Hantavirus Pulmonary Syndrome (HPS) has emerged as a new infectious disease in the United States. Although it is still very rare with only around 160 cases to date, HPS is extremely dangerous with a 50% mortality rate. The disease was unknown until the late spring of 1993 when a cluster of cases appeared in the Southwestern United States. Like other hantaviruses, the Sin Nombre Virus (SNV) has a rodent host, the deer mouse, as its primary reservoir. Humans contract the virus by coming in contact with a rodent carrier directly or indirectly through inhaling their excreta as saliva, feces or urine. This can occur during agricultural work, cleaning of animal sheds, entering rarely-used structures and trapping and handling rodents, all of which have been identified as common risk factors. Case studies seem to demonstrate that infection with SNV occurs uniquely for each individual. Epidemiological and genetic studies have been able to document in only a few cases the exact moment of transmission and origin of the virus. Evidence does seem to suggest that transmission requires exposure to rodent excreta in an enclosed area and rarely occurs out in the wild outdoors.

In May 1993, a cluster of patients in western New Mexico and eastern Arizona entered hospitals and clinics with a fatal acute respiratory illness. They presented with a previously unidentified syndrome of pulmonary edema, fever, elevated WBC count and low oxygen saturation ( 17). On May 14 the NMDOH was notified of a couple in the same household, a 21-year old woman and 19-year old man, who had died within 5 days of each other of acute respiratory failure after an abrupt onset of fever, myalgias, headache and cough (14). Termed "The Four Corner's Disease" because of its appearance around the Four Corners area where Utah, Arizona, New Mexico and Colorado meet in the southwestern United States, the bizarre disease was identified in 24 case-patients by June 7, 1993. The onset began in December 1992, though 14 had symptom onset in May. The median age was 34; fourteen case-patients were Native American, nine were white and one was Hispanic. Twelve of the first 24 cases (50%) died (4).

The Centers for Disease Control and Prevention (CDC) quickly became involved in investigating the cause of these cases. Their lab's analysis of clinical and autopsy specimens found high titers of antibodies to hantaviruses. Six of the nice initial case-patients' serum IgG and IgM antibodies cross-reacted with Puumala, Prospect Hill and Seoul hantaviruses, suggesting a previously unknown virus (5). The CDC soon isolated the causative agent from deer mice lung tissue and visualized it by electron microscopy (15). RNA nucleotide sequence studies of virus glycoproteins and nucleocapsid of the M and S segments amplified with PCR from case-patients identified a new hantavirus. The virus was closely related to Prospect Hill and Puumala viruses but not a recombinant genetic shift of the three segments of RNA as suspected (16). During the week of June 6 rodents, known to be carriers of hantavirus, were collected from peridomestic settings of several case-patients. Of the 42 tested, 12 (29%) were positive of which all were deer mice, Peromyscus maniculatus (5). Hantavirus sequences amplified in 6 of the Ab positive mice were found to be closely related to sequences from 3 human cases, providing a direct genetic link between the rodents and the infected humans (6). By testing serum samples of people from the outbreak area in 1991 and 1992 for hantavirus Ab, seroprevalence was found to be 1% (3 /270) which further supported an etiological role for hantavirus as the causal infectious agent in the outbreak (5). By July cases were identified outside of the outbreak area in Nevada and Texas (7). In August two cases were confirmed in California and one Louisiana who had a hantavirus distinct from the four-corners virus (8) By December 1994, 108 cases of HPS from 21 states were confirmed, 56% of which were from the four corners area (20).

The virus has been coined the Sin Nombre Virus (SNV) after being known as the Four Corners virus, the Muerto Canyon virus and Convict Creek virus. Unlike the rest of the bunyaviridae viruses which are primarily transmitted via anthropod vectors, hantaviruses primary reservoirs are rodents which are persistently infected, shedding virus without clinically apparent symptoms. The most common worldwide hantavirus is the Seoul virus whose host is the Norwegian rat. Europe has the Puumala virus carried by the black vole, while in Asia the stripped field mouse serves as the reservoir for the Hantaan virus (17). These hantaviruses all clinically manifest as hemorrhagic fever with renal syndrome (HFRS) with Hantaan being the most severe and Puumala the mildest form. HFRS is a clinically distinct disease from HPS attacking the kidneys instead of the lungs. Otherwise, the viruses have similar epidemiologies and transmission routes. Early responses and prevention strategies to SNV relied on findings on Seoul, Puumala and Hantaan hantaviruses.

