Poisson Modeling Predicts Acute Telestroke Patient Call Volume

Area of Science:

• Neurology
• Health Services Research
• Biostatistics

Background:

• Effective staffing is crucial for managing time-sensitive strokes.
• Predicting telestroke and teleneurology consultation frequency is essential for resource allocation.

Purpose of the Study:

• To evaluate the suitability of Poisson distribution for modeling hourly telestroke call volumes.
• To develop a predictive model for telestroke call frequency over a 24-hour period.

Main Methods:

• Retrospective cohort review of 8,499 telestroke activations (January 2019-December 2022).
• Analysis of call frequency using Poisson probability mass function (PMF) and cumulative distribution function (CDF).
• Application of univariable Poisson regression to predict future call volumes.

Main Results:

• The Poisson distribution demonstrated a close fit to the observed telestroke call data.
• Mean calls/day ranged from 0.07 (5-6 a.m.) to 0.45 (7-8 p.m.).
• Predicted probability of two or fewer calls/day ranged from 98.9% to 99.9%; future calls projected to increase.

Conclusions:

• Poisson modeling is a robust method for predicting telestroke call volumes.
• This predictive capability can inform staffing decisions and physician allocation in real-time.
• The model is adaptable to any healthcare system with established mean call volume data.