Heat-RelatedDeaths—Four States, July-August 2001, and United States, 1979-1999

Each year in the United States, approximately 400 deaths are attributed to excessive natural heat; these deaths are preventable. This report describes heat-related deaths in Missouri, New Mexico, Oklahoma, and Texas when elevated temperatures were recorded for several consecutive days during July-August 2001; summarizes heat-related deaths in the United States during 1979-1999; and presents risk factors and preventive measures associated with heat-related illness and death, especially in susceptible populations.

witness reported that he had complained about abdominal pain and vomiting. He arrived at an emergency department in New Mexico 3 hours after he was found. His rectal temperature was 105.7°F (40.9°C). The patient had laboratory evidence of rhabdomyolysis, severe dehydration, and renal failure. Blood alcohol level and a screen for drugs were negative. He died 3 hours after arrival at the hospital. Cause of death was attributed to hyperthermia due to environmental heat exposure. High temperature at the border that day was 90°F (32°C).
heat exhaustion include heavy sweating, muscle cramps, fatigue, weakness, paleness, cold or clammy skin, dizziness, headache, nausea or vomiting, and fainting. Untreated heat exhaustion can progress to heatstroke. 4 Even with prompt medical care, 15% of heatstroke cases are fatal. 5 Symptoms of heatstroke include a high body temperature (oral temperature of Ն103°F [Ն39.4°C] or a rectal temperature of 106°F [41.1°C]); red, hot, dry skin and no sweating; rapid pulse; throbbing headache; dizziness; nausea; confusion; disorientation; delirium; and coma. Heatstroke can occur in the absence of physical exertion. Infants, elderly persons, socially isolated persons, bedridden persons, and persons with certain mental and chronic illnesses are at highest risk. 6,7 The elderly, especially those aged Ն80 years, are susceptible to heat-related illness because they are less able to adjust to physiologic changes (e.g., vasodilation) that occur with exposure to excessive heat and are more likely to be taking medication for chronic illness (e.g., tranquilizers and anticholinergics) that increase the risk for heat-related illness. 5 Infants also are sensitive to heat. Conditions such as mild fever can progress quickly to heatstroke if heat stress occurs. Parents and other caregivers should provide adequate hydration during summer months and refrain from dressing children too warmly. 5 Adults also should keep well hydrated during summer months.
Heatstroke also can occur in young, healthy persons who are exercising, 6 because physical exertion during hot weather increases the likelihood of fainting and cramps caused by increased blood flow to the extremities. 5 Onset of heatstroke can be rapid and is considered a medical emergency.
The findings in this report are subject to at least three limitations. First, information on decedents is provided by surrogates, who might not accurately describe characteristics or behavior of the decedents. Second, heat-related deaths due to weather conditions or exposure to excessive natural heat might repre-sent only a portion of actual heatrelated deaths. These deaths often are a diagnosis of exclusion and can be misclassified as a stroke or heart attack. Deaths attributed to cardiovascular and respiratory disease increase following heat waves. 8 In addition, jurisdictions might use different definitions of heatrelated death. Finally, ICD-10 coding was introduced in 1999 and might not be comparable with previous data for 1979-1998.
To reduce morbidity and mortality from heat-related illness, many cities have developed emergency response plans. Local officials use meteorologic information and assess population characteristics to implement prevention strategies. 7 Spending time in an airconditioned area is the strongest factor in preventing heat-related deaths. 1,9 The use of fans does not appear to be protective during periods of high heat and humidity. 1 If exposure to heat cannot be avoided, prevention measures should include reducing or eliminating strenuous activities or rescheduling them for cooler parts of the day; drinking water or nonalcoholic fluids frequently; taking cool showers frequently; wearing lightweight, lightcolored, loose-fitting clothing; and avoiding direct sunshine. 9 Public health messages disseminated to all age groups can make the public aware of the signs and symptoms of heat-related illness. Prevention messages delivered as early as possible in the media can prevent heat-related illness, injury, and death. 1 Because many heat-related illnesses and deaths occur among the elderly population, older persons should be encouraged to take advantage of airconditioned environments (e.g., shopping malls, senior centers, and public libraries) for part of the day. Parents and other caregivers should be educated about the heat sensitivity of children aged Ͻ5 years. 5 *The National Weather Service issues a heat advisory when the maximum daytime heat index is expected to be Ն105°F (40.6°C) and the minimum nighttime heat index is expected to be 80°F (26.7°C) for 2 or more consecutive days. The heat index takes into account air temperature and relative humidity and indicates the actual feel of the temperature to the body.

