Environmental-temperature injury in a Canadian Metropolis.

Background

Epidemiology and Impact of Environmental-Temperature Injuries

Environmental injury due to ambient temperature is commonly under-appreciated by health care workers and public health practitioners alike (Centers for Disease Control, 1985). Mortality data from the U.S. National Center for Health have revealed that during the 13-year period from 1968 to 1980, 6,460 deaths were attributed to the effects of cold (Centers for Disease Control, 1985). In the following 13 years, from 1980 to 1992, U.S. deaths from hypothermia increased to 10,550 (Centers for Disease Control and Prevention [CDC], 1996). The effects of heat waves on morbidity and mortality were recently investigated in two major U.S. cities. In July 1995, Chicago sustained a heat wave that resulted in more titan 600 excess deaths, 3,300 excess emergency department visits, and a substantial number of intensive-care-unit admissions for near-fatal heat stroke (Dematte et al., 1998). A smaller heat wave struck Philadelphia in 1993, with proportionately fewer deaths and hospitalizations (Mirchandani, McDonald, Hood, & Fonsea, 1996). To date, little is known about environmental-temperature-injury morbidity and mortality in Canada.

Range of Environmental-Temperature Injuries

Health effects of extreme temperature are not restricted to hypothermia and heat stroke, but include a myriad of disorders. Cold injury can result in frostnip, frostbite, immersion foot, chilblains, and hypothermia. Heat exposure can lead to heat cramps, heat fatigue, heat edema, heat exhaustion, heat syncope, and heat stroke. Although hypothermia and heat stroke are commonly thought of as the most life-threatening of temperature injuries, the morbidity of conditions such as immersion foot and frostbite should not be overlooked. The sidebar on page 41 details diagnoses of environmental-temperature injuries according to the International Classification of Disease coding system (ICD-9) for effects of heat and reduced temperature (ICD-9 codes 991 and 992) (U.S. Department of Health and Human Services, 1990).

Determinants of Environmental-Temperature Injury

Several risk factors are postulated in temperature injury. The risk of death by hypothermia varies by both age and sex (Centers for Disease Control, 1985). From 1979 through 1992, approximately half of all hypothermia deaths in the United States occurred among persons aged [greater than or equal to] 65 years (CDC, 1996). The annual death rate for hypothermia in this age group was 1.3 per 100,000 persons compared with the annual, age-adjusted death rate for hypothermia of 0.3 per 100,000 for the population as a whole (CDC, 1996). Men are at greater risk than women for hypothermia-related deaths (CDC, 1996; Hirvonen, 2000).

Studies of the regulation of body temperature in the elderly have shown that physiologic components of thermo-regulation contribute to an increased vulnerability to hypothermia (Centers for Disease Control, 1985; Hirvonen, 2000). Ethanol abuse, which is strongly associated with hypothermia, results in vasodilation and interferes with peripheral vasoconstriction, an important physiologic mechanism of defense against the cold (CDC, 1996; Hirvonen, 2000). Neuroleptic drugs predispose to hypothermia by inducing vasodilation and suppressing the shivering response. The hypothermic effects of neuroleptic drugs are amplified by lower ambient temperatures. Other risk factors for hypothermia include hypothyroidism, mental illness, dehydration and starvation, homelessness and poverty, immobilizing illness, and sustained contact with materials that promote conductive heat loss (e.g., water, solvents, and metals) (CDC, 1996: Hirvonen, 2000).

Similar risk factors exist for heat stroke (Keatinge et al., 2000). During the 1993 heat wave in Philadelphia, most of the heat-related deaths were in elderly people with preexisting natural diseases who lived alone and had no air-conditioning (Mirchandani et al., 1996). In the 1995 heat wave in Chicago, the majority of excess hospital admissions were due to dehydration, heat stroke, and heat exhaustion, and occurred among people with underlying medical conditions (Semenza, McCullough, Flanders, McGeehin, & Lumpkin, 1999).

Canadian data on determinants of environmental-temperature injury are not available.