In the United States, since the identification of SNV many other strains of hantaviruses have been identified that also give rise to a HPS like syndrome in humans. El Moro Canyon virus infects R.eithrodontomys megalotis. Microtus californicus and M. pensylvanicus , the meadow vole, are carriers of the Isla Vista and Prospect Hill viruses (26). In Florida the cotton rat (Sigmodon hispidus) is associated with the Black Creek Canal virus and Muleshoe viruses (B. Hjelle, personal communication). The rice rat (Orysomys paulstris) has been found to carry the Bayou virus in Louisiana. Cases of HPS from the Florida and Louisiana viruses were discovered after the 1993 SNV outbreak. A variant of SNV found in the white-footed mouse (P. leucopus) has been discovered in New York but no known cases of infected humans exist (20).

Peromyscus maniculatus , the deer mouse, is endemic to most of the continental US excluding the Southeast and the Atlantic seaboard, with a range that spans into Canada and Mexico prevalent in mostly rural areas. It has been shown to be the primary rodent reservoir for the Sin Nombre virus. To find this, researchers trapped rodents from June to August 1993 in the four corners region in households of human cases of HPS, nearby homes or locations where case-patients spent time. The sera of the 1687 rodents of 29 species was tested with ELISA for Ab to PHV, HTNV, PUUV and SEOV. P. maniculatus had an antibody prevalence of 30.4%, while P. truei and P. boylii had 19.6% and 5.9% respectively. The pattern of prevalence was mass-specific suggesting that hantavirus was acquired with age and horizontal transmission of virus. Peromyscus maniculatus was also the most frequent rodent captured accounting for 48% of the rodent community indicative its abundance (12). SNV was identified in spleen and lung tissues of mice from Mono County, CA caught in 1983. The high association of SNV to P. maniculatus , an indigenous species to North America, supports a long, developed co-evolution. The genetic diversity of SNV in the Western U.S. was up to 14%, suggesting that SNV did not recently emerge but is well-established in a virus-rodent relationship. (3)

The 1993 outbreak localized to the Four Corners region seems to be a result of increased vegetation and thus rodent population in the spring of 1993. Ecological factors such as weather can induce a higher incidence of host and human interaction, enhancing opportunities for viral transmission. After an usually wet winter there was a large crop of nuts and other rodent food. The local pinon tress for only the third time this century bore nuts according to Navajo Indians (28). The abnormally high precipitation originated in El Nino , a climatic phenomenon in the Western Pacific. The extra food led to a blossoming of rodents, documented to be a tenfold increase in areas of New Mexico. More rodents afforded more opportunities for contact with humans. (24) The crowding of rodents may have facilitated the transmission of virus among rodents and to increased human contacts. As more rodents entered human dwellings, more humans happened to aerosolize rodent excreta harboring virus (17). Thus, ecological factors play a significant role in the appearance of the disease.

The ecology of the rodent host also has an enormous impact on the epidemiology of the ensuing human disease. Deer mice, unlike the other rodents, invade human dwellings. As they inhabit human homes and structures they leave their excreta for humans to disturb, aerosolized and inhale. Beyond its behavioral differences, the abundance of P. maniculatus , may be the reason that it has such significance in the epidemiology of HPS. In most rodent trapping studies, the deer mouse is the most common, 53% of all rodents trapped (26). Certain areas have a higher prevalence of infected deer mice. Seroprevalences among deer mice range from essentially zero up to 80%. Most surveys in the West show between 2 and 20%. East of the Mississippi, however, it is rare to find prevalences above 2-3%. (B. Hjelle, personal communication). In the homes of case-patients, SNV seroprevalence for trapped deer mice was found to be 30%. .Studies in 1985 found that 11 of 218 (5%) P. maniculatus rodents collected in CA, NM and CO were positive for hantavirus antibodies (6). A more recently study of rodents trapped in rural areas of CA and NV found P. maniculatus had antibody prevalence to SNV of 12.5% with pockets of high prevalence (30%) and other areas of none positive. M. motaus, R. megalotis, P. truei had Ab prevalences of 11%, 8% and 4% respectively. (26) Prevalences vary as certain patches of rodent populations are infected while others in a similar habitat none are. This may explain why outbreaks are localized to certain areas even though the deer mice is endemic to a much larger range.

