Skip Navigation

Vibrio parahaemolyticus Infections in the United States, 1973–1998

  1. Nicholas A. Daniels1,3,5,
  2. Linda MacKinnon2,
  3. Richard Bishop2,
  4. Sean Altekruse4,
  5. Beverly Ray6,
  6. Roberta M. Hammond7,
  7. Sharon Thompson8,
  8. Susan Wilson9,
  9. Nancy H. Bean2,
  10. Patricia M. Griffin1 and
  11. Laurence Slutsker1
  1. 1Foodborne and Diarrheal Diseases Branch, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia
  2. 2Biostatistics and Information Management Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia
  3. 3Epidemic Intelligence Service, Epidemiology Program Office, Atlanta, Georgia
  4. 4Food and Drug Administration, Center for Food Safety and Applied Nutrition, Atlanta, Georgia
  5. 5Division of General Internal Medicine, Department of Medicine, University of California, San Francisco
  6. 6Texas Department of Health, Austin
  7. 7Florida Department of Health, Tallahassee
  8. 8Alabama Department of Public Health, Montgomery
  9. 9Louisiana Department of Health, New Orleans
  1. Reprints or correspondence: Dr. Nicholas A. Daniels, University of California, San Francisco, Department of Medicine, Division of General Internal Medicine, 1701 Divisadero St., Ste. 500, San Francisco, CA 94115 (ndaniels{at}medicine.ucsf.edu).
  1. Presented in part: 39th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, California, 26–29 September 1999.

Abstract

Vibrio parahaemolyticus infections are associated with consumption of raw or undercooked shellfish, contaminated food, and exposure of wounds to warm seawater. Foodborne outbreaks and sporadic infections from Vibrio species in 4 Gulf Coast states are reported routinely to the Centers for Disease Control and Prevention (CDC). Between 1988 and 1997, 345 sporadic V. parahaemolyticus infections were reported: 59% were gastroenteritis, 34% were wound infections, 5% were septicemia, and 2% were from other exposures. Forty-five percent of patients suffering from these conditions were hospitalized for their infections, and 88% of persons with acute gastroenteritis reported having eaten raw oysters during the week before their illness occurred. Between 1973 and 1998, 40 outbreaks of V. parahaemolyticus infections were reported to the CDC, and these outbreaks included >1000 illnesses. Most of these outbreaks occurred during the warmer months and were attributed to seafood, particularly shellfish. The median attack rate among persons who consumed the implicated seafood was 56%. To prevent V. parahaemolyticus infections, persons should avoid consumption of raw or undercooked shellfish and exposure of wounds to seawater.

Vibrio parahaemolyticus is a gram-negative, halophilic bacterium that naturally inhabits marine and estuarine environments and causes 3 major syndromes of clinical illness—gastroenteritis (the most common syndrome), wound infections, and septicemia [1]. V. parahaemolyticus was first identified as a cause of foodborne illness in Japan in 1950. At that time an outbreak investigation confirmed that infection was associated with eating sardines; 272 persons became ill, and 20 died [2]. Since then, V. parahaemolyticus has been recognized as a common cause of foodborne illness in Japan and throughout Asia. During the past 10 years in the United States, V. parahaemolyticus has been the most common Vibrio species isolated from humans, as well as the most frequent cause of Vibrio-associated gastroenteritis [3]. Recent V. parahaemolyticus outbreaks in the United States have been associated with consumption of raw or undercooked shellfish [46]. Despite an increasing number of Vibrio infections, V. parahaemolyticus and other noncholera Vibrio infections have not been reported in most states.

In this article, we describe the epidemiology of V. parahaemolyticus infections in the United States from 1973 to 1998, including data on sporadic infections and foodborne outbreaks reported to the US Centers for Disease Control and Prevention (CDC). These epidemiologic data, along with environmental factors that may have contributed to the occurrence of recent V. parahaemolyticus outbreaks, suggest potential control measures to reduce the burden of illness caused by this organism in the United States.

