Mortality in Portugal associated with the heat wave of August 2003: Early estimation of effect, using a rapid method

During the first two weeks of August 2003, Portugal was affected by a severe heat wave. Following the identification in Portugal of the influence of heat waves on mortality in 1981 and 1991 (estimated excess of about 1900 and 1000 deaths respectively), the Observatório Nacional de Saúde (ONSA) - Instituto Nacional de Saúde Dr. Ricardo Jorge, together with the Vigilância Previsão e Informação - Instituto de Meteorologia, created a surveillance system called ÍCARO, which has been in operation since 1999. ÍCARO identifies heat waves with potential influence on mortality [1]. Before the end of the 2003 heat wave, ONSA had produced a preliminary estimate of its effect on mortality. The results based on daily number of deaths from 1 June to 12 August 2003 were presented within 4 working days. Data was gathered from 31 National Civil registrars, covering the district capitals of all 18 districts of mainland Portugal, and representing approximately 40% of the mainland’s mortality. The number of deaths registered in the period 30 July to 12 August was compared with the ones registered during 3 comparison periods: (in July): 1-14 July, 1-28 July, and 15-28 July). 15-28 July, the period best resembling the heat wave in time and characteristics, produced an estimation of 37.7% higher mortality rate then the value expected under normal temperature conditions. From this value, an estimate of 1316 death excess was obtained for mainland Portugal. The main purpose of this article is to present the method used to identify and assess the occurrence of an effect (excess mortality) during the heat wave of summer 2003.