Clostridium difficile infection is recognized as a significant contributor to health care-associated morbidity, causing up to 30% of antibiotic-associated diarrhea.1 However, it is thought to play a secondary role in outpatient bacterial colitis. Enteric organisms, such as Campylobacter, Salmonella, Shigella, and Escherichia coli are considered to be the most common infectious causes of outpatient bacterial colitis. Recent reports from Quebec have documented a significant increase in the incidence of C. difficile toxin detection with a 3-fold increase in all patients, rising to a 10-fold increase in patients older than 65 years in 2003 compared with 1991 to 2002.2 This increase in C. difficile toxin detection was accompanied by a more than 2-fold increase in mortality and complications. These infections were reported both in the inpatient and in the outpatient populations. A more recent study of hospitalized Quebec patients reported a 23% mortality in hospitalized patients with C. difficile infection compared with 7% in matched hospital controls.3 This strain of C. difficile toxin resulting in higher morbidity and mortality has been identified as toxinotype III, North American PFGE type 1, and PCR-ribotype 027 (NAP1/027); it carries the binary toxin gene cdtB and an 18-bp deletion in tcdC, resulting in increased peak median concentrations of toxins A and B by 16- and 23-fold, respectively.4
We recently observed an increasing number of C. difficile infections in our community including a few outpatients without risk factors such as the use of antibiotics or recent hospitalization. This observation has also been noted by a recent Center for Disease Control report documenting cases of severe C. difficile diarrhea in patients thought to be at low risk for the infection. Low risk includes patients who have not been in a health-care setting or used antibiotics.5 The purpose of the present study was to determine the relative role of bacterial pathogens in patients with infectious colitis who were undergoing testing for C. difficile toxin and bacterial stool cultures at a large community hospital laboratory.
MATERIALS AND METHODS
We retrospectively reviewed the hospital records of all patients who had a positive C. difficile toxin A or B by enzyme-linked immunosorbent assay, or a positive stool culture for bacterial pathogens over a 12-month period (January 1, 2003-December 31, 2003) at a large community hospital (more than 100,000 emergency visits per year). The study was approved by the Institutional Review Boards of William Beaumont Hospital, Royal Oak, Mich. Our definition of a recurrent C. difficile infection was a positive toxin detected at more than 4 weeks after the initial positive test. A relapsed C. difficile infection was defined by a positive toxin assay at 11 to 28 days after the initial episode.
The computerized hospital records were examined for demographic data, hospital outcome, patient location (inpatient/outpatient), and results of microbiology studies. For the patients who died, we recorded the time elapsed between the last episode of C. difficile infection and death. While reviewing the electronic charts, we noticed that a positive C. difficile toxin assay was frequently preceded by a urine culture, and therefore, as a secondary objective, we collected data on the urine cultures done within the 2 months preceding the diagnosis of C. difficile. A urinary tract infection (UTI) was confirmed when the urine culture was positive for 1 or 2 organisms, at more than 100,000 colony-forming units for at least one of the pathogens.
For statistical analysis, differences between groups were analyzed using univariate ANOVA. The proportions were compared with χ2 test or Fisher exact test when the numbers were small. We defined statistical significance as a value of <0.05.
A total of 1434 patients met the criteria for inclusion into this study. Of these, 1372 had positive C. difficile toxin assays and 67 had positive stool cultures for bacterial pathogens (36 for Salmonella, 20 for Campylobacter, 9 for Shigella, and 2 for E. coli O157:H7). Five patients had coinfections. There were 3 coinfected with C. difficile toxin and Salmonella (all patients were females; ages 22, 53, and 63). One 16-year-old girl was coinfected with C. difficile and Campylobacter, and there was 1 coinfection with C. difficile and Shigella in a 40-year-old man. Eighty patients with C. difficile toxin were diagnosed as outpatients from the community. In the group with positive stool cultures, there were 60 patients referred from community and 7 inpatients.
The patients with bacterial diarrhea had a mean age of 33.6 years, significantly younger than the patients with C. difficile infection. This group included 37 women and 30 men. Of these, 12 were inpatients and 55 were outpatients. The diagnosis of bacterial infectious colitis was more frequent in the summer months (May-August), as shown in Figure 1, without reaching a statistically significant level.
In patients with bacterial colitis, 24 (35.8%) had both a C. difficile assay and ova and parasites stool testing with the initial evaluation, 14 (20.9%) had neither test, 18 (26.9%) were only tested for C. difficile toxin and 11 (16.4%) were only tested for ova and parasites. Of the community-referred patients with positive C. difficile toxin assay, 35 patients (43.75%) had only a stool culture and ova and parasites testing at the initial evaluation, 29 (36.25%) had neither test, 15 (18.75%) had only a stool culture, and 1 (1.25%) had only ova and parasites testing.