Methods

Since 1988, the CDC has collected epidemiologic and clinical information about sporadic V. parahaemolyticus infections in states that participate in the Gulf Coast Vibrio Surveillance System. These states include Florida, Alabama, Louisiana, and Texas. Sporadic infections are reported voluntarily to the CDC on a standard reporting form completed by state or local health departments after they have finished the investigation of a culture-confirmed illness. The surveillance report includes data on demographics (e.g., age, sex, race/ethnicity, and occupation of victims); data on the sites of Vibrio infection; clinical information (e.g., symptoms, preexisting medical conditions, duration of illness, sequelae of infection, and patient clinical outcomes); and epidemiologic information (e.g., recent travel experiences, seafood exposure, and contact with seawater).

To obtain clinical and epidemiological information (including food history) on V. parahaemolyticus infections, we extracted data for 1988–1997 from the Vibrio database that is maintained at the CDC. For analysis of data, wound infections were considered to be the source of infection when a patient incurred a wound before or during exposure to seawater or seafood drippings and when a culture of the wound, blood, or of a normally sterile site yielded V. parahaemolyticus. Primary septicemia was defined as a systemic illness characterized by fever or shock (<90 mm Hg systolic blood pressure), in which V. parahaemolyticus was isolated from either the blood or a normally sterile site (specifically when no wound infection preceded the illness). Gastroenteritis was defined as an illness with attendant diarrhea, vomiting, or abdominal cramps; no evidence of wound infection; and isolation of V. parahaemolyticus from stool samples alone.

Since 1973, the CDC has received reports of foodborne disease outbreaks, including those caused by Vibrio species, from state and local health departments. A V. parahaemolyticus outbreak was defined as an incident in which ⩾2 persons experienced a similar illness resulting from the ingestion of a common food and in which ⩾2 persons had laboratory-confirmed V. parahaemolyticus infection or as an incident in which ⩾105 Kanagawa-positive V. parahaemolyticus organisms per gram were isolated from epidemiologically implicated food. We reviewed V. parahaemolyticus outbreaks that occurred between 1 January 1973 and 31 December 1998 and that were reported on standard investigation forms through the foodborne disease—outbreak surveillance system.

Results

Sporadic disease

Between 1988 and 1997, a total of 345 cases of V. parahaemolyticus infection were reported to the CDC by the states participating in the Gulf Coast Vibrio Surveillance System. Cases were reported from Florida, Alabama, Louisiana, and Texas. Of the 345 patients whose cases were reported to the CDC, 202 (59%) had gastroenteritis, 118 (34%) had wound infections, and 17 (5%) had septicemia. Eight other infections were reported, including ear, eye, urinary tract, and peritoneal infections. Ear infections developed after swimming in the Gulf of Mexico, an eye infection occurred after a penetrating corneal injury and exposure to warm seawater, and the peritoneal infection was diagnosed after surgery for acute appendicitis. Two hundred twenty-one (64%) of the reported infections occurred in males. Patients' ages ranged from 1 month to 93 years (median, 36 years). Between 1988 and 1997 in the Gulf Coast states, a median of 34 infections occurred per year (range, 26–41). Among the 88 patients with V. parahaemolyticus gastroenteritis and known food history, 77 (88%) reported having eaten raw oysters in the week before their illness occurred. Among the 11 patients with septicemia and known food history, 10 (91%) had eaten raw oysters.