The Situation in Montreal

Although Montreal has cold, harsh winters and hot, humid summers, little is known about the incidence of environmental-temperature injury in this city. The situation applies throughout Canada, where data on these diseases are limited and not readily accessible. The United States also is not immune to these limitations. In 1997, the Council of State and Territorial Epidemiologists (CSTE), in collaboration with the Association of Schools of Public Health and CDC, conducted a survey of environmental epidemiologists in the 50 states, the District of Columbia, and Puerto Rico (CDC, 1998). Of the 174 identified environmental public health surveillance systems in these jurisdictions, only four included heat stroke and hypothermia, making those conditions the least frequently monitored environmental diseases (CDC, 1998).

The authors of this study conducted a preliminary assessment of environmental-temperature injury in Montreal Island, Quebec, Canada, with three objectives in mind. The first was to approximate the incidence of environmental-temperature injury among Montreal Island residents. The second was to contextualize to Montreal Island known risk factors associated with environmental-temperature injury. The third objective was to lay the groundwork for the development of a surveillance system for environmental-temperature injury on Montreal Island.

Methods

Population

This study targeted the resident population of the Montreal-Center administrative region of Quebec, Canada. This region includes Montreal Island and adjacent Ile-Bizard. The number of people residing in this region was 1,802,309 in 1999.

Incidence, Risk Factor Assessment, and Data Methods Evaluation

The authors assessed environmental-temperature injury among Montreal Island (and Ile-Bizard) residents through coroner's reports, death certificate data, hospital discharge data, and hospital chart reviews. The various agencies that collect these data were asked to provide records for the three-year period prior to December 31, 1999. The time periods provided by each agency varied according to the availability of data, as described below. Environmental-temperature injuries were identified with the International Classification of Disease coding system (ICD-9) for effects of heat and reduced temperature (ICD-9 codes 991 and 992) (U.S. Department of Health and Human Services, 1990). The ICD-9 remained in use in Quebec until the year 2000. The sidebar above details all diagnoses in these ICD-9 categories.

The Chief Coroner's Office for Quebec provided reports of investigated deaths due to temperature injury among all Montreal Island residents during the four-year period between January 1, 1994, and December 31, 1998. Investigated deaths caused by temperature injury were extracted according to ICD-9 codes 991 and 992. Information available in the reports included age, gender, date of death, cause of death, other diagnoses, and circumstances of death. It is important to note that not all environmental-temperature-injury deaths are investigated by the coroner's office--only a minority.

Death certificate data for all of Montreal Island were available for the period beginning April 1, 1994, and ending March 31, 1997, through the Quebec provincial mortality database. This database collects all death certificates from every physician and health institution in the province. Causes of death are coded by all hospital centers according to the ICD-9 coding system. Deaths recorded as caused by temperature injury (ICD-9 codes 991 and 992) during this time were extracted. Information available in this database included age, gender, year of death, and cause of death.

Discharge data for all hospital centers on Montreal Island were accessed through the provincial hospital discharge database known as MED-ECHO (Maintenance el Exploitation des Donnees pour l'Etude de la Clientele Hospitaliere). This database collects discharge summary forms from medical admissions in all hospital centers in the province (not including emergency department visits). The forms are completed by physicians at patient discharge. Discharge diagnoses are verified and coded by hospital archivists according to the ICD-9 system. For the three-year period from April 1, 1993, and Marcia 31, 1996, hospitalizations with diagnosis of temperature injury (ICD-9 codes 991 and 992) were extracted for all 22 Montreal Island area acute-care hospitals. Information available in this database included age, gender, year of admission, principal diagnosis, and secondary diagnoses.

Finally, a hospital chart review was conducted at the Montreal General Hospital (MGH) for all admissions with a diagnosis of temperature injury (ICD-9 codes 991 and 992) among residents of Montreal Island from January 1, 1992, to November 30, 1999. The MGH receives approximately 10 percent of all Montreal hospital admissions (all causes) and is one of the busiest hospital centers on the island. The MED-ECHO database revealed that the MGH had the largest number of temperature injury admissions of all Montreal hospital centers over the three-year period described above. The authors selected the MGH for conducting a chart review. Information extracted from the hospital charts included age, gender, admission date, principal diagnosis, and potential risk factors for all cases of temperature injury.