Acknowledgments
†Underlying cause of death during 1979-1998 is classified according to the International Classification of Disease, Ninth Revision (ICD-9). Excessive heat has three categories: E900.0 "due to weather conditions," E900.1 "of man-made origins," and E900.9 "of unspecified origin." The data for 1999 are from ICD-10; code X30 "exposure to excessive natural heat" was added to the 1979-1998 ICD-9 code E900.0, "excessive heat due to weather conditions." (four), Ohio (two), and Connecticut (two). Antimicrobial-susceptibility testing of three isolates by CDC revealed resistance to amoxicillin/clavulanate, ampicillin, cefoxitin, ceftiofur, cephalothin, chloramphenicol, streptomycin, sulfamethoxazole, and tetracycline. In addition, two of three isolates were resistant to kanamycin; two had decreased susceptibility or resistance to ceftriaxone. To determine the cause of the outbreak, the New York State Department of Health (NYSDOH) and CDC conducted a case-control study. This report summarizes the results of this investigation, which implicated exposure to raw or undercooked ground beef as the vehicle of transmission. The findings also highlight the emergence of multidrug-resistant S. Newport in the United States. These strains exhibit decreased susceptibility or resistance to ceftriaxone, thereby complicating empiric therapy for serious Salmonella infections. Clinicians should be informed of the emergence of these S. Newport strains, and persons should refrain from eating undercooked ground beef and wash their hands after handling raw ground beef.

Outbreak of Multidrug-Resistant
The outbreak was identified on February 11, when a county health department notified NYSDOH of seven cases of S. Newport infection. Pulsed-field gel electrophoresis (PFGE) testing by the NYSDOH laboratory revealed that six isolates had an indistinguishable pattern, and one isolate had a single band difference. NYSDOH defined a case as isolation of S. Newport with a PFGE pattern that was indistinguishable or one band different from the outbreak pattern. Additional cases were reported from Connecticut, Michigan, Ohio, and Pennsylvania through the National Molecular Subtyping Network for Foodborne Disease Surveillance (PulseNet).
A total of 47 cases from the five states was identified. The median age of infected persons was 45 years (range: 2-81 years); 33 (70%) were females. Symptom onsets occurred during January 1-April 4, with 33 (73%) occurring during February 1-15. Of the 47 patients, 46 were interviewed. The median du-ration of illness was 9 days (range: 3-60 days). Predominant symptoms included diarrhea (100%), abdominal pain (91%), fever (78%), blood-tinged stools (52%), and vomiting (48%). Six (13%) patients reported other symptomatic household members. A total of 33 (72%) patients received antimicrobial agents, and 17 (37%) were hospitalized. One patient from New York with leukemia developed sepsis and died; S. Newport was identified in both blood and stool cultures from this patient. A total of 44 isolates had an indistinguishable PFGE pattern after analysis with two enzymes (XbaI and AvrII); three isolates differed by one band.
To identify exposures associated with illness, NYSDOH and CDC compared 36 patients (28 from New York, four from Michigan, and four from Pennsylvania) with 85 controls, who were interviewed through random-digitdialing in case-patients' home area codes and frequency-matched by age group. A multivariate logistic regression analysis indicated that 22 (67%) of 35 case-patients had eaten ground beef during the 3 days before illness onset compared with 31 (53%) of 58 controls (odds ratio [OR] = 2.3; 95% confidence interval [CI] = 0.9-5.7). Case-patients and controls were asked about eating raw or undercooked ground beef during the 3 days before illness onset. Of the 26 case-patients who answered definitively, 12 (46%) had eaten raw or undercooked ground beef compared with one (1%) of 80 controls (OR=50.9; 95% CI=5.3-489.0). A total of 11 patients recalled the type of ground beef eaten; seven (64%) had eaten lean or extra-lean ground beef. The U.S. Department of Agriculture (USDA) Food Safety and Inspection Service (FSIS) was notified after this investigation implicated ground beef as a potential vehicle for exposure.
One New York patient had a leftover, frozen, uncooked meatloaf prepared with the same package of ground beef that was used to prepare meals eaten during the 3 days before onset of symptoms. A culture of the meatloaf yielded S. Newport with a PFGE pattern indis-tinguishable from the outbreak pattern. Traceback by FSIS of ground beef eaten by 12 New York patients identified a meat packing plant that could have supplied the meat eaten by all those identified in the outbreak. Review of distribution records, grinding logs, and purchasing information did not identify any specific lot of ground beef, and no intact ground beef sample processed by the plant during the outbreak period was available for testing by FSIS. On April 19, USDA issued a Public Health Alert reminding consumers of food safety guidelines. FSIS is examining practices that might contribute to contamination of meat by this pathogen. Since 1996, the National Antimicrobial Resistance Monitoring System (NARMS) for Enteric Bacteria has identified an increasing number of S. Newport isolates that are resistant to at least nine of 17 antimicrobial agents tested: amoxicillin/clavulanate, ampicillin, cefoxitin, ceftiofur, cephalothin, chloramphenicol, streptomycin, sulfamethoxazole, and tetracycline. In addition, these isolates exhibit decreased susceptibility (minimal inhibitory concentrations [MIC] Ն16mg/ml) or resistance (MIC Ն64mg/ml) to ceftriax-FROM THE CENTERS FOR DISEASE CONTROL AND PREVENTION one, an antimicrobial agent commonly used to treat serious infections in children. Isolates with this resistance pattern have plasmids that carry a bla CMY gene. These genes produce AmpCtype enzymes, which confer resistance to penicillin-inhibitor combinations (e.g., amoxicillin/clavulanate), cephamycins (e.g., cefoxitin), and expanded-spectrum cephalosporins (e.g., ceftiofur and ceftriaxone). To distinguish this type of resistance from other multidrug-resistant strains, these strains are referred to as Newport MDR-AmpC. In 1998, one (1%) of 78 S. Newport isolates tested in NARMS was Newport MDR-AmpC compared with 33 (26%) of 128 in 2001. Although the full clinical significance of Newport MDR-AmpC is unknown, treatment of these infections with ceftriaxone might be ineffective. In addition, antimicrobialresistant Salmonella infections have been associated with an increased hospitalization rate, morbidity, and mortality. 2,3 During 2001-2002, several state health departments, including California, Connecticut, and Massachusetts, documented association of exposure to dairy farms, ill cattle, and cheese made from unpasteurized milk with increased human Newport MDR-AmpC infections. [4][5][6] In the outbreak described in this report, most patients for whom information is available ate lean or extralean ground beef; dairy cattle are an important source of lean or extra-lean ground beef. 7 These data suggest that cattle, particularly dairy cattle, might be a source for human Newport MDR-AmpC infection.
This report is the first to associate eating of ground beef, specifically raw or undercooked ground beef, with Newport MDR-AmpC infection. Recent U.S. surveys indicate that 11%-28% of persons report eating raw or undercooked ground beef, and approximately one third of persons do not use safe food-handling practices to prevent cross-contamination in the kitchen. 8 The USDA Pathogen Reduction/ Hazard Analysis and Critical Control Points (PR/HACCP) inspection system in meat and poultry plants has reduced Salmonella prevalence in raw ground beef from 7.5% in 1998 to 2.8% in 2001. 9 The emergence of Newport MDR-AmpC suggests that further measures might be necessary. Potential strategies include (1) evaluating practices on the farm to determine factors that might contribute to multidrugresistant S. Newport and developing interventions to eliminate these factors; (2) implementing the Public Health Action Plan to Combat Antimicrobial Resistance 10 ; (3) encouraging industry to implement processes such as steam pasteurization or irradiation of ground beef; and (4) increasing efforts to educate consumers on the importance of safe handling and cooking practices.
State health departments and veterinarians should investigate clusters of S. Newport and perform antimicrobialsusceptibility testing to determine if isolates are Newport MDR-AmpC. Epidemiologic investigations and PFGE comparison of outbreak isolates will help to identify food vehicles associated with Newport MDR-AmpC and to identify control points for reducing these infections. Because treatment with ceftriaxone might be ineffective, clinicians should be informed of the emergence of Newport MDR-AmpC strains. Persons should not eat undercooked ground beef and should wash their hands after handling raw ground beef.  4 During the DTaP vaccine shortage beginning in 2000, 5 ACIP recommended that health-care providers vaccinate infants with the initial 3 DTaP doses, if they did not have sufficient supply of DTaP to vaccinate all children in their practice. ACIP also recommended deferral of the fourth and fifth DTaP doses if supplies were still inadequate. 6 Supplies are now adequate to resume the full 5-dose schedule for DTaP vaccine. 1,3