All human hantavirus infections are viral zoonoses. Humans have never been documented to transmit the virus to each other. A study of health care workers who had exposure to hantavirus either by caring for HPS patients, by doing laboratory work or by performing autopsies found that of the 396 participants none had any serological evidence of recent or past hantavirus infection. This shows that person-to-person transmission of SNV is unlikely, at least in health care workers (30). Nevertheless, precautions in dealing with SNV are taken as it is classified as a biosafety level 3 virus. Other hantaviruses such as Seoul have been transmitted in the laboratory in Japan and Europe (25). Preliminary results of studies of SNV-positive mice kept in cages with uninfected show that the virus is not easily spread. After four months of cohabitation none of the other animals became positive. Of the mice maintained in a 3'x2' terrarium for a year only 13% were seropositive, low considering the suspected spread by urine and feces (26). Thus it appears that the virus is not readily transmitted by aerosolized excreta which these mice were exposed to.

Seroprevalence data support the finding of rare transmission. A study in Baltimore of patients with community-acquired pneumonia of unknown etiology investigated the possibility of that they were HPS cases. The patients had symptoms of comparable to HPS, yet none of the patients had any evidence of SNV infection. This demonstrates that hantavirus cases do not go undetected and that HPS is indeed very rare, at least in the Baltimore area (2). Since rodents are known carriers, it seems likely that people with exposure to them would have a higher incidence of the disease. A study of foresters and park workers in area of the Southwest where SNV is endemic in mice, demonstrated that transmission of the virus is rare even among those with a high level of exposure. 143 park worker 64% of whom had repeated exposure to rodents, rodent nests and rodent droppings, showed no serological evidence of hantavirus infection. The parks polled were in NW NM and NE AZ, the area where the 1993 HPS outbreak occurred (29). This seems to imply that transmission requires more severe exposure. However, the forest workers sampled could have been from a pocket with low SNV seroprevalence since no tests were done.

Specific HPS cases can provide further insight into transmission routes. Epidemiological data on their habits, lifestyle and exposure patterns may give clues to a common route of infection. A 61-year-old hiker in the Appalachian Trail in Virginia came down with HPS. He reported mice excreta and rodent traps in shelters and bunk houses, which may be the possible source of his infection (10). A Rhode Island man died of HPS who had not traveled out of the Northeast within 2 months of his death. RT-PCR amplified from his lung showed a SNV-like viral sequence. Rodents trapped from his home, vacation house and two warehouses that had rodent infestation were not seropositive (9). Two cases of HPS occurred in South Dakota among co-workers of a cattle feedlot. The feedlot did not have rodent control and offered many opportunities to come into contact with rodent excreta or handling dead rodents with hay and straw piles, abandoned buildings and excess dirt, debris and spilled feed (11). A 14-year-old boy from North Dakota died of SNV caused HPS. Rodents trapped around his house were seropositive, specifically a deer mouse's RT-PCR matched the patient's (21). A family on a farm/ranch in rural New Mexico had two cases of HPS, first a four-year-old boy came down with a mild case of what was thought to be bronchitis, then his mother died of acute cardiopulmonary failure. This represented the first case of mild HPS and a rare appearance in a child. The house had numerous signs of rodent infestation, rodent droppings, burrows and gnawing sites (22). These cases demonstrate the common elements of rural settings and a history of rodent exposure in hantavirus cases. What also seems prevalent is an indoor exposure with rodent infestation of the home or nearby structures.

Genetic studies by B. Hjelle et al. demonstrate a direct link of specific P. maniculatus rodents to humans with HPS, providing some clues as to the path of infection. Investigators were able to find the source of infection by comparing sequences of viral SNV from case-patients with that isolated from rodents captured in areas frequented by the patient. Using RT-PCR with nested primers for the G1 protein and S segment, the researchers found exact, perfect matches for 4 of 6 patients investigated. One case was a carpenter from rural CA who traveled to NM after cleaning his cabin on May 26, 1995. It was unclear whether he contracted the virus in CA or NM. Rodents were captured near the home he visited in NM and in his cabin in CA. Only one, which was trapped in Mono County CA, of the 6 seropositive rodents matched the G1 glycoprotein sequence of the patient, pinpointing the sight of infection (18). These studies enable investigators to prove epidemiologic linkage of rodent SNV to human cases of HPS. Another study by Jay et al. of a utility worker showed a definite case of occupational exposure. After a 56 year-old man of Mono County CA developed HPS, rodents were trapped around his home, the control room, substations and outbuildings where he spent his time. A deer mouse from substation A had a nucleotide sequences identical to the patient. Two weeks before his illness, the patient had rodent droppings and dust fall on his desk from the replacing of ceiling panels in substation A (19). Therefore, it seems very likely from the genetic and exposure data that this was the source of his infection. These epidemiological studies seem to be the most promising way to directly link an HPS case with the infection source and establish how transmission occurred.

