Autoregressive With Exogenous Input (ARX) Decision Support for Blood Pressure Maintenance During Cesarean Delivery...

Area of Science:

• Anesthesiology
• Maternal-Fetal Medicine
• Health Informatics

Background:

• Spinal anesthesia for cesarean delivery frequently causes maternal hypotension.
• Hypotension can compromise uteroplacental perfusion and maternal comfort.
• Current vasopressor management is often reactive, despite guidelines recommending proactive strategies.

Purpose of the Study:

• To evaluate an autoregressive with exogenous input (ARX) decision-support algorithm for real-time maternal mean arterial pressure (MAP) forecasting.
• To assess the algorithm's ability to support vasopressor management during cesarean delivery under spinal anesthesia.
• To determine if ARX-guided care impacts hypotension duration, vasopressor use, and maternal/neonatal outcomes.

Main Methods:

• A prospective pilot study involving 20 pregnant patients undergoing cesarean delivery under spinal anesthesia.
• Patients received standard care supplemented with ARX-generated MAP predictions at 1-, 2-, and 3-minute horizons.
• ARX-guided patients were matched 1:2 to nonconcurrent controls (n=40) for comparison.

Main Results:

• One-minute-ahead ARX predictions showed accuracy (RMSE 3.71 mmHg, MAE 2.75 mmHg).
• ARX-guided patients experienced shorter hypotension duration (0.8 vs 3.0 min; P=.005) and lower incidence (25.0% vs 52.5%; P=.033).
• ARX-guided care was associated with higher phenylephrine dose (1823 vs 974 µg; P=.001) and lower maternal nausea (5% vs 35%; P=.014).

Conclusions:

• An ARX decision-support algorithm demonstrated accurate real-time MAP forecasting during spinal anesthesia for cesarean delivery.
• ARX-guided management was associated with reduced hypotension duration and incidence.
• These preliminary findings support further investigation in larger randomized trials to confirm clinical benefits.