Table 1 summarizes the above sources of data, including the information gathered and the time periods of review. The years of data for the different sources vary with availability: To maximize the amount of available information, this report presents results from all sources and all timeframes.

Results

Coroner's reports for Montreal Island during the four-year period between January 1, 1994, and December 31, 1998, identified 13 investigated deaths due to environmental-temperature injury (Table 2). Of these, 12 resulted from cold exposure, with hypothermia listed as the cause of death. The single death due to hyperthermia occurred in an elderly female. The average age of victims for all recorded deaths was 58, and seven of the cases occurred in males. Four victims were described as having a history of schizophrenia or depression. Alcoholism was also an underlying condition in three deaths. The lack of heat due to the 1998 ice storm power outages was noted in two cases. Previous cardiac conditions and exertion also were listed as precipitating factors for hypothermia.

Death certificate data tot Montreal were available for the three-year period beginning April 1, 1994, and ending March 31, 1997. This database identified 18 deaths due to environmental-temperature injury. Four deaths (all in females) were classified as due to extremes of heat, and 14 to extremes of cold. Age was an important factor in the heat-related deaths, with all cases occurring in victims over 80 years old. Thirteen of the 14 cold injury deaths were due to hypothermia; 11 of these occurred in males. Among fatalities, the average age of victims was 61 years.

MED-ECHO discharge data for the 22 acute-care hospitals on Montreal Island was accessed for the three-year period from April 1, 1993, to March 31, 1996. During that time, 165 admissions related to environmental-temperature injury were recorded in Montreal Island hospitals. Of the 165, 88 were coded with environmental-temperature injury as the principal diagnosis, and 77 were coded with temperature injury as a secondary diagnosis. One hundred and thirty-nine hospitalizations were related to cold injury, and 26 were related to heat extremes. Forty-two victims were female, and 123 were male. The average age of all victims admitted to acute-care hospitals was 50 years.

The chart review at the Montreal General Hospital (MGH) revealed 26 admissions with a diagnosis of environmental-temperature injury (ICD-9 codes 991 and 992) from January 1, 1992, to November 30, 1999. Of the 26 admissions over this seven-year period, 24 were a result of cold injury and two of heat injury. The two cases of heat injury were diagnosed as heat stroke, and both occurred in elderly women. Of the 24 cases of cold injury, 16 were diagnosed as frostbite, and the remaining eight were diagnosed as hypothermia. Two-thirds of the cold injury cases occurred in males. Two cases of cold injury resulted in death. The average age of all cold injury victims was 41 years. Table 3 summarizes the antecedent factors noted in the hospital charts of all 26 patients admitted for environmental-temperature injury. Alcohol intoxication was a prominent risk factor in 11 of the 26 incidents; 82 percent of the admissions involving alcohol occurred among males. Four admissions resulted from substance intoxication other than alcohol (cocaine and tricyclic antidepressants). Five people were described as suffering from mental illness (psychosis, suicide attempts), and two were homeless.

Using MED-ECHO discharge data, the authors estimated the incidence of environmental-temperature injuries requiring hospitalization on Montreal Island at 55 cases per year (165 hospitalizations in three years). Using the death certificate database, the authors estimated mortality from all environmental-temperature injuries on Montreal Island at six deaths per year (18 deaths in three years). Using 1999 population statistics, the authors estimated the incidence rate of environmental-temperature injury requiring hospitalization on Montreal Island at 3.1 per 100,000 person years, and the mortality rate from all environmental-temperature injuries on Montreal Island at 0.3 per 100,000 person-years. Eighty-four percent of all environmental-temperature injuries were a result of cold injury; only 16 percent were due to heat injury.