Recently, case-control studies have attempted to deduce risk factors for hantavirus infection. Studies conducted by Childs at the CDC compared the environment of 17 case households with selected homes near and far from the home, matched for location. Inspections were done of the outside habitat for rodents and of inside the home for food storage and rodent feces. Rodents were also trapped in and around the home to detect rodent density; the rodent sera was tested for hantavirus antibody. No significant differences were found in house type or outside habitat. Significantly higher levels of infestation appeared in case homes compared to both controls (p < .05) based on trap data. The rodent feces index approached statistical significance (p = .07) with more in case than control households. However, there was no significant difference in the SNV seroprevalence of the deer mice between case and control homes (13). Thus the key risk factor for contracting HPS in these rural settings appears to be rodent infestation of the home not from pockets of high seroprevalence. However, the small amount of case households and low statistical significance limits what conclusions can be drawn. Trapping is not an absolute measurement of rodent density.

At the same time, a parallel study by Zietz of the same 17 case patients with the same near and far controls was done to investigate particular activities done by patients placing them at risk for SNV infection. Twenty-two exposure variables were identified when cases were compared to the 236 controls matched for sex, race and age. Case patients were more likely to trap rodents (p=.02) and handle dead mice (p=.03) as 94% of patients had done so. The peridomestic activities (in and around one's house) associated with HPS were cleaning food storage areas or outbuildings. Agricultural activities such as using a hand plow (p=.003) and planting crops (p=.05) were also associated. The occupation of a herder also had a greater risk for HPS (p=.005). Also case-patients had a higher frequency of chronic medical condition than controls (p=.02) suggesting host factors had an affect (33). These activities all involve exposure to aerosolized rodent excreta as the debris is disturbed. If these households also have higher rodent infestation then they have a greater likelihood of contracting the virus by performing them. However, these data are based on patient recall and surveys which are inherently problematic. The small sample of cases again limits the findings. Brian Hjelle notes that, "the question of 'hand-plowing' drew a 'hit' in the Zeitz study, but personally I am very suspicious that this activity had anything to do with any person becoming infected... it is a marker for some other activity or characteristic that lead to infection," (personal communication, 3/7/97). Epidemiological questions have assumptions built in of transmission routes.

CDC Survey data on HPS cases show that rodent exposure is very ubiquitous risk factor. Of 27 patients, 25 (93%) reported exposure within 6 weeks of illness onset. This exposure most commonly occurred at home as 70% (48/69) had exposure to rodents associated with peridomestic activities in houses with rodent infestation. Combined occupational and peridomestic exposure was the next most frequent at 19% (13/70). Less common were occupational (4%) and recreational (9%) exposures (20). Brian Hjelle disputes this, saying that, "the CDC is undercalling work-related risks...the large number of cases associated with the Four Corners outbreak is leading to an overestimate of the importance of peridomestic exposures, " (3/7/97). Pets may possibly play a role in SNV exposure, as they did with other hantaviruses: cats in Austria had a 5% seroprevalence and were a risk factor in Scandinavia for PUUV infection and in China for HFRS (31). Also documented by people at the CDC was an association of HPS with entering or cleaning a rarely used, rodent infested structure. They have 6 confirmed cases of HPS who had a history of this activity. A 29-year-old woman developed HPS one to two weeks after cleaning a vacant house. A fatal HPS case had entered a converted storage shed heavily infested with rodent droppings that had been closed for many months, two weeks before illness onset (1). In these cases, the people acquired SNV probably by aerosolizing the virus from the rodent debris. The limited circulation in the enclosed area made it more likely that they inhaled the aerosolized particles. Brian Hjelle says, "I've personally interviewed quite a number of [HPS] patients...indoor exposures are nearly universal," (3/7/97). He and other N. American epidemiologists agree that indoor exposures to rodent excreta are very important to transmission.

