Skip Navigation

Characteristics of O157 versus Non-O157 Shiga Toxin-Producing Escherichia coli Infections in Minnesota, 2000–2006

  1. Erin B. Hedican1,
  2. Carlota Medus1,
  3. John M. Besser2,
  4. Billie A. Juni2,
  5. Bonnie Koziol2,
  6. Charlott Taylor2, and
  7. Kirk E. Smith1
  1. 1Acute Disease Investigation and Control Section and Public Health Laboratory, Minnesota Department of Health, St. Paul
  2. 2Public Health Laboratory, Minnesota Department of Health, St. Paul
  1. Reprints or correspondence: Erin Hedican, Minnesota Dept. of Health, 625 Robert St. N, PO Box 64975, St. Paul, MN 55164-0975 (erin.hedican{at}state.mn.us).
 
Next Section

Abstract

Background. Escherichia coli O157:H7 (O157) is the Shiga toxin-producing E. coli (STEC) serotype most frequently isolated and most often associated with hemolytic uremic syndrome (HUS) in the United States. Non-O157 STEC serotypes can also cause serious illness, but their impact as pathogens remains undefined. We compared characteristics of non-O157 and O157 STEC infections identified through sentinel surveillance.

Methods. Sentinel sites included a metropolitan health maintenance organization laboratory and a hospital laboratory serving a small city and rural area. We received sorbitol-MacConkey agar plates from every stool culture performed at both sites during 2000–2006. Colony sweeps were screened for stx1 and stx2 by polymerase chain reaction. E. coli identity, serotype, and presence of stx1 and/or stx2 were confirmed on individual isolates.

Results. Two hundred six STEC isolates were identified: 108 (52%) were non-O157 serotypes, and 98 (48%) were O157. Of non-O157 cases, 54% involved bloody diarrhea, and 8% involved hospitalization. Non-O157 isolates with at least stx2 were not more likely to cause severe illness (bloody diarrhea, hospitalization, or HUS) than were non-O157 isolates with only stx1. O157 cases were more likely than non-O157 cases to involve bloody diarrhea (78% vs 54%; P<.001), hospitalization (34% vs 8%; P<.001), and HUS (7% vs 0%; P=.005). When including only isolates with at least stx2, O157 cases were still more likely to involve bloody diarrhea (78% vs 56%; P=.02) and hospitalization (33% vs 12%; P=.01) than non-O157 cases.

Conclusions. Differences in severity among STEC infections could not be explained by stx2, suggesting that additional factors are important in STEC virulence.

Shiga toxin-producing Escherichia coli (STEC) is a group of pathogenic E. coli that cause diarrhea, bloody diarrhea, and hemorrhagic colitis. Hemolytic uremic syndrome (HUS) develops in 5%-10% of individuals with STEC-associated diarrhea and has a case-fatality rate of 3%-7% [1].

Cattle are a reservoir of STEC, and foods originating from or contaminated by cattle are important vehicles for human disease. Other routes of infection include consumption of contaminated water, direct contact with infected ruminant animals, and person-to-person transmission [1–3].

More than 200 E. coli serotypes produce Shiga toxin, and >100 of these have been linked with human illness [1]. E. coli O157 (O157) is the serotype most frequently isolated and most often associated with HUS in the United States. Non-O157 STEC (non-O157) serotypes can also cause serious illness and have been implicated in outbreaks in the United States and elsewhere [4–8].

Variability exists between non-O157 STEC serotypes in their associations with outbreaks and disease severity. Differences in virulence factors produced by different strains likely help explain this variability [9, 10]. Shiga toxin has been considered the primary virulence factor responsible for causing severe illness, including bloody diarrhea and HUS, but not all STEC infections lead to these conditions. This suggests that other determinants, in addition to host factors, are involved [9]. Shiga toxins are classified into 2 main categories—Shiga toxin 1 (Stx1) and Shiga toxin 2 (Stx2)—and each group contains variants. Previous studies have found that STEC strains producing Stx2 are more likely to cause HUS than are strains that produce Stx1 alone [11–13].

Detection of O157 in stool specimens relies mainly on its inability to rapidly ferment sorbitol; therefore, sorbitol-MacConkey (SMAC) agar is used to culture O157. However, most non-O157 STEC do ferment sorbitol and cannot be differentiated from non-pathogenic E. coli strains on SMAC agar. Therefore, detection of non-O157 STEC involves either testing directly for Shiga toxins or the genes that encode them [1, 3].