Discussion

Summary of Results

The incidence rate of environmental-temperature injury requiring hospitalization on Montreal Island was estimated at 3.1 per 100,000 person-years, and the mortality rate from all environmental-temperature injuries on Montreal Island was estimated at 0.3 per 100,000 person-years. The majority of hospitalizations and deaths were due to cold injury. All deaths due to heat injury, occurred in elderly females. Male gender, alcohol intoxication, psychiatric illness, and older age were found to be important risk factors for cold injury Younger extremes of age were not noted as posing a risk of temperature injury. Incidence measures, mortality measures, and risk factors that were associated in this study with temperature injury correspond to previous, albeit limited, knowledge (Centers for Disease Control, 1985; CDC, 1996; Hirvonen, 2000; Keatinge et al., 2000: Mirchandani et al., 1996; Semenza et al., 1999).

Anecdotally, chief emergency department physicians from six university-affiliated hospitals on Montreal Island were interviewed in November and December of 1999 to provide preliminary information on the incidence, diagnosis, and treatment of environmental-temperature injuries presented to Montreal Island emergency rooms. Each chief physician estimated that two to three cases of cold injury, were seen in the emergency department of his or her institution per year, and that heat injury consultations were less frequent (zero to one per hospital emergency department per year). The number of cases quoted by the six emergency department physicians reveals environmental-temperature injuries on an order of magnitude similar to that found in the four data sources used for this study.

Furthermore, data from Montreal's single ambulance service provider (Urgences Sante) are consistent with these findings, showing environmental-temperature injuries on an order of magnitude similar to that found in the four data sources used for the study. From January 1, 1997, to December 31, 1999, approximately 150 ambulance transports to Montreal hospitals were related to cold injury. Ambulance transport data for heat injury are not available.

The Data Sources

Each of the four main sources of data in this study provided unique strengths. Both MED-ECHO discharge databases and death certificate databases provided population coverage for the entire Montreal Island and were thus indispensable in estimating the rate of hospital admissions and deaths due to temperature injury. Coroner's reports contained greater detail pertaining to precipitating factors in deaths. For example, in one death, laboratory investigation revealed a methotrimeprazine level three times greater than the therapeutic level in a schizophrenic individual. The Montreal General Hospital chart review provided detailed environmental and clinical information for hospital admissions. For instance, frostbite of the foot was described in a transsexual male who had been wearing high heels with an open toe and standing outdoors for a prolonged period of time.

All four data sources complemented each other in providing a global understanding of both incidence and determinants of environmental-temperature injury. Although the yearly structure of the administrative databases caused a lag in available data, the information available was current, and the time periods available for all four data sources were long enough to support stable estimates. Because available time periods were staggered among the four data sources, each source provided unique information.

Characteristics and Limitations

Even though Canada's diverse climatic conditions warrant investigation of environmental-temperature injury, to the authors' knowledge, this study is the first in the scientific literature attempting to describe rates and determinants of these injuries in a Canadian city.

Given the sources of data assessed, this report does, however, have inherent limitations. Diagnostic imprecision may have limited the labeling of true cases of environmental-temperature injury For example, secondary diagnoses such as heart failure may have been more commonly recorded as primary diagnoses in hospital discharge or death, even though the underlying cause may have been hypothermia. Recognition of "accidental" hypothermia related to drowning, trauma, and severe exertion also was limited by minimal details recorded in the available data sources. Furthermore, data available from death certificate databases contained only a primary diagnosis and no secondary diagnoses. Information on potential risk factors for temperature injury abstracted from Montreal General Hospital medical charts describes factors elicited by physicians during medical care, and these factors therefore should be recognized as suggestive of potential risk rather than as clearly causal. Coroner's reports suffer less from this handicap because of the coroners' mandate to thoroughly identify potential risk factors.

Matching and correlating cases among hospital charts, coroner's reports, death certificates, and MED-ECHO data proved difficult. The difficulty was due to the lack of a unified identifier code in the three databases. Coroner's data did not include all temperature-related deaths and thus may be unrepresentative of all deaths.

