Summer Mortality Surveillance from Selected City and County Medical Examiners
Stephen D. VonAllmen Special Studies Branch Chronic Diseases Division Center for Environmental Health
Epidemiologic investigation of the 1980 heat-wave-related deaths in St. Louis and Kansas City, Missouri, revealed that the number of total deaths reported by medical examiners during heat-wave periods increased dramatically over expected numbers (1-4t). Although only a portion of those excess deaths were classified as heat-related (heat stroke, heat exhaustion, etc.), the increased numbers provided presumptive evidence of a heat-wave-induced public health problem. In 1981, CDC established the summer mortality surveillance project, based on total numbers of deaths reported by medical examiners in selected cities, to provide a simple means for early warning of excess, potentially preventable mortality associated with heat waves.
Participating medical examiners furnished baseline data consisting of tabulated total deaths, by day of report, for June, July, and August of 1979 and 1980 and for June of 1981. In order to obtain information regarding deaths in July and August 1981, CDC telephoned medical examiners' offices each day to obtain the previous day's total. (Each Monday, all deaths reported on the previous Friday, Saturday, and Sunday were recorded). Initially (mid-July 1981), nine cities (Kansas City, Memphis, Charlotte, Atlanta, Jacksonville, Dallas, Phoenix, Las Vegas, and Birmingham) had furnished baseline data and were being telephoned for prospective data. At that time, medical examiners in several cities were reporting current deaths but had not furnished baseline data. By early August, 13 cities were participating fully, and in early September, three additional cities were enrolled and supplied complete data sets for the summers of 1979, 1980, and 1981.
The Climate Analysis Center, National Weather Service, Camp Springs, Maryland, provided weather data twice a week by telecopier for the summer of 1981. We used these data, along with retrospective data for the summers of 1979 and 1980 furnished by the National Climatic Center, Asheville, North Carolina, in the succeeding analyses, tables, and figures. Statistical significance was defined at p less than 0.05 using one- or two-tailed t tests and Pearson correlation coefficients.
If changes in jurisdictional population bases, reporting procedures, and policies for autopsy selection are taken into account, only in July 1980 were there substantially larger numbers of deaths than in the comparable month for the two other years (Table A-1). For July 1979, the average increase in the normal daily temperature above normal for the 16 cities was 0.3OF (Table 1). For July 1980, an exceptionally hot month, the mean departure from normal was 4.0OF, significantly higher than the corresponding figure for 1979. During July 1981, all of the surveyed cities averaged higher-than-normal temperatures. The average daily temperature was 1.9OF above normal. July 1981 temperatures were significantly higher than those of July 1979, but they were significantly lower than 1980 temperatures.
In Kansas City, Memphis, Dallas, Atlanta, Oklahoma City, Phoenix, New York, and Philadelphia, all those cities with daily departures from the mean temperature that were greater in July 1980 than in July 1979, a larger number of medical-examiner-reported (MER) deaths were reported for July 1980 than for July 1979 (Table 2).* This observation was also true relative to July 1981, another non-heat-wave month. The first five of these cities had a heat wave in July 1980, with daily mean temperatures averaging more than 7.0F above the 30-year mean. In these five cities, MER deaths showed an average increase of 75% over July 1979 and 51% over July 1981. In Las Vegas, Phoenix, and Los Angeles, the heat wave had no apparent effect (Figures 1-3), while daily cumulative reports of MER deaths during the 1980 heat wave in St. Louis, Kansas City (Missouri), and Memphis showed a large heat-related effect (Figures 4-6).
The percentage increase in daily MER deaths in the heat-wave month of July 1980 over the relatively normal month of July 1979 correlated poorly with the average daily maximum temperature for July 1980. However, the percentage increase in July 1980 deaths over those in July 1979 did correlate well with the departure from normal July temperatures. In contrast with the extremely hot summer of 1980, the percentage increase in MER deaths seen during the somewhat warmer-than-normal month of July 1981 over July 1979 did not correlate well with the 1981 departure from normal July temperature. *Birmingham and St. Louis also had larger average daily departures from the mean temperature in July 1980 than in either July 1979 or July 1981, but changes in reporting practices prevented appropriate comparisons of mortality.
Data from the severe heat wave of July 1980 support the rationale for surveillance of MER deaths during hot weather. Substantial case-load increases occurred, correlating with the higher-than-normal temperature. Daily reports of total MER deaths early in a heat wave appear to provide a simple, timely, and sensitive index of the adverse effect of a severe heat wave on human health.
Data collected retrospectively for 1979 and 1980 verified the previous observation that increases in total mortality closely paralleled the sudden onset of heat-related deaths at the beginning and peak of the July 1980 heat wave in the affected cities. Similarly, the prospectively collected data from the summer of 1981 included sporadic and isolated reporting of heat-related deaths but showed no significant concurrent increases in total mortality. Thus, such surveillance appears to be sensitive to substantial numbers of heat-related deaths.
In general, the summer of 1981 was hotter than normal in the cities in which surveillance was maintained, but not nearly as hot as the severe, heat-wave summer of 1980. Although upward departure from normal temperature was strongly associated with the increase in total MER deaths for 1980, the milder above-normal temperatures in 1981 were not associated with such an increase. At no point during the summer of 1981 was a very substantial increase in the number of MER deaths seen; therefore, no early warning of a possible heat-related health hazard was required.
Correlating the occurrence of MER deaths with various weather conditions for purposes of surveillance has some limitations. First, as currently structured, the medical-examiner system is not specific and data are not uniform. On the other hand, weather data are highly structured and standardized and are based on years of data collected through a specialized network of reporting points. Such differences in quality of data could produce artifactual associations between weather parameters and health. Second, because there were no historical data for deriving normal values for MER death rates, it was necessary to assume in this study that the 1979 figures provided a reasonably normal baseline. The validity of this assumption requires further study. Third, no account was taken of the effect of humidity and wind velocity on biologic heat stress; however, analyses in which temperatures were corrected for humidity have yielded similar results (3t). Finally, an MER-based surveillance system is not sensitive to less severe increases in ambient temperature and cannot measure the full impact of a heat wave on morbidity and productivity.
Despite these limitations, the summer mortality surveillance project has proven to be a useful public health activity and was reinstated in those previously participating medical-examiner jurisdictions in 1982. Continual refinement of this system should assist public health officials in their efforts to implement timely and effective prevention measures in the event of severe heat stress in the community.
Disclaimer All MMWR HTML documents published before January 1993 are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.**Questions or messages regarding errors in formatting should be addressed to firstname.lastname@example.org.
Page converted: 08/05/98
This page last reviewed 5/2/01