Changes in vital signs under general anesthesia
Conditions
Sponsors
West China Hospital of Sichuan University
Eligibility
Sex/Gender
All
Age
18 Years to 65 Years
Inclusion criteria
Inclusion criteria: This study presents the development and external validation of vital signs prediction models, constructed using the publicly available Korean VitalDB database and externally validated at our center. Model Development Cohort Study subjects for model development were derived from the VitalDB database. The inclusion criteria were as follows: (a) underwent general anesthesia with mechanical ventilation; (b) had relatively complete intraoperative monitoring data for bispectral index (BIS), heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP); and (c) had infusion records for major anesthetic drugs, including intravenous anesthetics and opioids. External Validation Cohort Study subjects for external validation were obtained from West China Hospital, Sichuan University. The inclusion criteria were as follows: (a) patients who underwent elective surgery under general anesthesia with mechanical ventilation; and (b) relatively complete intraoperative continuous monitoring data, enabling acquisition of major vital signs including BIS, HR, SBP, DBP and MAP, along with records of major anesthetic drug administration and related perioperative clinical data.
Exclusion criteria
Exclusion criteria: The exclusion criteria for the model development cohort derived from the VitalDB database were as follows: (a) severe missingness in key monitoring tracks, precluding the formation of complete input sequences; (b) prolonged interruptions in primary target variables, rendering them unsuitable for label construction in future time-point prediction tasks; and (c) obviously abnormal raw data, such as monitoring values persistently exceeding physiologically plausible ranges, disordered timestamps, or logically inconsistent variable recordings. The exclusion criteria for the external validation cohort obtained from West China Hospital, Sichuan University were as follows: (a) significant interference in intraoperative key vital signs signals, resulting in distorted monitoring values or inability to continuously reflect physiological status; (b) severe missingness in primary input variables, precluding the construction of unified time series; and (c) incomplete or misaligned temporal information across different data modalities.
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Model predicts future 3-minute BIS performance; | — |
Secondary
| Measure | Time frame |
|---|---|
| Model predicts future 3-minute HR performance;Model predicts future 3-minute HR performance;Accuracy of the model in predicting adverse events related to vital signs;Ability to predict adverse events in vital signs; | — |
Countries
China
Contacts
Public ContactWensheng Zhang
West China Hospital of Sichuan University
Outcome results
None listed