Of the 1372 patients with C. difficile infection, 568 were diagnosed as outpatients and 804 as inpatients (Table 1). Of the outpatients, 80 (14%) were referred from a community physician and 488 (86%) from extended-care facilities (ECF). The ECF residents were significantly older than the community residents (77.8 years vs. 52.8 years, P < 0.001) and the hospitalized patients (77.8 years vs. 65.7 years, P < 0.001). The ECF female residents were significantly older than the ECF male residents (79.2 years vs. 75.6 years, P < 0.001). There was no significant gender difference with regard to age in the inpatient population (65.7 years vs. 66 years, P = 0.8) and in the community residents (52 years vs. 54 years, P = 0.73). The age distribution among these groups is illustrated in Figure 2. As expected, most cases of C. difficile colitis occurred in the older age groups, both in absolute numbers (Fig. 2A) and as a percentage of the cases in the cohort (Fig. 2B). The female/male ratio was 1.5:1 in the outpatient group and 1:1 in the inpatient group.
In the outpatient group, there were 85 deaths (Table 2). All deaths occurred in the C. difficile group (none in the bacterial diarrhea group). The 1-month mortality after the detection of C. difficile toxin was 5.1% in the ECF patients and 1.25% in the community. The mortality at 3 months was 11.45% in the ECF group and 3.75% in the community. For the patients who died in the first 30 days after the diagnosis of C. difficile infection, we speculate that this infection could have been a contributing factor to the clinical decline leading to the patient demise. The deaths that occurred at greater than 90 days after the last documented episode of C. difficile were probably unrelated to the infection. The average age of the expired patients was 83.1 years in the ECF and 78.4 years in the community (P = 0.2, not significant). Although there was a trend of a higher mortality in the ECF group, this was not statistically significant.
In the outpatient group, 17% (97/568) had more than 1 episode of C. difficile infection (Table 3). There was no statistically significant difference in the relapsed or recurrent infection in either group. No patient in the community had more than 1 relapse and only 1 patient had 2 recurrences. ECF patients had more than 1 relapse in 2% (8 patients with 2 relapses, 1 with 3 relapses, and 1 with 4 relapses) and more than 1 recurrence in 1.4% (6 patients with 2 recurrences, and 1 patient with 3 recurrences).
The detection of C. difficile toxin was preceded within 2 months by a urine culture in 32.6% (159/488) of the ECF patients compared with 10% (8/80) of the community residents; the difference in a positive urine culture preceding the detection of C. difficile toxin between the ECF and community groups was statistically significant (Table 4). However, the urine culture was positive, confirming a UTI, only in 84/159 (52.8%) of the ECF residents and in 4/8 (50%) of the community residents. UTI was identified in 18.4% of the outpatient female (63/342) and in 11% of the outpatient males (25/226); the difference was statistically significant (P < 0.025). There was an increased proportion of C. difficile-positive men with a negative urine culture (60.6% of the urine culture checked in the male outpatients were negative for C. difficile, compared with 38.7% in women; P < 0.01).
Infectious diarrhea is a major public concern.1-5 Although most bacterial or food-borne diarrheal diseases are self-limiting, C. difficile colitis can be associated with significant morbidity and mortality. Recent trends show an increased incidence in both nosocomial and community-acquired C. difficile colitis.6-12 This epidemic imposes a significant economic burden on the health system. Recent reports of an aggressive strain in the United States, Canada, and England document that C. difficile may be becoming a more lethal infection.3,4
Our study confirms that C. difficile infection with a toxin-producing strain is the most common identifiable cause of infectious diarrhea, both in the inpatient and in the outpatient setting. C. difficile affects predominantly the older age groups. In our outpatient setting, there were 1.5 times more women. This observation most likely relates to the high frequency of infection in patients from extended care facilities and the fact that most of these patients are women. The retrospective nature of this study limits information on patient presentation, type of antibiotics used, concomitant medications such as proton pump inhibitors, and symptoms in the outpatient setting.
The incidence of recurrent/relapsed C. difficile infection in our series was 17%, less than the incidence reported in the literature (20-58%).9 The mortality was also lower in our cohort than that cited in the literature (13.8-23%).2,3 The mortality in our ECF residents was higher than that in the community residents, without reaching the level of significance. These observations in our study may again reflect the retrospective nature of the study and the comorbid conditions in our older cohort.
The patients with non-C. difficile bacterial diarrhea were younger than the patients with C. difficile toxin. The most frequent pathogens were Salmonella, Campylobacter, and Shigella. The proportion of patients with bacterial colitis is comparable with the figures cited in the literature.1,13-16 A prospective study by Svenungsson et al14 in Sweden in patients with acute diarrhea identified Campylobacter in 13% of the patients, C. difficile in 13%, E. coli in 8%, Salmonella in 7%, and Shigella in 4%. In a study by Maraki et al,15 bacterial pathogens other than C. difficile were isolated in 14% of the patients with acute diarrhea. The most frequent pathogen was Salmonella (43% of isolates), followed by Campylobacter (30% of isolates), enteropathogenic E. coli (13% of isolates), and Shigella (2% of isolates). C. difficile toxin was detected in 14.4% of the patients. In our cohort, we detected 80 C. difficile toxin-positive outpatients (non-ECF), more than all other forms of infectious colitis (67) combined.
