Share this article on:

Which Days of Hot Weather Are Considered Dangerous by Heat-Health Watch Warning Systems?: A Comparison of the Predictive Capacity of Different Systems

Hajat, Shakoor*†; Sheridan, Scott; Allen, Michael; Pascal, Mathilde§; Laaidi, Karine§; Tobias, Aurelio; Yagouti, Abderrahmane**; Beckis, Ugis**; Bourque, Denis††; Armstrong, Ben; Kosatsky, Tom*

doi: 10.1097/01.ede.0000362247.26824.dd
Abstracts: ISEE 21st Annual Conference, Dublin, Ireland, August 25–29, 2009: Oral Presentations

*British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada; †London School of Hygiene & Tropical Medicine, London, United Kingdom; ‡Kent State University, Kent, Ohio, United States; §Institut de Veille Sanitaire, Paris, France; ¶Instituto de Salud Carlos III, Madrid, Spain; **Health Canada, Ottawa, Ontario, Canada; and ††Environment Canada, Ottawa, Ontario, Canada.

Abstracts published in Epidemiology have been reviewed by the organizations of Epidemiology. Affliate Societies at whose meetings the abstracts have been accepted for presentation. These abstracts have not undergone review by the Editorial Board of Epidemiology.


Back to Top | Article Outline

Background and Objective:

Prompted by growing concerns about global warming, many countries worldwide have introduced Heat Health Warning Systems (HHWS) to minimize the public health impacts of exposure to hot weather. These systems issue alerts in response to forecasts of adverse weather conditions. Fundamentally different approaches are currently in place in the various HHWS systems around the world in terms of setting the thresholds of weather parameters which, when forecast to be breached, are expected to be associated with unacceptable levels of adverse health impacts. We compared the alternate approaches for setting HHWS thresholds as measured by how well they predicted heat-associated mortality in a common set of historical weather and mortality records.

Back to Top | Article Outline


Four different threshold-setting approaches which are currently in operation in HHWS were compared:

  • Synoptic classification into air-mass types.
  • Epidemiological analysis of retrospective data.
  • Physiologic approach based on heat-budget models.
  • Empiric set of temperature/humidity indices, e.g. Humidex.

Each approach was calibrated on four 10-year datasets of daily temperature and mortality counts (Chicago, London, Madrid, Montreal) in order to identify the weather conditions associated with adverse health impacts. These parameters were then applied to a further set of 10-year weather only data to provide a ranking of the top 50 most “heat adverse” days occurring in this second dataset as identified by each separate approach. The extent of overlap in the 50 days identified across the approaches was assessed; and temperature, observed mortality, and excess mortality occurring on the identified days was compared.

Back to Top | Article Outline


In each of the 4 cities, there was very little agreement in the most “heat adverse” days identified across the four approaches.

Back to Top | Article Outline


The identification of “heat adverse” days, and therefore the days on which an alert is called and protection measures initiated, is very dependant on the particular approach used to establish the thresholds.

© 2009 Lippincott Williams & Wilkins, Inc.