Preoperative Identification of Patient-Dependent Blood Pressure Targets Associated With Low Risk of Intraoperative Hypotension During Noncardiac Surgery
Models can be created for various patient populations, using preoperative variables, that can predict intraoperative blood pressure ranges with low risk for subsequent hypotension (i.e. MAP < 65 mmHg).Â
Methods
Single-center
MAP data
> 160,000 surgeries over 9 years
Patients 18 years and older
7.3 million MAP measurements via arterial lines
Urological and gynecological surgeries excluded
Primary Outcome
Hypotension risk
The proportion of instances that a given MAP value (65 mmHg - 100 mmHg) was followed by a MAP value < 65 mmHg in the next 5 minutes
Training set
> 121,000 surgeries with 5.4 million MAP measurements from 2011 - 2018
Validation set
> 45,000 surgeries with 1.9 million MAP measurement from 2018 - 2020
Results
Selected model found to be predictive for 36/80 possible patient subgroups analyzed
Preoperative variables used in model (predictive of intraoperative hypotension risk)
Age
ASA PS
Gender
Common intraoperative MAP values (ex: 80 mmHg) found to be high-risk for some subgroups but not for others
Risk of subsequent hypotension increases exponentially as MAP approaches 65 mmHg
Risk of subsequent hypotension increases both when BP has been increasing or decreasing
Strengths
Large # of subjects
Generalizability
Multiple models evaluated
Best model can be used to evaluate other areas of interest
Novel concept
Weaknesses
Single-center
Only 36 of 80 possible subgroups had models validated
Some types of cases (gynecological and urological) not included
Only general anesthesia cases included
Some potentially contributory data not included in models (i.e. procedure type, pressor use)
Retrospective
Key Point: Preoperative variables can be used to determine intraoperative blood pressure ranges that are low-risk for subsequent hypotension.