Predicting intra-abdominal hypertension using anthropometric measurements and machine learning.

Area of Science:

• Critical Care Medicine
• Medical Measurement
• Patient Monitoring

Background:

• Intra-abdominal hypertension (IAH) affects nearly 25% of critically ill patients.
• Accurate intra-abdominal pressure (IAP) measurement is crucial for early IAH detection and complication avoidance.
• Current IAP measurement relies on invasive bladder techniques.

Purpose of the Study:

• To investigate the relationship between anthropometric parameters and IAP.
• To determine if IAP and IAH can be predicted non-invasively using body measurements.
• To identify reliable anthropometric predictors for IAP and IAH.

Main Methods:

• Retrospective observational study of 96 critically ill patients.
• Recorded IAP and 28 body parameters.
• Utilized statistical analyses (Pearson's correlation, mutual information) and simulation modeling.

Main Results:

• Three anthropometric parameters (difference between convex and horizontal xiphoid-pubis distance, sagittal abdominal diameter, abdominal compliance) predicted intra-bladder pressure (IBP) with R²=0.56.
• The difference between convex and horizontal xiphoid-pubis distance and sagittal abdominal diameter reliably predicted IAH (R²=0.79).

Conclusions:

• Specific anthropometric measurements can reliably predict IAP and IAH non-invasively.
• The difference between convex and horizontal xiphoid-pubis distance is a key predictor for IAH.
• Further research with larger, diverse populations is needed to validate these findings.