Because non-O157 STEC are not routinely detected through standard stool culture methods in clinical laboratories, their impact as pathogens remains largely undefined. We used sentinel surveillance in Minnesota to determine the burden of non-O157 infections relative to O157, and to compare clinical and epidemiological characteristics of O157 and non-O157 STEC infections identified through the sentinel sites.

Previous SectionNext Section

Methods

Sentinel surveillance sites. Data were collected through 2 sentinel surveillance sites in Minnesota during 2000–2006. One was the laboratory for a large Minneapolis-St. Paul metropolitan area health maintenance organization; the other was a hospital laboratory that serves a small city and surrounding rural area that is rich in food animal agriculture, particularly dairy production. The estimated maximum population sizes served by these systems were 656,000 (urban site) and 643,000 (rural site). All stool samples submitted to the 2 laboratories for enteric bacterial culture were plated on SMAC enteric stool culture plates; after completion of testing, the SMAC plates were sent to the Minnesota Department of Health Public Health Laboratory (St. Paul).

Laboratory methods. Upon receiving a SMAC plate from a sentinel laboratory, template DNA was prepared from colony sweeps. Six sweeps were made through representative areas of growth using a 1.0-µL disposable loop. Sweeps included all visible colony morphologies and avoided the primary inoculation area. Individual sweeps from each primary specimen were mixed. One loopful of the mixed sweep material and 200 µL of molecular grade water (Sigma) were heated for 15 min in boiling water and centrifuged at 16,000 g for 2 min. Clear supernatants containing bacterial DNA were withdrawn for polymerase chain reaction analysis.

From January 2000 through July 2005, Shiga toxin genes stx1 and stx2 were detected using previously described primers and amplification methods [14]. In July 2005, a different method [15] was implemented to increase Shiga toxin gene detection sensitivity.

Up to 24 individual colonies from samples that initially tested positive for stx1 or stx2 were retested by polymerase chain reaction. Shiga toxin gene-positive isolates were identified by standard biochemical methods [16]. Somatic and flagellar antigens were determined using Denka Seiken antisera. Isolate identity was further confirmed at the Centers for Disease Control and Prevention (Atlanta, GA). If no individual colony was found to contain Shiga toxin gene sequences, the sample was classified as a polymerase chain reaction-positive STEC of unknown serotype. If an individual E. coli colony was Shiga toxin gene-positive and tested negative for O157, it was classified as a non-O157 STEC even if the serogroup could not be determined.

Data collection. If STEC was detected, the submitting clinic provided case demographic and clinical information. Interviews were attempted for all STEC cases to collect additional demographic, symptom, exposure, and treatment information. Similar demographic and treatment information was collected for HUS cases reported to MDH. Only patients with HUS for whom STEC was confirmed through the sentinel surveillance during 2000–2006 were included in this investigation.

Statistical analyses. Comparisons were made between non-O157 and O157 STEC cases in Minnesota residents at the 2 sentinel sites. Cases were excluded if the serogroup of the identified STEC was undetermined, unless an E. coli isolate was Shiga toxin gene positive and tested negative for O157; in that instance, the corresponding case was classified as a non-O157 STEC and included. Cases for which samples yielded another gastrointestinal pathogen were excluded because of the potential impact on illness severity. Secondary and outbreak-associated cases were included for clinical comparisons but excluded for risk factor analyses.

Results were obtained using SAS statistical software, version 9.1 (SAS Institute). The χ2test for linear trend was used to analyze Shiga toxin type changes over time. Categorical variables were analyzed using Fisher's exact test. The Kruskal-Wallis test was used to compare continuous variables. Multivariate analysis using logistic regression was performed following a stepwise selection process. Two-sided P values ⩽.05 were considered to be statistically significant.

Previous SectionNext Section

Results

There were 302 Shiga toxin gene-positive SMAC plates identified from the 28,380 plates tested from the sentinel sites during 2000–2006 (median number of plates tested per year, 3984; range, 3612–4842). Of these, Shiga toxin-producing E. coli colonies were isolated for 226 (75%); therefore, for the other 76 (25%), the serogroup was unknown. Among the 226 isolated STEC, non-O157 serogroups accounted for 119 (53%), O157 accounted for 100 (44%), and the serogroup was defined as “rough” for 7 (3%).