MMR Vaccine
A temporary shortage of MMR vaccine in the United States resulted from a voluntary interruption of manufacturing operations of Merck & Co., Inc., the only manufacturer of this vaccine in the United States. 7 During the vaccine shortage, ACIP recommended deferral of the second dose of MMR vaccine at age 4-6 years if health-care providers were unable to obtain sufficient amounts of vaccine. The first dose at age 12-15 months was not to be delayed because of the severity of measles in young children. 7 Supplies are now adequate to resume the second dose of MMR vaccine. 2,3

Vaccine Supply
Health-care providers should review the vaccination status of their patients and administer DTaP and MMR vaccines, as appropriate. For at least the next 2 months, providers should order DTaP and MMR vaccine in amounts sufficient for a Յ30-day supply to ensure that current supplies can meet requests. Recall or special initiative programs can be instituted when DTaP and MMR vaccine supply improves further but should be deferred during this transition period. However, if children who need these vaccines seek medical care for other reasons, they should be administered vaccine provided no contraindications exist. Furthermore, vaccine should be offered to children who need vaccination and whose parents requested vaccination. CDC will continue to monitor DTaP and MMR vaccine supply and, if necessary, allocate vaccine. Updates regarding vaccine supply and shortages can be found at http://www.cdc.gov/nip/.