A previously unreported observation in our series notes that one third of the ECF patients who had a positive C. difficile toxin assay also had a urine culture checked in the previous 2 months; only half of these cultures were positive. Men were more likely to have a negative urine culture (60.6% vs. 38.7%), whereas women were diagnosed more frequently than men with a UTI (18.4% vs. 11%). Recent reports have implicated an emerging resistance of C. difficile to cephalosporins and fluoroquinolones, antibiotics that are often used to treat UTIs.17,18 Based on our data, we speculate that detection of C. difficile infection in our cohort suggests that some patients with suspected UTIs were started empirically on antibiotics before the result from the urine culture was available. This speculation could not be confirmed as we could not review the paper charts from the private offices or from the nursing homes regarding the administration of antibiotics. However, if the patients with suspected UTI were truly treated empirically while awaiting the result of a urine culture, it could potentially have avoided the development of C. difficile colitis in 79 (13.9%) outpatients. Given the trend toward a higher 30-day mortality in the ECF patients diagnosed with C. difficile colitis (~5%), our data suggest that it would be prudent to avoid empirical treatment of UTI.
In a community setting, C. difficile is the most common identifiable bacterial pathogen among inpatients and outpatients with bacterial colitis. Our observation agrees with the current recommendations of the Center for Disease Control to consider C. difficile infection in the differential diagnosis of patients with symptoms of diarrhea for more than 3 days.5 One third of patients in an ECF and 10% of outpatients in the community with a positive C. difficile toxin assay are likely to have had a suspected UTI. Empirical antibiotic treatment of UTI may contribute to the C. difficile toxin detection in the extended-care facility cohort.
1. Bartlett J. Antibiotic-associated diarrhea. NEJM
2. Pepin J, Valiquette L, Alary ME, et al. Clostridium difficile
-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ
3. Pepin J, Valiquette L, Cossette B, et al. Mortality attributable to nosocomial Clostridium difficile
-associated disease during an epidemic caused by a hypervirulent strain in Quebec. CMAJ
4. Warny M, Pepin J, Fang A, et al. Toxin production by an emerging strain of Clostridium difficile
associated with outbreaks of severe disease in North America and Europe. Lancet
5. Tan T, Booker C, Bresnitz E, et al. Severe Clostridium difficile
-associated disease in populations previously at low risk-four states. MMWR
6. Archibald L, Banerjee S, Jarvis W. Secular trends in hospital-acquired Clostridium difficile
disease in the United States, 1987-2001. J Infect Dis
7. Sohn S, Climo M, Diekema D, et al. Varying rates of Clostridium difficile
-associated diarrhea at prevention epicenter hospitals. Infect Control Hosp Epidemiol
8. Muto C, Pokrywka M, Shutt K, et al. A large outbreak of Clostridium difficile
-associated disease with an unexpected proportion of deaths and colectomies at a teaching hospital following increased fluoroquinolone use. Infect Control Hosp Epidemiol
9. Pepin J, Alary ME, Valiquette L, et al. Increasing risk of relapse after treatment of Clostridium difficile
colitis in Quebec, Canada. Clin Infect Dis
10. Eggertson L. C. difficile: by the numbers. CMAJ
11. Martirosian G, Szczesny A, Cohen SH. Analysis of Clostridium difficile
-associated diarrhea among patients hospitalized in tertiary care academic hospital. Diagn Microbiol Infect Dis
12. McEllistrem MC, Carman RJ, Gerding DN, et al. A hospital outbreak of Clostridium difficile
disease associated with isolates carrying binary toxin genes. Clin Infect Dis
13. Rohner P, Pittet D, Pepey B, et al. Etiological agents of infectious diarrhea: implications for requests for microbial culture. J Clin Microbiol
14. Svenungsson B, Lagergren A, Ekwall E, et al. Enteropathogens in adult patients with diarrhea and healthy control subjects: a 1-year prospective study in a Swedish clinic for infectious diseases. Clin Infect Dis
15. Maraki S, Georgiladakis A, Tselentis Y, et al. A 5-year study of the bacterial pathogens associated with acute diarrhea on the island of Crete, Greece, and their resistance to antibiotics. Eur J Epidemiol
16. Liesenfield O, Weinke T, Hahn H. Three year prevalence of enteropathogenic bacteria in an urban patient population in Germany. Infection
17. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi-institutional outbreak of Clostridium difficile
-associated diarrhea with high morbidity and mortality. NEJM
© 2007 Lippincott Williams & Wilkins, Inc.
18. McDonald LC, Killgore GE, Thompson A, et al. An epidemic, toxin gene-variant strain of Clostridium difficile