The number and proportion of STEC cases identified each year displayed a downward trend during 2000–2004, but then increased from 2004–2006 (2000, 53 cases [1.2% of plates submitted]; 2001, 53 cases [1.3%]; 2002, 36 cases [0.9%]; 2003, 25 cases [0.7%]; 2004, 22 cases [0.6%]; 2005, 50 cases [1.3%]; and 2006, 63 cases [1.3%]). Most STEC samples were collected during July-October (183 samples [61%]). Both the O157 and non-O157 serogroups exhibited a similar summer/early fall seasonality (figure 1).

Figure 1
View larger version:
Figure 1

Number of Shiga toxin-producing Escherichia coli cases identified through sentinel surveillance in Minnesota, by serogroup and month of specimen collection, 2000–2006.

Twenty-two (7%) of the 302 patients with STEC were infected with another gastrointestinal pathogen and were excluded from further analyses. Of these, 8 had enterotoxigenic E. coli infection, 7 had Campylobacter infection, 3 had Cryptosporidium infection, 3 had Giardia infection, and 1 had Salmonella infection. Remaining isolates for which the serogroup was unknown or rough (n=74) were also excluded.

Of the remaining 206 STEC cases, 108 (52%) involved non-O157 serotypes, and 98 (48%) involved O157. The urban site yielded 28 O157 and 54 non-O157 STEC isolates, whereas the rural site yielded 70 O157 and 54 non-O157 STEC isolates.

Five serogroups represented 74% of the 108 non-O157 isolates: O26 (29 isolates [27%]), O103 (23 [21%]), O111 (20 [19%]), O145 (5 [5%]), and O45 (3 [4%]). The remaining serogroups included O1, O21, O22, O51, O71, O76, O80, O88, O91, O119, O121, O128, O137, O159, O165, O166, and O175. The serogroup could not be determined for 6 non-O157 isolates (6%).

STEC case demographic characteristics. Interviews were conducted for ∼90% of patients with non-O157 and O157 infection. Individuals infected with non-O157 and O157 STEC were similar with regard to age, sex, race, ethnicity, and antibiotic treatment status (defined as treatment with any antibiotic). The median age for all patients with STEC infection was 15 years (range, 15 days to 88 years), and 118 (57%) individuals were female. Race information was available for 186 patients (90%); 177 (95%) were white, 6 (3%) were black, 2 (1%) were Asian/Pacific Islander, and 1 (1%) reported “other.” Ethnicity was reported for 182 patients (88%); 177 (97%) were non-Hispanic. Time from illness onset to specimen collection and from onset to interview was significantly longer for patients with non-O157 cases than for those with O157 cases (median, 4 vs 3 days [P=.006] and 45 vs 14 days [P<.001], respectively).

Risk factor analysis. In the 7 days before illness onset, there was no significant difference between O157 and non-O157 groups with regard to the proportion of patients who consumed unpasteurized milk; drank untreated or “raw” water; reported swimming in a lake, river, ocean, or pool; consumed meat from a place other than a grocery store (eg, a butcher shop or by private slaughter); lived on a farm; visited a farm or petting zoo; or ate at a restaurant (table 1). Patients with non-O157 infection were more likely to have travelled internationally (12% vs 1%; P=.01), and those with O157 infection were more likely to have consumed ground beef (76% vs 58%; P=.03) (table 1). In the rural population, patients with O157 and non-O157 infection were similar with regard to all exposures (data not shown). In the urban population, patients with non-O157 infection were more likely to have consumed raw water than were those with O157 infection (24% vs 0%; P=.02) before illness onset.

Figure 2
View larger version:
Figure 2

Type of Shiga toxin produced by O157 and non-O157 Shiga toxin-producing Escherichia coli isolates, by year of specimen collection, 2000–2006. Stx1, Shiga toxin 1; Stx2, Shiga toxin 2.

Table 1
View larger version:
Table 1

Comparison of Potential Exposures for O157 and Non-O157 Shiga Toxin-Producing Escherichia coli Cases Identified in Minnesota through Sentinel Surveillance, 2000–2006.