The periods of data availability differed for the four data sources, the emergency department physician interviews, and the ambulance transport information. As previously mentioned, lack of data availability meant that this study could not present results for a specified timeframe without sacrificing the relatively limited data already available. The time periods of all sources are comparable, however, and complement each other in revealing environmental-temperature injury information for Montreal Island over the past decade.

Finally, individuals with environmental-temperature injuries who consulted physicians in the emergency departments (without admission), the community, and non-hospital centers could not be captured by this study, unless mortality was the final outcome. This group of patients is likely to be much larger than the groups of patients who were admitted to a hospital or who died from environmental-temperature injuries. The range of diagnoses of temperature injury seen in these patients, would not be equivalent to those found through hospitalization and death records. For example, patients with minor frostbites that would not require admission may simply have been sent home after a brief emergency department visit. Thus, few minor frostbites would be retrievable in the four main data sources used by this study.

Timeliness of the information also was of concern because the databases available are updated only on an annual basis. In addition, these databases (Death Certificate database, MED-ECHO) contain limited information on risk factors.

Information gleaned from the data sources is unlikely to overestimate the true rates of environmental-temperature injury on Montreal Island, but the degree of underestimation needs further elaboration.

Conclusion: Future Directions

The morbidity and mortality of environmental-temperature injuries, and the climatic variability found in Canada underscore the need to know more about the incidence of these conditions and better understand their determinants. The assessment presented here for Montreal Island provides a first step, from which environmental-temperature injury can be better explored and understood. The limitations encountered in this study indicate the importance of accruing better data.

One unique feature of Montreal Island is its novel ambulance transport record database, which was tapped for the first time in this study and has great potential for future use. Furthermore, the authors have recently developed a new design for surveillance of environmental-temperature injury in Montreal Island. The groundwork for this design was laid by the study reported here. The design involves physician reporting of cases to the Montreal Regional Public Health Department, followed by case investigation through patient interviews and chart reviews. This design will, the authors hope, circumvent many problems raised in the present study and provide more accurate measures of environmental-temperature-injury incidence and determinants. An initial pilot project is being planned to target emergency department physicians in all Montreal hospital centers. Increased knowledge about the incidence and determinants of environmental-temperature injuries may suggest areas for effective intervention to decrease the morbidity and mortality associated with these diseases.

TABLE 1

Sources of Data on Environmental-Temperature Injury
on Montreal Island

                                 Available Time
Source                               Period

Coroner's reports       January 1, 1994-December 31, 1998

Death certificate       April 1, 1994-March 31, 1997
mortality database

MED-ECHO Hospital       April 1, 1993 March 31, 1996
discharge database

Montreal General        January 1, 1992-November 30, 1999
Hospital chart review

                        Identification
                        of Temperature
Source                      Injury       Available Variables

Coroner's reports           ICD-9        Age, gender, date of death,
                                         cause of death, other
                                         diagnoses, circumstances
                                         of death

Death certificate           ICD-9        Age, gender, year of death,
mortality database                       cause of death

MED-ECHO Hospital           ICD-9        Age, gender, year of
discharge database                       admission, principal
                                         diagnosis, secondary
                                         diagnosis

Montreal General            ICD-9        Age, gender, admission
Hospital chart review                    date, principal diagnosis,
                                         potential risk factors for
                                         temperature injury

TABLE 2

Risk Factors Associated with Coroner-Investigated
Environmental-Temperature-Injury Deaths in the
Montreal-Centre Region, 1994-1998

                                                       Associated
                                                         Deaths
Risk Factor                                             (n = 13) *

Male gender                                               54%
Psychiatric illness (psychosis, depression, suicide)      31%
Age >65 years                                             31%
Alcohol intoxication                                      23%
Previous cardiac condition/exertion                       23%
Substance intoxication (cocaine, antidepressants)         15%
Homeless                                                  15%
Lack of home heat (ice storm-power outage)                15%
Unknown (information unavailable)                         8%
Impaired cognition (Alzheimer's disease, dementia)        8%

* Several deaths were associated with more than one risk factor.

