Nielsen, Nathan D. MD*; Layton, Blythe A. BA†; McDonald, L. Clifford MD‡; Gerding, Dale N. MD§; Liedtke, Laura A. RN†; Strausbaugh, Larry J. MD†∥¶; Infectious Diseases Society of America Emerging Infections Network
For almost 3 decades, Clostridium difficile has been recognized as the leading cause of antibiotic-associated colitis throughout the world.1 Despite the absence of population-based studies to define the true incidence of C. difficile-associated disease (CDAD), several reports suggest that its incidence has increased in recent years.2-4 Moreover, other reports5-8 suggest that CDAD has become more severe and less responsive to therapy. Since 2001, concerns about changes in the epidemiology and clinical features of CDAD appeared in postings on the electronic mail conference of the Infectious Diseases Society of America (IDSA) Emerging Infections Network (EIN). Responding to these postings, the EIN formally surveyed its members in the spring of 2004 to characterize their experience with CDAD. In the fall of 2004, the EIN issued a second survey to clarify some interim verbal reports from members. This report summarizes the results of the 2 surveys, which provide additional perspective on changing patterns of CDAD in the United States.
MATERIALS AND METHODS
The IDSA EIN is a voluntary network of adult and pediatric infectious disease consultants (IDCs) who belong to either the IDSA or the Pediatric Infectious Diseases Society.9 During March 2004, the IDSA EIN distributed a 1-page introduction and a 1-page questionnaire via e-mail and facsimile to all 841 members practicing in the United States. Nonrespondents received a second and third mailing, 2 and 4 weeks after the original, respectively. The survey inquired about member experience with CDAD during the preceding 6 months and specifically sought perceptions about the relative frequency of cases, severity, recurrences, and refractoriness to therapy. Members were asked to compare their perceptions from the previous 6 months with those of prior years. Along with the perceptions of changed frequency or character of disease, IDCs were asked to estimate the number of all CDAD cases encountered in the past 6 months and the number of severe cases, treatment-refractory cases, and recurrent cases. The survey instrument did not provide definitions for severe, refractory, or recurrent CDAD.
In November 2004, the EIN issued a second survey similar to the first. However, members who had responded to the first survey did not complete it unless their observations and impressions had changed. Members noting changes in the period from May to November and members not responding to the first survey completed the second survey in its entirety.
In combination, the surveys reflected the experience of EIN members during the period from September 2003 to November 2004 in blocks of 6 months as captured by the 2 surveys. Results of both surveys were analyzed separately and in comparison to each other and, where appropriate, analyzed by US census divisions (Fig. 1).10 When respondents provided a range of observed cases to answer a question, only the lower value of the range was used for analysis. χ2 Tests were used to identify statistically significant differences in frequencies between geographic divisions.
Characteristics of Respondents
Overall, 456 (54.2%) of the 841 EIN members who received the first survey provided analyzable data, and 387 (43.9%) of 882 who received the second survey provided analyzable data. Altogether, 532 unique individuals (523 United States based) responded to one or both surveys (Fig. 2). Data from prior surveys completed by 468 respondents indicated that approximately half of them work as IDCs in university-affiliated hospitals and half in community hospitals.
First Survey: Experience and Perceptions
The first survey targeted member experience during the period from September 2003 to March 2004. During this period, the 447 US-based respondents (of 456 total respondents) reported seeing a total of 4803 cases of CDAD (Table 1). The number of cases reported by geographic division ranged from 189 in the East South Central Division to 990 in the South Atlantic Division. In all but one division (East South Central), the average number of CDAD cases reported by each EIN members ranged from 10 to 12.2 cases.
The overall complication rate for the 4803 cases reported by EIN members averaged 14.1% (Table 1). In the preceding 6 months, 129 respondents saw 251 patients with toxic megacolon not requiring colectomy; 96 saw 162 patients with toxic megacolon requiring colectomy; 64 saw 94 patients with colonic perforations; and 108 saw 169 fatal cases. The proportion of CDAD cases with one or more of these 4 complications varied from 3.9% in the Mountain Division to 22.8% in the East South Central Division (Table 1). Divisional data for the 4 complications exhibited similar variation.
Approximately one third of EIN respondents perceived changes in the epidemiology of CDAD during the 6-month survey period (Table 2). Specifically, 130 EIN members (29.1%) reported seeing increased numbers of cases in the preceding 6 months; 130 respondents (29.1%) reported seeing more severe or fulminant CDAD; 151 respondents (33.8%) reported seeing more recurrent cases; and 138 respondents (30.9%) reported encountering more cases refractory to therapy. Of note, greater proportions of EIN members in the mid-Atlantic, West North Central, South Atlantic, and West South Central Divisions perceived an increase in their CDAD caseload (χ2 = 13.8, P = 0.008). Larger proportions of respondents in the New England, West North Central, and East South Central Divisions reported an increase in more severe or fulminant cases (χ2 = 11.7, P = 0.019). In addition, larger proportions of EIN respondents in the New England, mid-Atlantic, East North Central, West North Central, and South Atlantic Divisions reported increases in recurrent CDAD cases (χ2 = 9.9, P = 0.041). Finally, larger proportions of EIN members in the New England, mid-Atlantic, and West South Central divisions noted increases in CDAD cases that were refractory to therapy (χ2 = 18.4, P = 0.001).
Second Survey: Experience and Perceptions
The second survey captured EIN member experiences during the 6-month period from May to November 2004. Of the 387 respondents, 221 (57%) reported no change in their experiences with CDAD from that reported in the first survey. Ninety members (23%) who had responded to the first survey also completed the second one because their experience with CDAD had changed in the interim. Of the remainder, 76 (20%) had not responded to the first survey (Fig. 2).
