Accuracy of Quantifying Hypotension During Surgery Using Physiological Sensor Data

Area of Science:

• Anesthesiology and Critical Care Medicine
• Biomedical Engineering
• Surgical Monitoring

Background:

• Arterial blood pressure monitoring is vital during surgery to detect hypotension, a condition linked to adverse post-surgical outcomes.
• Quantifying hypotension events (number, duration, area under threshold) aids in risk assessment.
• Data simplification through downsampling (e.g., to 15-second intervals) is common for manageability but may introduce inaccuracies.

Purpose of the Study:

• To quantify the errors introduced by downsampling arterial blood pressure data in the assessment of surgical hypotension.
• To evaluate the effectiveness of error mitigation techniques in reducing these downsampling-induced inaccuracies.
• To compare the accuracy of downsampled data analysis with continuous blood pressure waveform data.

Main Methods:

• Analysis of arterial blood pressure data from 419 open abdomen surgery patients.
• Calculation of hypotension quantification errors at various downsampling rates, focusing on 15-second intervals.
• Testing of error mitigation strategies and comparison with continuous waveform data.

Main Results:

• Downsampling to 15-second intervals resulted in significant relative biases: 27.0% for hypotensive event count, 30.0% for cumulative duration, and 10.8% for cumulative area under threshold.
• Error mitigation techniques generally did not improve the accuracy of hypotension quantification.
• Continuous blood pressure waveform data provides a more accurate basis for hypotension assessment.

Conclusions:

• Significant errors arise from downsampling arterial blood pressure data for hypotension quantification.
• Awareness of these quantitation errors is essential when interpreting historical studies relying on downsampled data.
• High-frequency arterial blood pressure data is necessary for precise surgical hypotension monitoring and risk assessment.