Contrasting patterns of mortality and hospital admissions during hot weather and heat waves in Greater London, UK

Background: Epidemiological research has shown that mortality increases during hot weather and heat waves, but little is known about the effect on non-fatal outcomes in the UK. Aims and Methods: The effects of hot weather and heat waves on emergency hospital admissions were investigated in Greater London, UK, for a range of causes and age groups. Time series analyses were conducted of daily emergency hospital admissions, 1 April 1994 to 31 March 2000, using autoregressive Poisson models with adjustment for long term trend, season, day of week, public holidays, the Christmas period, influenza, relative humidity, air pollution (ozone, PM10), and overdispersion. The effects of heat were modelled using the average of the daily mean temperature over the index and previous two days. Results: There was no clear evidence of a relation between total emergency hospital admissions and high ambient temperatures, although there was evidence for heat related increases in emergency admissions for respiratory and renal disease, in children under 5, and for respiratory disease in the 75+ age group. During the heat wave of 29 July to 3 August 1995, hospital admissions showed a small non-significant increase: 2.6% (95% CI -2.2 to 7.6), while daily mortality rose by 10.8% (95% CI 2.8 to 19.3) after adjusting for time varying confounders. Conclusions: The impact of hot weather on mortality is not paralleled by similar magnitude increases in hospital admissions in the UK, which supports the hypothesis that many heat related deaths occur in people before they come to medical attention. This has evident implications for public health, and merits further enquiry.