Shiga toxin type. Among the 108 non-O157 STEC isolates, 65 (60%) had stx1 alone, 14 (13%) had stx2 alone, and 29 (27%) had both toxin genes; therefore, 43 non-O157 isolates (40%) had at least stx2. The proportion of non-O157 STEC samples that contained only stx1 increased significantly from 2000 to 2006, whereas the proportion that had both toxins decreased significantly (figure 2). Among the 98 O157 isolates, 1 (1%) had stx1 alone, 30 (31%) had stx2 alone, and 67 (68%) produced both toxins. There was no significant change in Shiga toxin gene type since 2000 for O157 isolates (figure 2). Non-O157 isolates were more likely to have stx1 alone than were O157 isolates (65 [60%] vs 1 [1%]; P<.001), whereas O157 isolates were more likely to have stx2 alone (30 [31%] vs. 4 [13%]; P=.03) or both toxins (65 [66%] vs 29 [27%]; P<.001).

Clinical comparison. Patients infected with any STEC (i.e., O157 and non-O157 combined) that had at least stx2 were more likely to develop bloody diarrhea (71% vs 53%; P=.03) and to be hospitalized (27% vs 8%; P=.001) than were those infected with any STEC that had stx1 alone, regardless of serogroup (table 2). However, for non-O157 STEC only, there were no significant differences in illness severity between cases involving isolates that had at least stx2 and those that had only stx1 (table 2). Patients infected with any STEC that had stx2 alone were not significantly different from patients infected with STEC that had stx1, alone or in combination with stx2, with regard to severity of illness (data not shown).

Table 2
View larger version:
Table 2

Clinical Outcomes in Shiga Toxin-Producing Escherichia coli Cases Identified in Minnesota through Sentinel Surveillance by Shiga Toxin Gene Type, 2000–2006.

Cases with O157 infection were more likely to involve bloody diarrhea (78% vs 54%; P<.001), hospitalization (34% vs 8%; P<.001), and HUS (7% vs 0%; P=.005), compared with cases of non-O157 infection (table 3). Five of the patients with HUS were 1–3 years old, 1 was 20 years old, and 1 was 87 years old. Patients with O157 infection reported having a greater maximum number of stools per 24-h period than did those with non-O157 infection (median, 15 vs 10; P<.001). Patients with non-O157 infection reported a longer duration of diarrhea than did those with O157 infection (median, 7 vs 6 days; P=.004) (table 3). However, only 56 patients with O157 infection (57%) had recovered when interviewed, compared with 81 (75%) of those with non-O157 infection.

Table 3
View larger version:
Table 3

Clinical Outcomes in Shiga Toxin-Producing Escherichia coli Cases Identified in Minnesota through Sentinel Surveillance by Serogroup, 2000–2006.

When we restricted the analysis to include only STEC isolates that had stx2, either alone or with stx1 (n=140), patients with O157 infection were still significantly more likely to develop bloody diarrhea (78% vs 56%; P=.02) and to be hospitalized (33% vs 12%; P=.01) than were those with non-O157 infection (table 3). Patients with O157 infection also reported having a greater maximum number of stools per 24-h period (median, 15 vs 10; P=.002) (table 3). When we included only isolates that had stx2 alone, patients with O157 infection (n=30) were significantly more likely than those with non-O157 infection (n=14) to develop bloody diarrhea (83% vs 42%; P=.02).

Demographic, serogroup, and Shiga toxin type were analyzed using logistic regression for the outcomes of bloody diarrhea and HUS. No factors were statistically associated with HUS in the multivariate analysis. The O157 serogroup was the only variable associated with bloody diarrhea (adjusted odds ratio, 3.1; 95% confidence interval, 1.5–6.4; P=.003).

Previous SectionNext Section

Discussion

Non-O157 STEC isolates were recovered from stool specimens obtained from ill patients slightly more frequently than O157 when results from the sentinel sites were combined. This suggests that non-O157 serotypes account for a substantial proportion of STEC infections in Minnesota and is consistent with other studies in the United States on the relative incidence of O157 versus non-O157 [17–21]. Three serogroups (O26, O103, and O111) accounted for 67% of the non-O157 isolates of known serogroup in this study. These serogroups are among the 6 most common serogroups in the United States [12]. Most non-O157 cases in Minnesota were identified during the summer months, following a seasonal trend consistent with that of O157 [12].