Overall, 119 (34%) persons reported having a preexisting illness, and 156 (45%) persons were hospitalized. Of the 301 patients for whom information on survival was available, 12 (4%) died as a result of their infections. Of the 12 patients who died, 5 (29%) had primary septicemia, 4 (2%) had gastroenteritis, and 3 (3%) had wound infection. Of the 9 patients who died as a result of severe gastroenteritis or septicemia, food histories existed for 5 of them; all of them had eaten raw oysters during the week before the onset of illness. Of the 12 patients who died, 10 (83%) had a known preexisting medical condition; these conditions included alcoholism, liver disease, renal disease, vascular disease, and diabetes. The clinical characteristics of patients with V. parahaemolyticus infection by clinical syndrome are summarized in table 1. Patients with primary septicemia (10 [83%] of 12 patients) were more likely than those with gastroenteritis (11 [19%] of 58 patients) to have a known history of alcoholism or liver disease. All syndromes of V. parahaemolyticus infection were more common in the warmer months (figure 1); 94% of cases occurred between April and October.

Figure 1

Vibrio parahaemolyticus infections in the United States, as reported to the Centers for Disease Control and Prevention, by month of culture date, 1988–1997.

Figure 2

Vibrio parahaemolyticus outbreaks in the United States, by month of occurrence, 1973–1998.

Figure 3

Vibrio parahaemolyticus outbreaks in the United States, by region, 1973–1998.

Table 1

Characteristics of 337 patients with Vibrioparahaemolyticus infections, by clinical syndrome, 1988–1997.

Foodborne outbreaks

Between 1973 and 1998, the CDC received reports of 40 outbreaks of V. parahaemolyticus infection in 15 states and the Guam Territories; these outbreaks resulted in 1064 illnesses (table 2). The majority of these outbreaks occurred during the warmer months (figure 2), with 80% occurring between April and October; the median month of occurrence was July. Outbreaks were reported along all coastal areas in the United States and the Guam Territories. The range for attack rates was 3%–100%, with a median of 56%. The median incubation period was 17 h (range, 4–90 h). The median number of ill persons involved in these outbreaks was 5 (range, 2–296 persons). Diarrhea was the most common symptom and often was associated with abdominal cramps, nausea, and vomiting. The median reported duration of illness was 2.4 days (range, 8 h to 12 days). One death was reported. The food vehicle in all the outbreaks was seafood or cross-contamination with seafood, particularly raw or undercooked shellfish. Seafood was eaten raw in 15 (38%) of the outbreaks.

Table 2

Epidemiological characteristics of outbreaks of Vibrio parahaemolyticus infection reported to the Centers for Disease Control and Prevention, United States, 1973–1998.

Twelve (30%) of the 40 V. parahaemolyticus outbreaks were reported in 1997 and 1998, which may suggest a resurgence of this pathogen (figure 3). Three recent large multistate outbreaks occurred in 1997 and 1998. During July–August 1997, an outbreak linked to consumption of raw or undercooked shellfish harvested from waters off the coasts of California, Oregon, Washington, and British Columbia resulted in >200 illnesses in 7 states and Canada [4]. The most common V. parahaemolyticus serotypes isolated from patients involved in this outbreak were O4:K12 and O1:K56. During May—July 1998, 416 persons in 13 states reported having gastroenteritis after eating oysters harvested from Galveston Bay, Texas (these patients included those who reported 296 cases of V. parahaemolyticus infection in Texas and 120 cases of infection from other states). All patient isolates were V. parahaemolyticus serotype O3:K6, which commonly causes outbreaks in Asia but had not been identified previously in the United States. Later in 1998, another multistate outbreak of V. parahaemolyticus O3:K6 infection occurred; 23 culture-confirmed cases of infection were reported from New York, New Jersey, and Connecticut. Infection was linked to consumption of raw oysters and clams harvested from Oyster Bay, Long Island, New York [5].

Discussion

Our clinical and epidemiologic review of V. parahaemolyticus infections and foodborne outbreaks demonstrates that this infection causes significant morbidity in the United States and occasionally can be fatal. Nearly 30% of reported patients with gastroenteritis had bloody stools, which suggests that only the most severe cases came to the attention of medical personnel. The high case fatality rate (4%) of sporadic infections and the high attack rate (56%) during outbreaks may also suggest a reporting bias toward severe cases. Most sporadic infections and outbreaks were linked to consumption of contaminated, raw molluscan shellfish. During the past several years, the number of reported outbreaks has increased steadily; there was a sharp rise after 1997. A possible explanation for this observed increase may be that warmer sea temperatures affect the emergence of more-virulent serotypes in oyster-harvesting areas.