Overall, the data does seem to support the belief that indoor exposures to aerosolized rodent excreta are a major source of SNV transmission. The forestry workers and health care workers, mysteriously without SNV seropositivity though they both were exposed, lacked indoor exposure to rodent excreta. The case studies of recent HPS patients, from the hiker's shelters to the New York warehouse, all had some sort of exposure to rodents and their excreta in an enclosed area. The indoor requirement also explains the outbreak in 1993 due to the explosion in the rodent population. The mice were forced to invade homes in rural areas with food, leaving their droppings full of virus for people to come into contact with. With such high numbers the rodent infested more homes in greater numbers, bringing the virus into people's homes and the gruesome disease of hantavirus pulmonary syndrome.

Future research should be done to investigate precisely how and when SNV is transmitted from the rodent hosts to humans. Perhaps using the same genetic, epidemiologic technique as in the Hjelle study (18), further studies can clarify this. More case-control studies should be done with a larger sample of HPS cases from a more diverse setting to more accurately reflect possible risk factors. Research on SNV is also needed to identify the respiratory tract tropism and membrane protein for SNV entry into the cells. Understanding the internal pathway of the virus may help to elucidate how it is transmitted. Studies of rodent excreta could show how long the virus persists outside the organism and where, in the urine, feces or saliva it is shed in the highest concentration. This could explain how exactly the rodents transmit the Sin Nombre virus and what places humans at risk to contract it. This could also confirm that the virus is spread by rodent excreta, as this is only assumed based on knowledge from other hantaviruses.

The policy implications of these findings are enormous. An emerging disease like hantavirus pulmonary syndrome with such extreme morbidity spread by a highly ubiquitous rodent is a major threat. Knowledge of transmission would be extremely helpful in prevention guideline development, potentially saving lives. With an understanding of how HPS-patients contracted SNV, specific strategies could be undertaken to reduce the spread of the virus. The government could start a rodent extermination program to eradicate rodent populations with high seroprevalences in an effort to reduce the amount of SNV. A milder intervention effort would be a campaign to inform people on the danger of HPS and its rodent carriers especially in high risk rural areas. It would encourage households to implement rodent control programs. Understanding the virus and its glycoproteins could be valuable for drug development of for this deadly disease hantavirus pulmonary syndrome.

1. Armstrong, L. et al., "Hantavirus Pulmonary Syndrome Associated With Entering or Cleaning Rarely Used, Rodent-Infested Structures," The Journal of Infectious Disease, 1995, 172: 1166

 2. Auwaerter, P., et al., "Hantavirus Serologies in Patients Hospitalized with Community-Acquired Pneumonia," The Journal of Infectious Disease, 1996, 173: 237-9

 3. Butler, J. C. and C. J. Peters, "Lessons Learned from the Hantaviruses and Other Hemorrhagic Fever Viruses," The American Journal of the Medical Sciences, 1996, 311(1): 55-59

 4. CDC, "Outbreak of Acute Illness--Southwestern United States, 1993," MMWR, June 11, 1993, 42 (22): 421-424

 5. CDC, "Update: Outbreak of Hantavirus Infection--Southwestern United States 1993," MMWR June 18, 1993 42 (23): 441-443

 6. CDC, "Update: Outbreak of Hantavirus Infection--Southwestern United States 1993," MMWR, June 25, 1993, 42 (24): 477-479

 7. CDC, "Update Hantavirus Disease--Southwestern United States, 1993," MMWR, July 30, 1993, 42 (29): 570-571

 8. CDC, "Update: Hantavirus Disease--United States, 1993," MMWR, Aug. 13, 1993, 42 (31): 612-614

 9. CDC, "Hantavirus Pulmonary Syndrome--Northeastern United States, 1994", MMWR, Aug. 5, 1994, 43 (30): 553-55

 10. CDC, "Hantavirus Pulmonary Syndrome--Virginia 1993," MMWR Dec. 2, 1994, 43 (47): 876-77

 11. CDC, "Hantavirus Pulmonary Syndrome--United States, 1995 and 1996," Journal of American Medical Association, 1996, 275 (18): 1395-1397