It was hypothesized that individuals in rural settings would be more likely to contact farm animals and settings than urban populations and, therefore, that STEC infection would occur more frequently in the rural population. The rural population accounted for 63% of STEC cases in this study. Non-O157 cases were identified with the same frequency in the urban and rural populations. However, the distribution of O157 cases differed markedly by site. The rural site accounted for 71% of all O157 cases in the study. O157 cases occurred more frequently than non-O157 cases in the rural population but were far less common than non-O157 cases in the urban population. This suggests that farm-related factors may be more important for O157 than non-O157 STEC. However, our risk factor analysis did not provide additional evidence for that. In fact, patients with non-O157 infection were similar to those with O157 infection with regard to exposure to several established O157 risk factors, including living on or visiting a farm, consumption of unpasteurized milk, and procuring meat through private slaughter [22, 23]. Non-O157 cases were significantly more likely to have travelled internationally. When stratifying the population by sentinel site, patients with non-O157 STEC infection from the urban site were significantly more likely to have consumed raw water. Conversely, a significantly higher proportion of patients with O157 reported consuming ground beef. These results suggest that whereas some risk factors for O157 and non-O157 STEC infection are the same, the epidemiology of these 2 groups may differ in important ways. Future studies to evaluate risk factors for non-O157 STEC infection independently are necessary.

Non-O157 infections caused substantial morbidity in our study cases; 54% of patients reported bloody diarrhea, 8% were hospitalized, and the median duration of diarrhea was 7 days. However, patients with O157 infection were significantly more likely to develop severe illness, as measured by bloody diarrhea, HUS, hospitalization, and maximum number of stools per 24-h period. To assess the impact of stx2 on the difference in illness severity between O157 and non-O157 STEC, stratified analyses were performed using only isolates that had stx2, either in combination with stx1 or alone. Patients infected with stx2 -containing O157 STEC were still more likely to develop bloody diarrhea and be hospitalized than were those who were infected with stx2 -containing non-O157 STEC. If stx2 is the primary STEC virulence determinant, as has been proposed, these 2 STEC categories should have been similar with regard to illness severity after controlling for this factor. Furthermore, there were no differences when we compared patients infected with non-O157 serotypes that had only stx1 with those for which the serotypes had at least stx2. If toxin type is a stronger determinant of severe illness than serogroup, differences should have been seen within non-O157 cases by toxin type as well.

In our investigation, stx1 and stx2 were the only virulence factors evaluated. However, other genes or gene variants have been proposed to be associated with severe clinical outcomes [9, 24]. Among these is the hypothesis that certain Shiga toxin variants are more likely to cause severe disease (ie, the specific variant rather than the broader toxin group may have a greater influence on the course of illness) [24–27]. Subtyping of Shiga toxin gene variants to assess the association between toxin subtypes and clinical outcomes is warranted, particularly for stx2. In addition, the influence of other putative virulence factors on the pathogenicity of STEC infections, including the additive effects of these factors, should be evaluated [27].

This study had some potentially important limitations. Because of detection methods, patients with O157 infection were interviewed significantly closer to onset than were those with non-O157 infection. Poor recall or recall bias may have been introduced in the interviews with non-O157-infected patients as a result of the delay, causing patients to inaccurately report illness details and/or exposure histories. Therefore, some of the differences observed between the non-O157 and O157 groups may have been influenced by differences in recall. Another limitation was the change in STEC laboratory detection methods at the Minnesota Department of Health Public Health Laboratory that occurred during 2005. The new STEC detection method likely had an increased sensitivity for stx1. Therefore, the increase in non-O157 STEC found to contain stx1 in the last 2 years of the study may have been influenced by enhanced testing methods and not represent real trends.

The results of this study have clinical implications. Testing of samples to identify the presence of stx2 versus stx1 may not be reliable in predicting whether the infecting STEC strain is capable of causing severe illness. Serogroup (O157 vs non-O157) was an important predictor of severity, but it is apparent that more-specific testing (ie, for stx2 subtypes or other virulence factors) is necessary to identify the specific virulence factors that contribute to severe disease.

In summary, non-O157 STEC infections were slightly more common than O157 infections in our study. Overall, O157 infections were more severe, but non-O157 infections caused substantial morbidity as well. Improvements in diagnostic capabilities will lead to increased detection of non-O157 STEC cases. This should lead to the identification of additional outbreaks and risk factors for this group of organisms. It also should result in a better understanding of the determinants associated with severe disease caused by STEC.

Previous SectionNext Section

Acknowledgments

We thank the Public Health Laboratory staff and Acute Disease Investigation and Control Section staff at the Minnesota Department of Health who participated in this study and Craig Hedberg and Rhonda Jones-Webb at the University of Minnesota School of Public Health for reviewing this work.

