Clostridium difficile Infection in Outpatients, Maryland and Connecticut, USA, 2002–2007

Clostridium difficile, the most commonly recognized diarrheagenic pathogen among hospitalized persons, can cause outpatient diarrhea. Of 1,091 outpatients with diarrhea, we found 43 (3.9%) who were positive for C. difficile toxin. Only 7 had no recognized risk factors, and 3 had neither risk factors nor co-infection with another enteric pathogen.

I n the United States, ≈375 million episodes of acute diarrhea occur annually (1). Among hospitalized persons, toxin-producing Clostridium diffi cile is a primary diarrheagenic pathogen, usually as a consequence of normal bowel fl ora distortion caused by antimicrobial drug therapy (2,3). C. diffi cile infection (CDI) complicates and prolongs hospital stays, leading to increases in health care costs, illness, and death. Recent reports suggest increases in community-onset CDI among persons without recent antimicrobial drug treatment or hospitalization. We describe a prospective evaluation of CDI in persons with diarrhea who visited emergency departments (EDs) and ambulatory primary care clinics in Baltimore, Maryland, and New Haven, Connecticut, and identify microbiologic causes and epidemiologic characteristics of diarrhea. This report highlights cases of outpatient CDI, identifi es factors associated with infection, and describes molecular strain characterization.
Patients seeking medical attention for communityonset diarrheal illnesses were enrolled from May 2002 through September 2004 in the EDs and ambulatory clinics at Yale-New Haven Hospital (New Haven, CT, USA) and from May 2002 through July 2007 at EDs and clinics affi liated with the University of Maryland (Baltimore, MD, USA) (4).
Informed consent for stool sample collection, initial and follow-up patient interviews, and medical records review was obtained from primarily urban and suburban residents, or parents/guardians for minors, who sought treatment for self-identifi ed primary or secondary diarrhea. This research was approved by the institutional review boards at all participating institutions.
Participants were interviewed at outpatient clinics to assess health status, symptoms, and potential exposures to enteric pathogens, and at follow-up to determine the duration of diarrhea, whether treatment was administered, or whether hospitalization resulted from the initial visit. Stool samples collected during the visit or provided within 48 h and kept cool were homogenized and transferred into multiple vials for storage at -80°C.
An outpatient CDI case was defi ned in an outpatient with diarrhea whose stool was positive for C. diffi cile toxins by enzyme immunoassay (TOX A/B II ELISA; TechLab, Blacksburg, VA, USA). Presumptive non-health care-associated (NHA) CDI was defi ned by the absence of an overnight stay at an inpatient healthcare facility over the previous month.
Traditional risk factors for CDI that were investigated included antimicrobial drug use within the past month, age >65 years, serious underlying illness/weakened immune system, history of bowel or ulcer surgery, colon disease, previous CDI, and recent hospitalization. Statistical analysis was done by using SAS version 9.2 (SAS Institute, Inc., Cary, NC, USA). All p values reported are 2-sided, with no correction for multiple comparisons; p<0.05 was considered signifi cant.
C. diffi cile toxin-positive stool specimens, in 1-mL aliquots, were shipped frozen to the Centers for Disease Control and Prevention (Atlanta, GA, USA) anaerobe laboratory for culturing by direct inoculation onto cycloserine cefoxitin fructose agar (CCFA) or ethanol shock, followed by CCFA inoculation. Cultures were incubated for 48-72 h at 35°C under anaerobic conditions and examined for characteristic yellow-green fl uorescence under long-wave ultraviolet light and CCFAp-cresol odor. C. diffi cile colonies were confi rmed with indole (negative) and PRO disk (positive; Remel, Lenexa, KS, USA) tests.
C. diffi cile toxin tests were performed on 1,091/1,197 stool specimens; 43 (3.9%) of these case-patients met the case defi nition for outpatient CDI. The mean age of these  Control and Prevention for anaerobic culture and C. diffi cile was isolated from 31 samples. Binary toxin was identifi ed in 12 (38.7%). Pulsed-fi eld gel electrophoresis identifi ed 15 different types ( Figure). No associations were found between risk factors, including age, and strain or toxinotype (data not shown). NHA-CDI has been recognized for >12 years, and recent reports suggest that disease occurs without patient's known exposure to antimicrobial drugs or other previously identifi ed risk factors (2,(8)(9)(10)(11)(12)(13). Although we found a proportion of C. diffi cile-positive diarrheal stools similar to that of 2 other recent prospective studies that used confi rmatory culture (i.e., 1.5%-3.9%) for outpatient CDI (7,13,14), we also found a lower proportion of outpatient CDI cases without recognized risk factors of recent hospitalization, chronic medical conditions, recent antimicrobial drug exposure, or co-infection than did those studies.
One limitation of our study was using retrospective self-reporting for assessment of hospitalizations or antimicrobial drug use in the previous month, which potentially can result in recall bias. Also, antimicrobial drug therapy was assessed for only 4 weeks before diarrhea onset; exposure to antimicrobial drugs for a period longer than 1 month before patient seeks treatment may present a risk for CDI. In addition, this study was conducted at 2 urban centers in the eastern United States and may not be generalizable to other locations or clinical settings. Finally, although enzyme immunoassay detection for C. diffi cile was the standard of care at the time of the study, it is now considered too insensitive to be used as a stand-alone diagnostic test (15).
In summary, we detected toxigenic C. diffi cile in a similar proportion of patients to those reported in other studies of CDI. However, all but 3 patients had either known risk factors for CDI or other pathogens potentially responsible for their illness; 1 was <1 year of age. C. diffi cile isolates responsible for outpatient CDI are genetically diverse. An evolving picture of widespread, frequent CDI among outpatients without risk factors should be tempered by these fi ndings.
Funding for this study was provided by the National Center for Infectious Diseases of the Centers for Disease Control and Prevention (grant no. U01CI000296), the American Association of Medical Colleges (grant no. MM0205-02/02).
Dr Hirshon is an associate professor in the Departments of Emergency Medicine and Epidemiology and Public Health at the University of Maryland School of Medicine. His research interests include developing emergency departments as sites for surveillance and hypothesis-driven research in public health.