Derivation of a formula to predict patient volume based on temperature at college football games

Prehosp Emerg Care. 2007 Oct-Dec;11(4):453-7. doi: 10.1080/00207450701537043.


Objective: We sought to explore the relationship between temperature and spectator illness at Division I college football games by deriving a formula to predict the number of patrons seeking medical care based on the ambient temperature and attendance of the game.

Methods: A retrospective review was conducted of medical records from 47 Division I college football games at two outdoor stadiums from 2001 through 2005. Any person presenting for medical care was counted as a patient seen. Weather data were collected from the National Weather Service. A binomial model was fit to the spectator illness records by using the patients seen per attendance as the outcome measure, with temperature as the predictor.

Results: Using a binomial model, a formula was derived to estimate the number of patients needing medical attention based on the temperature and the number of spectators in attendance. Predicted # of Patients = exp (-7.4383 - 0.24439* Temperature C + 0.0156032 * Temperature C(2) - 0.000229196 * Temperature(3)) * number of spectators; all factors were highly significant (p < 0.0001). The model suggests that as the temperature rises, the number of patients seeking medical attention will also increase. The formula shows that an increase in temperature from 20 to 21 degrees C will show an increase in patient encounters from 3.64 to 4.05 visits per 10,000 in attendance (an 11% increase).

Conclusion: These results show that temperature is an important variable to consider when determining the medical resources needed in caring for spectators at outdoor football games. Our model may help providers predict the number of spectators presenting for medical care based on the forecasted temperature and predicted attendance.

MeSH terms

  • Emergency Medical Services / statistics & numerical data*
  • Football*
  • Forecasting
  • Humans
  • Medical Records
  • Models, Statistical
  • North Carolina
  • Ohio
  • Retrospective Studies
  • Temperature*