The occurrence of outbreaks that result from consumption of contaminated seafood, particularly shellfish harvested between February and December from the waters of all coastal regions of the United States and the Guam Territories, is consistent with the ecology of V. parahaemolyticus. The seasonality of infections, which occurred mainly during the warmer months, suggests that water temperature may be an important factor in the epidemiology of Vibrio infection. Water temperatures have been shown to influence the growth of V. parahaemolyticus [711]. Interestingly, 1997 and 1998 were El Niño years that resulted in warmer sea temperatures in coastal areas. These elevated temperatures have been shown to influence the incidence of V. cholerae [12] and may also explain the resurgence of V. parahaemolyticus during these years. Although we did not have information on the site of oyster harvest and on the water temperature at time of harvest, one recent CDC study showed that 89% of oysters that made persons ill with V. vulnificus were harvested in waters with a temperature >22°C [13]. The emergence of more-virulent serotypes in oyster-harvesting areas also could have contributed to an increase in V. parahaemolyticus outbreaks in the United States from 1997 to 1998. Several strains of V. parahaemolyticus (e.g., O3:K6, O4:K12, and O1: K56) have been associated with this recent upsurge. Whether warmer seawater temperature influences strain selection is unknown.

More than 95% of V. parahaemolyticus strains isolated from stool cultures of symptomatic persons are virulent, as determined by a positive Kanagawa reaction [14]. In contrast, ⩽1% of strains from the environment, including strains isolated from seafood implicated during outbreaks, have a positive Kanagawa reaction [1416]. The reason for this discrepancy in virulence between human and environmental isolates is unknown. Nevertheless, it suggests that environmental sampling of oyster beds to detect the presence of V. parahaemolyticus may not be a useful public health measure. In a volunteer feeding study, the infective dose of V. parahaemolyticus was determined to be between 105 and 107 organisms [17]. Participants were healthy young adults who were given antacid tablets before being fed V. parahaemolyticus. Recent outbreaks in the United States indicate that the infective dose may be much lower [4, 5].

V. parahaemolyticus in stool specimens is readily identified on selective media such as thiosulfate-citrate-bile salts (TCBS) agar; it usually appears as a blue-green colony [18], and identification is accomplished by using biochemical tests. In 1997, a survey of US clinical laboratories in FoodNet sites in California, Connecticut, Georgia, Minnesota, and Oregon found that only 31% of laboratories routinely cultured stool specimens specifically for Vibrio by using TCBS agar [19]. In a random survey of Gulf Coast clinical laboratories, we found that only 20% of laboratories used TCBS routinely to culture stool specimens (CDC, unpublished data). Blood agar and other non-selective media may also support the growth of Vibrio; however, detection is more difficult than with TCBS. Many laboratories are now using automated biochemical machines, such as Vitek (Vitek Systems, Hazelwood, MO), to identify microorganisms. These automated machines have not been validated for identification of Vibrio species. Misclassification of Vibrio with other enteric pathogens may occur, particularly in the cases of Aeromonas, Pleisiomonas, and Pseudomonas. Therefore, all Vibrio isolations should be confirmed using TCBS and standard biochemical tests at state public health laboratories.