Financial support. The Centers for Disease Control and Prevention as part of the Emerging Infections Program, Foodborne Diseases Active Surveillance Network (FoodNet; cooperative agreement U50/CCU511190).

Potential conflicts of interest. All authors: no conflicts.

  • Received December 18, 2008.
  • Accepted March 22, 2009.
Previous Section
 

References

  1. ↵
    1. Blaser MJ,
    2. Smith PD,
    3. Ravdin JI,
    4. et al.
    1. Griffin P,
    2. Mead P,
    3. Sivapalasingam S
    . Escherichia coli O157:H7 and other enterohemorrhagicE. coli. In: Blaser MJ, Smith PD, Ravdin JI, et al., editors. Infections of the gastrointestinal tract2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 627-42.
    1. Duffy G,
    2. Garvey PDM
    1. Blanco J,
    2. Blanco M,
    3. Blanco J,
    4. et al
    . Epidemiology of verocytotoxigenic Escherichia coli (VTEC) in ruminants. In: Duffy G, Garvey PDM, editors. Verocytotoxigenic Escherichia coli. Trumbull, CT: Food and Nutrition Press; 2001. p. 113-48.
  3. ↵
    1. Johnson R,
    2. Clarke R,
    3. Wilson J,
    4. et al
    . Growing concerns and recent outbreaks involving non-O157:H7 serotypes of Verotoxigenic Escherichia coli. J Food Prot 1996;59:1112-22.
    Web of Science
  4. ↵
    1. Banatvala N,
    2. Debeukelaer M,
    3. Griffin P,
    4. et al
    . Shiga-like toxin-producing Escherichia col i O111 and associated hemolytic-uremic syndrome: a family outbreak. Pediatr Infect Dis J 1996;15:1008-11.
    CrossRefMedlineWeb of Science
    1. Caprioli A,
    2. Luzzi I,
    3. Rosmini F,
    4. et al
    . Communitywide outbreak of hemolytic-uremic syndrome associated with non-O157 Verocytotoxin-producing Escherichia coli. J Infect Dis 1994;169:208-11.
    Abstract/FREE Full Text
    1. Brooks J,
    2. Bergmire-Sweat D,
    3. Kennedy M,
    4. et al
    . Outbreak of Shiga toxin-producing Escherichia coli O111:H8 infections among attendees of a high school cheerleading camp. Clin Infect Dis 2004;38:190-8.
    Abstract/FREE Full Text
    1. McCarthy T,
    2. Barrett N,
    3. Hadler J,
    4. et al
    . Hemolytic-uremic syndrome and Escherichia coli O121 at a lake in Connecticut, 1999. Pediatrics 2001;108:e59-66.
    Abstract/FREE Full Text
  8. ↵
    1. Werber D,
    2. Fruth A,
    3. Liesegang A,
    4. et al
    . A multistate outbreak of Shiga toxin-producing Escherichia coli O26:H11 infections in Germany, detected by molecular subtyping surveillance. J Infect Dis 2002;186:419-22.
    Abstract/FREE Full Text
  9. ↵
    1. Wickham ME,
    2. Lupp C,
    3. Mascarenhas M,
    4. et al
    . Bacterial genetic determinants of non-O157 STEC outbreaks and hemolytic-uremic syndrome after infection. J Infect Dis 2006;194:819-27.
    Abstract/FREE Full Text
  10. ↵
    1. Tarr P,
    2. Gordon C,
    3. Chandler W
    . Shiga toxin-producing Escherichia coli and the haemolytic uraemic syndrome. Lancet 2005;365:1073-86.
    MedlineWeb of Science
  11. ↵
    1. Ethelberg S,
    2. Olsen K,
    3. Scheutz F,
    4. et al
    . Virulence factors for hemolytic uremic syndrome, Denmark. Emerg Infect Dis 2004;10:842-7.
    MedlineWeb of Science
  12. ↵
    1. Brooks J,
    2. Sowers E,
    3. Wells J,
    4. et al
    . Non-O157 Shiga toxin-producing Escherichia coli infections in the United States, 1983–2002. J Infect Dis 2005;192:1422-9.
    Abstract/FREE Full Text
  13. ↵
    1. Nataro JP,
    2. Kaper JB
    . Diarrheagenic Escherichia coli. Clin Microbiol Rev 1998;11:142-201.
    Abstract/FREE Full Text
  14. ↵
    1. Persing D,
    2. Smith T,
    3. Tenover F,
    4. et al.
