Lukas Fenner,*1 10. Feuillet L, Carvajal J, Sudre I, Pelletier J, comprise a substantial percentage of
Thomassin JM, Drancourt M, et al. First STEC cases.
isolation of Bacteroides thetaiotaomicron
Pascal Ananian,† from a patient with a cholesteatoma and In other countries, nonculture-
and Didier Raoult*† experiencing meningitis. J Clin Microbiol. based methods are routinely used for
*Hôpital de la Timone, Marseille, France; 2005;43:1467–9. STEC detection (6). However, E. coli
and †Hôpital de la Conception, Marseille, O157:H7 culture methods remain the
France Address for correspondence: Didier Raoult, focus in the United Kingdom, Canada,
Hôpital de la Timone, 264 rue Saint-Pierre, and the United States (6). Reliance on
References 13385 Marseille, France; email: didier.raoult@ culture methods can result in mislead-
medecine.univ-mrs.fr ing interpretations of STEC preva-
1. Rautio M, Lonnroth M, Saxen H, Nikku R,
lence. For example, 93% of STEC
Väisanen ML, Finegold SM, et al. Charac-
teristics of an unusual anaerobic pigment- infections in Canada are reported to
ed gram-negative rod isolated from nor- be E. coli O157:H7, yet a Manitoba
mal and inﬂamed appendices. Clin Infect 1992 study showed that when toxin
Dis. 1997;25(Suppl 2):S107–10.
assays were used, 35% of the recov-
2. Rautio M, Saxen H, Siitonen A, Nikku
R, Jousimies-Somer H. Bacteriology ered STEC isolates were non-O157
of histopathologically deﬁned appen-
dicitis in children. Pediatr Infect Dis J.
Shiga Toxin– serotypes (6).
Analysis of reported non-O157
2000;19:1078–83. producing STEC cases in Idaho showed a simi-
3. Kimura M. A simple method for estimat-
ing evolutionary rates of base substitutions Escherichia coli, lar trend. From 2002–2004, 66% of
through comparative studies of nucleotide
sequences. J Mol Evol. 1980;16:111–20.
Idaho Idaho’s non-O157 cases originated
in Health District 7, where >70% of
4. Rautio M, Eerola E, Väisänen-Tunkel-
To the Editor: Data collected stool cultures are screened by enzyme
rott ML, Molitoris D, Lawson P, Collins
MD, et al. Reclassiﬁcation of Bacteroi- from expanded surveillance study immunoassay (EIA) for Shiga toxin
des putredinis (Weinberg et al., 1937) in suggest that more than half of Idaho (Premier EHEC, Meridian Bioscience,
a new genus Alistipes gen. nov., as Alis- Shiga toxin–producing Escherichia Cincinnati, OH, USA). This rate was
tipes putredinis comb.nov., and descrip-
coli (STEC) illnesses are caused by disproportionately higher than that
tion of Alistipes ﬁnegoldii sp. nov., from
human sources. Syst Appl Microbiol. non-O157 serotypes. Using data from of the remaining 6 health districts,
2003;26:182–8. a regional medical center whose stool which primarily use culture methods
5. Weisburg WG, Barns SM, Pelletier DA, culture protocol included Shiga toxin to screen for E. coli O157:H7. We
Lane DJ. 16S ribosomal DNA ampliﬁca-
testing, we predicted Idaho’s STEC hypothesized that this disproportion
tion for phylogenetic study. J Bacteriol.
1991;173:697–703. incidence to be signiﬁcantly higher if was due to differences in stool culture
6. Fenner L, Roux V, Mallet MN, Raoult D. non-O157 STEC E. coli were routine- protocol. To test this premise, we con-
Bacteroides massiliensis sp. nov., isolated ly detected by immunoassay. Recent ducted enhanced surveillance for 16
from blood culture of a newborn. Int J Syst
ﬁndings suggest that the prediction months in a “low” STEC incidence
Evol Microbiol. 2005;55:1335–7.