 12. Childs, et al. "Serologic and Genetic Identification of Permyscus maniculatus as the Primary Rodent Reservoir for a New Hantavirus in the Southwestern United States," The Journal of Infectious Diseases, 1994; 169: 1271-80

 13. Childs, et al. "A Household-Based Case-Control Study of Environmental Factors Associated with Hantavirus Pulmonary Syndrome in the Southwestern United States," American Journal of Medicine and Hygiene, 1995; 52(5) : 393-397

 14. Duchin, J. et al., "Hantavirus Pulmonary Syndrome: A Clinical Description of 17 Patients with a Newly Recognized Disease," The New England Journal of Medicine, 1994, 330 (14): 949-55

 15. Elliott, et al. "Isolation of the Causative Agent of Hantavirus Pulmonary Syndrome," American Journal of Medicine and Hygiene, 1994; 51(1): 102-108

 16. Hjelle, B. et al, "A Novel Hantavirus Associated with an Outbreak of Fatal Respiratory Disease in the Southwestern United States: Evolutionary Relationships to Known Hantaviruses," Journal of Virology, 1994, 68 (2): 592- 596

 17. Hjelle, B. et al., "Emergence of Hantaviral Disease in the Southwestern United States," Western Journal of Medicine, 1994 161 (5): 467-473

 18. Hjelle, B. et al. "Epidemiological Linkage of Rodent and Human Hantavirus Genomic Sequences in Case Investigations of Hantavirus Pulmonary Syndrome," The Journal of Infectious Disease, 1996, 173: 781-6

 19. Jay, M. et al., "Occupational Exposure Leading to Hantavirus Pulmonary Syndrome in a Utility Company Employee," Clinical Infectious Disease, 1996, 22: 841-4

 20. Khan, A. et. al, "Hantavirus Pulmonary Syndrome: The First 100 US Cases," The Journal of Infectious Diseases, 1996; 173: 1297-303

 21. Khan, A. et al. "Fatal Hantavirus Pulmonary Syndrome in an Adolescent," Pediatrics, 1995, 95 (2): 276-79

 22. "Mild Hantaviral Disease Caused by Sin Nombre Virus in a Four-Year-Old Child," The Pediatric Infectious Disease Journal, 1995, 14 (12): 1108-1110

 23. Morrison and Rathbun, "Hantavirus Pulmonary Syndrome: The Four Corners Disease," The Annals of Pharmacotherapy, 1995; 29: 57-63

 24. Morse, S., "Hantaviruses and Hantavirus Outbreak in the United States: A Case Study in Disease Emergence," Annals of New York Academy of Sciences, 1994 199-204

 25. Nolte, K. et al, "Hantaviral Biosafety Issues in the Autopsy Room and Laboratory: Concerns and Recommendations," Human Pathology, 1996, 27 (12): 1253-54

 26. Otteson, E. et al. "Occurrence of Hantavirus Within the Rodent Populatio of Northeastern California and Nevada," American Journal of Tropical Medicine and Hygiene, 1996, 54 (2): 127-33

 27. Simonsen, et.al "Evaluation of the Magnitude of the 1993 Hantavirus Outbreak in the Southwestern United States," The Journal of Infectious Diseases, 1995; 127: 1729-33

 28. Sinnott, J. et al, "Hantavirus: An Old Bug Learns New Tricks," Infection Control and Hospital Epidemiology, 1993, 14 (11): 661-664

 29. Vitek, C. et al, "Evidence Against Infection with Hantaviruses Among Forest and Park Workers in the Southwestern United States," Clinical Infectious Disease, 1996, 23: 283-5

 30. Vitek, C. et al., "Evidence Against Person-to-Person Transmission of Hantavirus to Health Care Workers,"Clinical Infectious Disease, 1996, 22: 824-6

 31. Weigler, B. "Zoonotic Hantaviruses: New Concerns for the United States," Journal of American Veterinary Medical Association, 1995, 206 (7): 979-986

 32. Zaki, et al. "Hantavirus Pulmonary Syndrome: Pathogenesis of an Emerging Infectious Disease," American Journal of Pathology, 1995; 146 (3): 552-556

 33. Zeitz, et al. "A Case-Control Study of Hantavirus Pulmonary Syndrome during an Outbreak in the Southwestern United States," The Journal of Infectious Diseases, 1995; 171: 864-70