    1. Olsvik O,
    2. Strockbine NA
    . PCR detection of heat-stable, heat labile, and Shiga-like toxin genes in Escherichia coli. In: Persing D, Smith T, Tenover F, et al., editors. Diagnostic molecular microbiology. Washington, DC: American Society for Microbiology Press; 1993. p. 271-6.
  15. ↵
    1. Paton A,
    2. Paton J
    . Detection and characterization of Shiga toxigenic Escherichia coli by using multiplex PCR assays for stx1, stx2, eaeA, enterhemorrhagic E. coli hylA, rfbO111, and rfbO157. J Clin Microbiol 1998;36:598-602.
    Abstract/FREE Full Text
  16. ↵
    1. Bergan T
    1. Orskov F,
    2. Orskov I
    . Serotyping of Escherichia coli. In: Bergan T, editor. Methods of microbiology. Vol. 14. London: Academic Press; 1984. p. 43-112.
  17. ↵
    1. Lockary V,
    2. Hudson R,
    3. Ball C
    . Shiga toxin-producing Escherichia coli, Idaho [letter]. Emerg Infect Dis 2007;13:1262-4.
    MedlineWeb of Science
    1. Fey P,
    2. Wickert R,
    3. Rupp M,
    4. et al
    . Prevalence of non-O157:H7 Shiga toxin-producing Escherichia coli in diarrheal stool samples from Nebraska. Emerg Infect Dis 2000;6:530-3.
    MedlineWeb of Science
  19. Centers for Disease Control and Prevention. Laboratory-confirmed non-O157 Shiga toxin-producing Escherichia coli —Connecticut, 2000–2005. MMWR Morb Mortal Wkly Rep 2007;56:29-31.
    Medline
    1. Jelacic J,
    2. Damrow T,
    3. Chen G,
    4. et al
    . Shiga toxin-producing Escherichia coli in Montana: bacterial genotypes and clinical profiles. J Infect Dis 2003;188:719-29.
    Abstract/FREE Full Text
  21. ↵
    1. Manning S,
    2. Madera R,
    3. Schneider W,
    4. et al
    . Surveillance for Shiga toxin-producing Escherichia coli, Michigan. Emerg Infect Dis 2007;13:318-21.
    MedlineWeb of Science
  22. ↵
    1. Kassenborg H,
    2. Hedberg C,
    3. Hoekstra M,
    4. et al
    . Farm visits and undercooked hamburgers as major risk factors for sporadic Escherichia coli O157:H7 infection: data from case-control study in 5 FoodNet sites. Clin Infect Dis 2004;38:271-8.
    Abstract/FREE Full Text
  23. ↵
    1. Voetsch A,
    2. Kennedy M,
    3. Keene W,
    4. et al
    . Risk factors for sporadic Shiga toxin-producing Escherichia coli O157 infections in FoodNet sites, 1999–2000. Epidemiol Infect 2007;135:993-1000.
    Medline
  24. ↵
    1. Bielaszewska M,
    2. Friedrich AW,
    3. Aldick T,
    4. et al
    . Shiga toxin activatable by intestinal mucus in Escherichia coli isolated from humans: predictor for a severe clinical outcome. Clin Infect Dis 2006;43:1160-7.
    Abstract/FREE Full Text
    1. Persson S,
    2. Olsen K,
    3. Ethelberg S,
    4. et al
    . Subtyping method for Escherichia coli Shiga toxin (verocytotoxin) 2 variants and correlations to clinical manifestations. J Clin Microbiol 2007;45:2020-4.
    Abstract/FREE Full Text
    1. Orth D,
    2. Grif K,
    3. Khan A,
    4. et al
    . The Shiga toxin genotype rather than the amount of Shiga toxin or the cytotoxicity of Shiga toxin in virto correlates with the appearance of the hemolytic uremic syndrome. Diagn Microbiol Infect Dis 2007;59:235-42.
    CrossRefMedlineWeb of Science
  27. ↵
    1. Friedrich A,
    2. Bielaszewska M,
    3. Zhang W,
    4. et al
    . Escherichia coli harbouring Shiga toxin 2 gene variants: frequency and association with clinical symptoms. J Infect Dis 2002;185:74-84.
    Abstract/FREE Full Text
| Table of Contents

This Article

  1. Clin Infect Dis. (2009) 49 (3): 358-364. doi: 10.1086/600302
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most