7. Rigottier-Gois L, Rochet V, Garrec N, was accurate in an expanded surveil- area, Health District 5. A total of 2,065
Suau A, Doré J. Enumeration of Bacteroi- lance area. stools submitted for culture were
des species in human faeces by ﬂuorescent Several studies have shown an screened for Shiga toxin by EIA. With
in situ hybridisation combined with ﬂow
increased incidence of non-O157 this approach, reported non-O157
cytometry using 16S rRNA probes. Syst
Appl Microbiol. 2003;26:110–8. STEC infections in the United States. STEC incidence rose from <1 case/
8. Brook I. Clinical review: bacteremia For example, a community hospital year/100,000 population to 11 cases/
caused by anaerobic bacteria in children. in Virginia detected non-O157 sero- year/100,000 population. Addition-
Crit Care. 2002;6:205–11.
types in 31% of patients with STEC ally, 56% of recovered STEC isolates
9. Wareham DW, Wilks M, Ahmed D, Bra-
zier JS, Millar M. Anaerobic sepsis due to from 1995–2002 (1). A 1998 Nebraska were non-O157 serotypes, mirroring
multidrug-resistant Bacteroides fragilis: study that analyzed 30,000 diarrheal the proportion of non-O157 detected
microbiological cure and clinical response stool samples found that non-O157 in District 7. Notably, this appears to
with linezolid therapy. Clin Infect Dis.
and O157:H7 STEC were equally be the endemic rate for District 5 be-
prevalent (2). Additionally, ﬁndings cause no non-O157 STEC outbreaks
from a Connecticut study of labora- or matching pulsed-ﬁeld gel electro-
tory-conﬁrmed cases (3), STEC sur- phoresis patterns were detected dur-
veillance results from Montana (4), ing the surveillance period. Although
Current afﬁliation: University
Hospital and a recent study from Michigan our study captured only a portion of
Basel, Basel, Switzerland (5) indicate that non-O157 serotypes stool cultures in Idaho, our ﬁndings
1262 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 8, August 2007
demonstrated increased prevalence of proportion of HUS cases attributable O157 in STEC disease will remain
non-O157 STEC in the region when to non-O157 STEC. obscured as long as screening methods
nonculture methods were used. Some evidence suggests that the focus on traditional culture methods.
Two barriers cited for not routine- testing focus may be changing in the
ly screening diarrheal stools for Shiga United States. We used US Census Acknowledgments
toxin are cost and perception of low Bureau population statistics to trans- We thank Richard Gelok and staff
non-O157 STEC incidence. While late reported O157:H7 and non-O157 at Eastern Idaho Regional Medical Cen-
toxin testing is more expensive than STEC cases for each state into inci- ter in Idaho Falls and Janie Palmer and
culture testing, the potential effects of dence data. Despite widespread varia- staff at St. Luke’s Magic Valley Regional
misdiagnosis may outweigh cost con- tion in STEC testing and incidence Medical Center in Twin Falls for their
cerns. A study estimating the ﬁnancial among states, there has been a signiﬁ- participation.
repercussions of E. coli O157 infec- cant statistical decline in the propor-
Partial support came from the Centers
tions in the United States suggested tion of E. coli O157:H7 among total
for Disease Control and Prevention, Epi-
that annual cost associated with this STEC cases every year since 2001
demiology and Laboratory Capacity grant
pathogen is $405 million, with the (Figure; p<0.001) (10). Consistent
cost per case varying from $26 for with this trend, the incidence of non-
those who do not seek medical care to O157 STEC in the United States has
$6.2 million for a patient with fatal he- increased (10). This may indicate that Vivian Marie Lockary,*
molytic uremic syndrome (HUS) (7). more laboratories are adopting Shiga Richard Frederick Hudson,*
Non-O157 STEC infections have been toxin testing protocols, as we are ad- and Christopher Lawrence Ball*
an important cause of HUS in many vocating in Idaho. Our ﬁndings sug- *Idaho Bureau of Laboratories, Boise,
countries. For example, a 3-year pro- gest that perceptions of low non-O157 Idaho, USA
spective study in Germany and Aus- STEC incidence in Idaho are probably
tria reported that non-O157 serotypes artifactual and due to overemphasis on References
comprised 90 (43%) of 207 STEC iso- culture methods for O157 STEC. Our
1. Park CH, Kim HJ, Hixon DL. Importance
lates from stools of 394 pediatric pa- ongoing EIA-based surveillance high- of testing stool specimens for Shiga tox-
tients with HUS (8). Further, a 6-year lights the need for continued investiga- ins [letter]. J Clin Microbiol. 2002;40:
Danish study of 343 registered STEC tion of the epidemiology of non-O157 3542–3.
