2 mmHg. 
The Pearson correlation coefficient for CAL1 is 0.83 for DIA and 0.87 for MAP and respectively 0.66 and 0.75 in CAL 2 setting. 
This study shows a good correlation between the measured optical BP and the invasive BP in both calibration scenarios. As awaited a 3-minutes interval calibration shows a stronger correlation than a unique calibration at the beginning.
 
For such a simple optical measurement setup, these performances are already very close to the ISO81060-2 requirements, in particular when using OBPM to continuously estimate BP in between two sphygmomanometer measurements.

Conclusion

Our results demonstrate feasibility of continuously estimating BP changes in the operating room via the analysis of pulse oximeter optical signals by the OBPM library of algorithms. Further studies are needed but this research validates a promising track towards non-invasive continuous monitoring of BP in the perioperative setting.


08 / Continuous optical blood pressure measurement in the perioperative setting