TABLE 3

Antecedent Factors Associated with Environmental-Temperature-Injury
Admissions at the Montreal General Hospital, 1992-1999

                                                       Associated
                                                       Admissions
Antecedent Factor                                      (n = 26) *

Male gender                                               62%
Alcohol intoxication                                      42%
Psychiatric illness (psychosis, depression, suicide)      19%
Age >65 years                                             11%
Substance intoxication (cocaine, antidepressants)         15%
Accidental (motor vehicle)                                8%
Homeless                                                  8%
Unknown (information unavailable)                         8%
Lack of home heat (ice storm power outage)                4%

* Several admissions were associated with more than one risk factor.

REFERENCES

Centers for Disease Control and Prevention. (1996). Hypothermia-related deaths--Vermont, October 1994-February 1996. Morbidity and Mortality Weekly Report, 45(50), 1093-1095.

Centers for Disease Control and Prevention. (1998). Monitoring environmental disease--United States, 1997. Morbidity and Mortality Weekly Report, 47(25), 522-525.

Centers for Disease Control, (1985). Perspectives in disease prevention and health promotion: Hypothermia-associated deaths--United States, 1968-1980. Morbidity and Mortality Weekly Report, 34(50), 753-754.

Dematte, J.E., O'Mara, K., Buescher, J., Whitney, C.G., Forsythe, S., McNamee, T., Adiga, R.B., & Ndukwu, I.M. (1998). Near-fatal heat stroke during the 1995 heat wave in Chicago. Annals of Internal Medicine, 129(3), 173-181.

Hirvonen, J. (2000). Some aspects on death in the cold and concomitant frostbites. International Journal of Circumpolar Health, 59(2), 131-136.

Keatinge, W.R., Donaldson, G.C., Cordioli, E., Martinelli, M., Kunst, A.E., Mackenbach, J.P., Nayha, S., & Vuori, I. (2000). Heat related mortality in warm and cold regions of Europe: Observational study. BMJ (Clinical Research ed.), 321(7262), 670-673.

Mirchandani, H.G., McDonald, G., Hood, I.C., & Fonsea, C. (1996). Heat-related deaths in Philadelphia--1993. American Journal of Forensic Medicine and Pathology, 17(2), 106-108.

Semenza, J.C., McCullough, J.E., Flanders, WD., McGeehin, M.A., & Lumpkin, J.R. (1999). Excess hospital admissions during the July 1995 heat wave in Chicago. American Journal of Preventive Medicine, 16(4), 269-277.

U.S. Department of Health and Human Services. (1990). ICD-9-CM: International classification of diseases, clinical modification (3rd ed., 9th rev.). Washington, DC: Author.

ICD-9 Coding for Diagnoses of Environmental-Temperature Injury

Effects of Heat

992.0 Heat stroke and sunstroke 

-- Heat apoplexy 

-- Heat pyrexia 

-- Ictus solaris

-- Siriasis

-- Thermoplegia

992.1 Heat syncope

-- Heat collapse

992.2 Heat cramps

992.3 Heat exhaustion, anhydrotic

-- Heat prostration due to water depletion

992.4 Heat exhaustion due to salt depletion

-- Heat prostration due to salt (and water) depletion

992.5 Heat exhaustion, unspecified

-- Heat prostration, not otherwise specified (NOS)

992.6 Heat fatigue, transient

992.7 Heat edema

992.8 Other specified heat effects

992.9 Unspecified

Effects of Cold

991.0 Frostbite of face

991.1 Frostbite of hand

991.2 Frostbite of foot

991.3 Frostbite of other and unspecified sites

991.4 Immersion foot

-- Trench foot 

991.5 Chilblains

-- Erythema pernio 

-- Perniosis

991.6 Hypothermia

-- Hypothermia (accidental)

991.8 Other specified effects of reduced temperature

991.9 Unspecified effect of reduced temperature

-- Effects of freezing or excessive cold, NOS

Dr. Athanasios Tom Koutsavlis, Vice-President, Scientific Affairs, Paladin Labs Inc., 102-6111 Royalmount Avenue, Montreal, Quebec, Canada, H4P 2T4. E-mail: [e-mail].