2. Fey PD, Wickert RS, Rupp ME, Safranek
patients found that 76% of STEC and STEC disease. We conclude that O157
TJ, Hinrichs SH. Prevalence of non-O157:
48% of HUS cases were attributable to STEC culturing has limited usefulness H7 Shiga toxin–producing Escherichia
non-O157 serotypes (9). In the United in areas like the Idaho health districts coli in diarrheal stool samples from Ne-
States, continued reliance on O157 investigated, where non-O157 sero- braska. Emerg Infect Dis. 2000;6:530–3.
3. Centers for Disease Control and Preven-
STEC culturing hinders our ability to types accounted for 55% of STEC ill-
tion. Laboratory-conﬁrmed non-O157
determine the ﬁnancial effects and the nesses. The true involvement of non- Shiga toxin–producing Escherichia coli.
Connecticut, 2000–2005. MMWR Morb
Mortal Wkly Rep. 2007;56:29–31.
4. Jelacic JK, Damrow T, Chen GS, Jelacic
S, Bielaszewska M, Ciol M, et al. Shiga
toxin-producing Escherichia coli in Mon-
tana: bacterial genotypes and clinical pro-
ﬁles. J Infect Dis. 2003;188:719–29.
5. Manning SD, Madera RT, Schneider W,
Dietrich SE, Khalife W, Brown W, et al.
Surveillance for Shiga toxin–producing
Escherichia coli, Michigan, 2001–2005.
Emerg Infect Dis. 2007;13:318–21.
6. Kaper JB, O’Brien AD, eds. Escherichia
coli O157:H7 and other Shiga toxin–pro-
ducing E. coli Strains. Washington: ASM
Press; 1998. p. 26, 55.
7. Frenzen PD, Drake A, Angulo FJ; Emerg-
ing Infections Program FoodNet Work-
ing Group. Economic cost of illness due
to Escherichia coli O157 infections in
Figure. Shiga toxin–producing Escherichia coli (STEC) incidence trends, United States, the United States. J Food Prot. 2005;68:
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 8, August 2007 1263
8. Gerber A, Karch H, Allerberger F, Ver- Importazione (GISPI) centers (Italian (33.3%) and thrombocytopenia (plate-
weyen HM, Zimmerhackl LB. Clinical network of Institutes of Infectious and let count <150,000/μL) in 1 patient
course and the role of Shiga toxin–pro-
ducing Escherichia coli infection in the Tropical Diseases). Serologic diagno- (8.3%). This ﬁnding may help distin-
hemolytic-uremic syndrome in pediatric sis was performed with a hemagglu- guish CHIKV infection from dengue
patients, 1997–2000, in Germany and tination-inhibition test and conﬁrmed fever (4). Anemia (hemoglobin level
Austria: a prospective study. J Infect Dis. by a plaque-reduction neutralization <12 g/dL) was found in only 1 patient
9. Ethelberg S, Olsen KE, Scheutz F, Jen- test (8). Demographic and epidemio- (8.3%). Alanine aminotransferase
sen C, Schiellerup P, Engberg J, et al. logic characteristics of these patients (ALT) and aspartate aminotransferase
Virulence factors for hemolytic uremic are reported in the Table. (AST) determination were available
syndrome, Denmark. Emerg Infect Dis. Cases were distributed through- for 12 patients. ALT and AST levels
10. Centers for Disease Control and Preven- out the year with a peak from March were elevated (>40 IU/L) in 5 (41.7%)
tion. Summary of notiﬁable diseases, to May 2006 (n = 10). Nine patients and 2 (16.7%) patients, respectively.