Recent outbreaks have highlighted the need to reevaluate regulations and policies concerning the safety of raw molluscan shellfish, and these outbreaks indicate that current policy and regulations for oyster harvesting do not protect the approximately 1 in 10 American adults who occasionally eat raw shellfish (CDC FoodNet, unpublished data). During recent outbreaks, the implicated oyster beds met current bacteriologic standards for fecal coliforms, which were established by the National Shellfish Sanitation Program [20], and the median enumerated V. parahaemolyticus counts were less than the regulatory threshold of 10,000 cfu/g [4, 5]. Molluscan shellfish harvesting is regulated by individual states. In most states, oyster beds are closed by state shellfish authorities when human illness is traced to harvest sites. Because surveillance for V. parahaemolyticus infections detects only a fraction of cases of human illness, many cases may occur before an outbreak is detected; 1 culture-confirmed case may represent a larger outbreak. Therefore, in addition to closing oyster beds after trace-backs from outbreaks of human illness, prevention strategies might include (1) monitoring oysters, to identify beds in which Vibrio counts are elevated; (2) identifying and implementing processing technologies to reduce Vibrio counts in oysters that are sold for raw consumption; (3) banning harvesting of oysters during the warmer months (e.g., months without the letter “R” in their names), when seawater temperatures and Vibrio counts are usually elevated; or (4) diverting oysters harvested during the warmer months for cooking, irradiation, or pasteurization. Development of environmental parameters (e.g., seawater temperatures and salinity levels) that could more precisely predict the risk of V. parahaemolyticus infection could be a useful additional way to enhance the safety of raw molluscan shellfish.

Prompt recognition of outbreaks by clinicians, clinical laboratories, and public health authorities is important for implementing control measures, such as closing implicated shellfish harvest sites to prevent further illness. During outbreaks, V. parahaemolyticus isolates should be referred to public health laboratories for confirmation and strain subtyping. Pulsed-field gel electrophoresis is more discriminatory than serotyping [21] and may be useful in linking common source outbreaks. Clinical laboratories in coastal areas are encouraged to use TCBS agar when culturing stool specimens, particularly during the summer months.

To prevent gastroenteritis attributable to V. parahaemolyticus, all consumers should avoid eating raw or undercooked molluscan shellfish. In particular, persons with liver disease should be counseled to avoid eating raw or undercooked molluscan shellfish, since they are at particularly high risk for V. parahaemolyticus and other severe Vibrio infections [22]. Cooking oysters and other shellfish could prevent illness by killing Vibrio. To prevent wound infections, people should avoid exposure of open wounds to seawater or raw shellfish products.

Clinicians should insist that a stool specimen be tested for Vibrio when patients seek medical attention for acute gastroenteritis within 48 h of ingesting raw or undercooked shellfish. Treatment of gastroenteritis with anything other than oral rehydration is seldom necessary, since gastroenteritis is usually mild and self-limited, although antimicrobial therapy may help those patients with severe diarrhea, wound infections, or septicemia. Antimicrobials effective against Vibrio infections include tetracycline, fluoroquinolones, third-generation cephalosporins, and aminoglycosides.

State shellfish programs should establish a protocol for closing and reopening oyster beds, and they should require regular monitoring of environmental factors, such as seawater temperature and salinity levels at oyster-harvest sites, particularly environmental monitoring during the warmer months and during periods when the number of cases of human illness increases. Data from improved monitoring should be linked to human-illness data, to help develop better criteria for the closing and reopening of oyster beds. Because of the recent increase in multistate outbreaks, all states should consider mandating that infections involving V. parahaemolyticus and other Vibrio species be reported, as they are in the 4 Gulf Coast states.

Acknowledgments

We thank the state epidemiologists, the directors of the state public health laboratories, and other state and local health department personnel who assisted with outbreak investigations and who reported results to the Centers for Disease Control and Prevention. We also thank Mary Evans.

Footnotes

  • This work is in compliance with the human experimentation guidelines of the US Department of Health and Human Services.

  • State epidemiologists can obtain copies of the Centers for Disease Control and Prevention's (CDC) “Cholera and other Vibrio Illness Surveillance Reporting Form” from the CDC, Foodborne and Diarrheal Diseases Branch, Mailstop A38, Atlanta, GA 30333.

  • Received October 28, 1999.
  • Revision received January 7, 2000.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
| Table of Contents