A Bivariate Autoregressive Integrated Moving Average Analysis of Combat Troop Casualty Rates

Abstract

Mathematical modelling of medical resource requirements during military operations requires analyzing the underlying relationships between Disease and Non-Battle Injury (DNBI) rates and Wounded-In-Action (WIA) rates. DNBI and WIA data were extracted from Marine Corps unit diaries for a 150-day period of the Korean War and from a go day period of the Okinawa operation during World War 2 and a 123 day period of the war in Vietnam. The time series data were set up in a bivariate autoregressive integrated moving average analysis. All the univariate models are best represented with a moving average term. This study has shown that WIA rates can be a useful predictor of DNBI rates when using a bivariate ARIMA model. High levels of WIA incidence will affect DNBI rates the immediate day and the following day. These results were consistent throughout the three military conflicts examined and should prove indicative for future military campaigns. ARIMA, Autocorrelation, Medical resources, Arrival rates, Time series.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1994
Accession Number
ADA278290

Entities

People

  • Christopher G. Blood
  • Edward R. O'donnell

Organizations

  • Naval Health Research Center

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Autocorrelation
  • Battles
  • Casualties
  • Combat Operations
  • Data Science
  • Diseases And Disorders
  • Information Science
  • Military Medicine
  • Military Operations
  • Statistics
  • Time Intervals
  • Time Series Analysis
  • Transfer Functions
  • War
  • Warfare
  • White Noise
  • Wounds And Injuries

Readers

  • Approximation Theory.
  • Computational Modeling and Simulation
  • Trauma or Military Medicine