United States. [cited 2007 Jan 11]. Avail- (53%) were men. Median age was Seven (46.7%) of 15 patients fully
able from http://www.cdc.gov/mmwr/ 43 years (range 31–66 years). Sev- recovered within 1 month; 8 patients
eral reasons for travel were reported: (53.3%) reported persistent arthralgia.
Address for correspondence: Vivian Marie tourism (64.6%), visits to relatives or Because the GISPI network pro-
Lockary, Idaho Bureau of Laboratories, 2220 friends (11.8%), business (11.8%), and vides regional coverage only, the
Old Penitentiary Rd, Boise, ID 83712, USA; missionary work (5.9%). One patient number of imported CHIKV cases in
email: firstname.lastname@example.org was a resident in the disease-epidemic all of Italy in 2006 was likely higher.
area. The median exposure time in the Moreover, most patients probably did
CHIKV-endemic area for the 15 trav- not seek medical care, and when they
elers was 15 days (range 9–93 days) did, physicians may have failed to
(missionary and resident patients were recognize the disease because of lack
excluded). The median delay before of familiarity with it or limited diag-
being seen at a clinic after return was nostic facilities. Differential diagno-
2 days (range 0–73 days). Only 7 pa- sis with other arthropodborne viruses
tients (41.2%) were hospitalized. The of the Alphavirus genus (Ross River,
Imported remainder were outpatients. Barmah Forest, o’nyong nyong, Sind-
Chikungunya All patients had fever; arthralgia bis, and Mayaro viruses) is difﬁcult,
Infection, Italy (88.2%, n = 15), weakness (70.6%, n but these are comparatively rare. In
= 12), headache (11.8%, n = 2), diar- contrast, dengue and CHIKV epidem-
To the Editor: Chikungunya vi- rhea (11.8%, n = 2), and gum bleeding ics may overlap, and potential patients
rus (CHIKV) infection is a self-lim- and epistaxis (5.9%, n = 1) were other should be screened for both.
iting illness characterized by fever, reported symptoms. The median dura- The potential risk for introduction
headache, weakness, rash, and arthral- tion of fever was 5 days (range 2–12 and establishment of CHIKV reser-
gia. Some patients have prolonged days). Only 7 of 16 patients (43.8%) voirs in areas with mosquito vectors
weakness or arthralgia lasting several were still febrile when ﬁrst seen. was discussed in March 2006 by a mul-
months. In 2006, several Indian Ocean Physical examination showed diffuse tidisciplinary European expert panel
states and India had an outbreak of macular erythematous rash in 13 pa- (9). In Italy, A. albopictus was ﬁrst
CHIKV infection (1,2). During the tients (76.5%), a similar rate to that recorded in 1990; it has since quickly
epidemic’s peak, some European and reported among French travelers (4). spread across the country. Scattered
American travelers returning from Hepatomegaly was found in 2 patients foci are now reported in almost all re-
these areas were infected (3–6). (11.8%), splenomegaly in 2 (11.8%), gions, mainly along the coastal plains,
Because the foci of Aedes al- and peripheral lymphadenopathy in 2 from the sea to the inlands, up to an
bopictus, 1 of the 2 main vectors of (11.8%). altitude of ≈500–600 m (7).
CHIKV, are now in Italy and many Twelve acute-phase patients were The ability of A. albopictus to
travelers visit CHIKV-epidemic ar- admitted to the hospital for blood colonize new areas and its adaptability
eas, surveillance for imported cases testing within 3 days of the initial ex- to the mild Italian climate allow vec-
is mandatory in Italy (7). From July amination. In contrast with results of tor populations to be active throughout
to September 2006, a total of 17 con- other studies, leukopenia and throm- the year (10). The patient is thought to
ﬁrmed cases of CHIKV infection were bocytopenia were uncommon in our be viremic for only 6–7 days (shortly
observed in travelers at 5 Gruppo di study. Leukopenia (leukocyte count before and during the febrile period)
Interesse e Studio delle Patologie di <4,000/μL) was present in 4 patients (6). We were unable to directly assess
1264 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 8, August 2007