Research ArticlesExcess hospital admissions during the July 1995 heat wave in Chicago
Introduction
H igh summer temperatures can result in increased mortality among at-risk populations, such as elderly people and urban dwellers.1, 2, 3, 4, 5, 6 However, only a few studies have characterized nonfatal heat-related illnesses during heat waves. During the heat wave of 1980, hospital admissions in St. Louis and Kansas City were found to be increased by 5.1% and 1.5%, respectively.2 Also in 1980, in Memphis, there were 483 emergency department visits for heat-related illnesses. The most common diagnoses were heat exhaustion (58%), heat stroke (17%), and heat cramps (6%).7 In Adelaide, South Australia, potential medical risk factors identified for heat illnesses included diabetes mellitus, heart failure, and alcoholism.8
The health impact of heat depends not only on weather conditions at any given time, but also on previously existing health conditions and socioeconomic status.1, 9 The heat wave in Chicago in July 1995 provided an opportunity to study the nonfatal health effects of hot weather for the population served by Cook County hospitals. Previous studies have focused on heat-related mortality in Chicago.1, 10, 11 Our objectives were to retrospectively determine if there was an excess of hospital admissions as a result of the hot weather, ascertain the principal reason for hospitalization, and define the underlying medical conditions of the susceptible population. We compared the number of hospital admissions during the heat wave week with an average number for comparison periods without a heat wave.
A key contribution of this study is that we characterize the population at risk for heat-related morbidity that can be targeted with public health interventions. We propose prevention measures for individuals with specific underlying medical conditions with the goal of preventing heat-related morbidity in the future.
Section snippets
Methods
Chicago weather data for O’Hare airport, the official Chicago measuring station, was obtained from the Mid Western Climate Center. For every day in July 1995, we plotted the highest hourly heat index (apparent temperature), which is a function of temperature (in degrees Fahrenheit) and the relative humidity. The heat index measures the evaporative heat between a typical human and the environment and is a better measure of the effect of heat on the body than temperature alone.
Results
In July 1995, there were 42,304 people admitted to non-VA hospitals in Cook County (which includes the city of Chicago), compared with 40,910 in the month of July 1994. During the heat wave study period (July 13 through 19, 1995), the number of hospitalizations exceeded the average for non heat wave weeks by 1072 (p = 0.054) (Table 1, Figure 1), but no morbidity displacement was observed to supersede this excess. Although the increase for all diagnoses and age groups combined did not reach
Primary discharge diagnoses: principal reason for hospitalization
The primary discharge diagnosis, defined as the main disease process that resulted in hospital admission, was analyzed alone, without taking into account underlying conditions. Of the estimated 1072 excess admissions, 443 (p < 0.001) were heat-related diagnoses (Table 1, Figure 2 ), including heat stroke and heat exhaustion. An excess of 245 (p < 0.001) primary discharge diagnoses of fluid disorders and electrolyte and acid-base imbalance were found, of which volume depletion alone accounted
Primary and secondary discharge diagnoses: susceptible population
Comorbid conditions of those individuals at increased risk during the Chicago heat wave can be described by an evaluation of primary and secondary discharge diagnoses combined. Excess primary and secondary conditions define underlying conditions of the susceptible population. Although no statistically significant excess of hospitalizations with a primary diagnosis of cardiovascular diseases was seen during the heat wave period (see above and Table 1) 461 admissions (Table 2) over expected were
Discussion
Overall, during the Chicago heat wave study period in July 1995, we found 11% more inpatient hospital admissions in Cook County than expected. One quarter of the excess primary discharge diagnoses were specifically coded as dehydration, followed by heat stroke (22%) and heat exhaustion (15%). The susceptible population had specific underlying or comorbid medical conditions that increased their risk for dehydration; effective public health interventions need to be directed toward these
Acknowledgements
We thank Norm Staeling for his expert assistance. We also thank our colleagues for all the stimulating discussions, in particular Drs. Lisa Weasel, Gib Parrish, Tom Sinks, Carol Rubin, Richard Brennan, Ed Kilbourne, Loretta DiPietro, Nina Stachenfeld, Larry Kenney, Andrew Young, Kent Pandolf, and John Wilhelm.
References (34)
Some effects of the urban structure on heat mortality
Environ Res
(1972)Mortality from heat illness and heat aggravated illness in the United States
Environ Res
(1972)Patterns of urban heat wave deaths and implications for preventiondata from New York and St Louis during July, 1966
Environ Res
(1972)- et al.
Increased platelet and red cell counts, blood viscosity, and plasma cholesterol levels during heat stress, and mortality from coronary and cerebral thrombosis
Am J Med
(1986) - et al.
Heat related death during the July 1995 heat wave in Chicago
N Engl J Med
(1996) - et al.
Morbidity and mortality associated with the July 1980 heat wave in St. Louis and Kansas City, MO
JAMA
(1982) - et al.
Geographic distribution of heat related deaths among elderly persons
JAMA
(1989) - et al.
Ambient temperature and mortality from unintentional cocaine overdose JAMA
(1998) - et al.
Analysis of the 1980 heat wave in Memphis
J Am Geriatr Soc
(1981) - et al.
The effect in the heatheat related hospital presentations during a ten day heat wave
Aust NZ J Med
(1995)