Deaths due to flash floods in Puerto Rico, January 1992: implications for prevention

Int J Epidemiol. 1994 Oct;23(5):968-75. doi: 10.1093/ije/23.5.968.

Abstract

Background: In January 1992, the Commonwealth of Puerto Rico sustained unusually heavy rainfall, which caused flash floods and deaths.

Methods: We conducted a descriptive study and a case-control study to determine the circumstances of these deaths and to identify mortality-prevention strategies. We describe the time, place, and circumstances of each death, and compare this information with water-level and rainfall data and the timing of warnings. Using controls selected from the affected population, we estimated the risk of death by age, sex, and vehicle occupancy during the flood.

Results: Within 7 hours, 23 people died in 17 incidents; 20 of the decedents (87%) were occupants of motor vehicles. The estimated risk of mortality was significantly elevated for motorists (odds ratio = 16, 95% confidence interval: 3.5-144). Being in a vehicle to evacuate a flash flood area was protective; however, being in a vehicle during the flood for other reasons further increased the risk of mortality. Deaths occurred early during the rapid rise of water and before official warnings were issued.

Conclusion: We recommend improving the sensitivity of the warning system and its ability to disseminate appropriate information rapidly. We also recommend educating officials and the public about the risks of driving on flooded roads and in potential flash flood conditions; and about the unique flash flood risks associated with specific topographical features in their region.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Automobile Driving
  • Case-Control Studies
  • Child
  • Child, Preschool
  • Disasters*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Mortality*
  • Puerto Rico / epidemiology