Forecasting Wounded-in-Action Casualty Rates for Ground Combat Operations

Abstract

(U//FOUO) The methodology to calculate wounded-in-action (WIA) casualty rates for ground combat operations was updated to implement recommendations from the Defense Health Agency Casualty Rate Development Work Group. The new model will be implemented in the Casualty Rate Estimation Tool (CREstT), which resides within the Department-of-Defense-accredited Medical Planners Toolkit. Data were collected from various sources for major combat and contingency operations, including World War II, Korea, Vietnam, Iraq, and Afghanistan. Building upon Trevor Dupuys extensive research on casualty rate estimation, we explored the effects of factors such as duration, population at risk or force size, terrain, climate, enemy capability, and tactical operation on WIA rates. We developed several regression and ensemble models, and selected the Ridge Regression model to estimate ground combat WIA rates. The Ridge Regression model outperformed the current model in CREstT (where a baseline-derived WIA rate is modified using adjustment factors) as well as Dupuy's model of estimating casualties. This work will refine WIA casualty projections in CREstT and improve resource estimates for medical planners.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 28, 2019
Accession Number
AD1105971

Entities

People

  • Andrew Olson
  • Edwin D'souza
  • James Zouris
  • Trevor Elkins
  • Vern Wing

Organizations

  • Leidos
  • Naval Health Research Center

Tags

DTIC Thesaurus Topics

  • Afghanistan Conflict
  • Amphibious Operations
  • Combat Casualty Care
  • Combat Injuries
  • Combat Operations
  • Contingency Operations (Military)
  • Data Analysis
  • Department Of Defense
  • Failed States
  • Information Science
  • Iraqi-War
  • Korean War
  • Machine Learning
  • Marine Corps
  • Second World War
  • Supervised Machine Learning
  • Test Sets
  • Training
  • United States
  • Warfare

Readers

  • Computational Modeling and Simulation
  • Regression Analysis.
  • Trauma or Military Medicine