In the first survey, the 221 respondents who reported no change between surveys had reported 2264 cases of CDAD, an average of 10.4 cases per respondent. In addition, they reported a total of 110 cases of toxic megacolon not requiring colectomy (4.9% of their cases), 77 cases of toxic megacolon requiring colectomy (3.4%), 30 cases of colonic perforation (1.3%), 67 fatalities (3.0%), and an overall complication rate of 12.5%. Within this group, 59 (27%) had reported increased numbers of CDAD cases, 56 (25.7%) more severe cases, 64 (38.6%) more recurrences, and 61 (36.7%) more refractoriness to therapy.
The 166 EIN members providing data for the second survey reported seeing 1976 CDAD cases, an average of 11.9 cases per respondent (Table 3). Respondents in the mid-Atlantic, South Atlantic, and Pacific Divisions reported the largest number of cases; however, the mean number of cases per respondent observed in the 9 geographic divisions did not differ significantly. The 166 respondents estimated that 309 (15.6%) of their CDAD cases experienced one or more of the 4 complications described above. Their frequency exhibited some geographic variation (Table 3). For example, the proportion of cases requiring colectomy for toxic megacolon ranged from 0% in the Mountain Division to 7.6% in the East South Central Division.
Between 41% and 65% of the 166 infectious disease consultants providing data for the period from May to November 2004 perceived changes in the epidemiology of CDAD (Table 4). Overall, 68 (41.0%) reported increased number of cases, 76 (45.8%) more severe cases, 107 (64.5%) more recurrences, and 80 (48.2%) more refractoriness to therapy. Perceptions varied geographically; however, more than 40% of respondents in all 9 divisions answered affirmatively to one or more of the questions about perceptions of increased caseloads, increased severity, increased recurrences, or increased refractoriness (Table 4).
First Survey Versus Second Survey: Experience and Perceptions
Data from the 90 EIN members who provided complete responses to both surveys lacked uniformity. Some reported increased caseloads of CDAD on the first survey but not on the second. Similarly, some respondents who reported increased caseloads on one or both surveys did not report increased frequencies of complications, relapses, or refractoriness to therapy on either survey. Nevertheless, many of the 90 infectious diseases consultants reported noteworthy changes in their experience. For example, 21 respondents who reported no increase in caseload on the first survey but reported an increase in the 6 months before the second survey collectively estimated seeing 292 cases in the second period (an average of 14 cases/respondent) compared with 201 cases in the first period (9.6 cases/respondent). Similarly, 23 respondents not reporting more severe or fulminant CDAD on the first survey but who did so on the second survey collectively reported 60 cases (2.6 cases/respondent), with one of the 4 complications during the second survey period. These same respondents had reported only 25 complicated cases (1.1 case/respondent) during the first survey period.
Diagnostic and Therapeutic Issues Addressed in Both Surveys
Respondents to both surveys indicated that enzyme immunoassays for toxin A or both toxins A and B were the most readily available tests in their hospitals for diagnosing CDAD. In the first survey, 393 (95%) of 411 respondents indicated that their hospital laboratories performed these tests. In contrast, less than 15% of respondents indicated that their laboratories performed cell cytotoxin assays or stool cultures for C. difficile. The second survey yielded comparable findings.
The results of both surveys indicated that some IDCs recently found CDAD more difficult to diagnose. In the initial survey, 74 (16%) of EIN members reported increased difficulty in diagnosis of CDAD cases during the previous 6 months. Of these respondents, 39 (53%) reported using endoscopy for diagnosis; 22 (30%), sending specimens to other laboratories for testing; and 67 (91%), resorting to empiric treatment as a therapeutic trial. The second survey yielded similar responses.
The results of both surveys also indicated that EIN members have experienced difficulties in treating patients with CDAD during the prior 6-month period. Of 140 respondents reporting more refractory CDAD cases in the first survey, 84 (60%) suspected that antimicrobial resistance was the cause of treatment failure. Furthermore, 53 members (38%) reported using vancomycin, rather than metronidazole, for initial treatment, and 46 (33%) used agents other than metronidazole or vancomycin for therapy. The second survey produced comparable results. Of 80 respondents reporting more refractory disease at the time of the second survey, 37 (46%) suspected antimicrobial resistance; 28 (35%) used vancomycin as initial therapy; and 32 (40%) used alternative therapeutic agents. Despite concerns about resistance, only 8 of 523 EIN members responding to either or both surveys indicated that they had requested susceptibility testing on C. difficile isolates recovered from their patients.
Several reports from North America have documented an increase in the incidence of CDAD during the last few years.2-5,7,8 The most dramatic reports have come from facilities experiencing outbreaks. For example, a hospital in Quebec reported that the incidence of CDAD had increased from 35.6 cases per 100,000 population in 1991 to 156.3 cases per 100,000 in 2003.8 A hospital in Pittsburgh reported nosocomial CDAD cases rising from 2.7 cases per 1000 discharges (0.46 cases/1000 patient care days) in 1999 to 6.8 cases per 1000 discharges (1.12 cases/1000 patient care days) in 2000 to 2001.5 Since 2000, outbreaks also have occurred in Georgia, Illinois, Maine, New Jersey, and Oregon.7
National data, probably more reflective of endemic disease, have also revealed an increasing incidence of CDAD. During the period from 1987 to 2001, hospitals with more than 500 beds participating in the Centers for Disease Control and Prevention's National Nosocomial Infections Surveillance System experienced significant increases in CDAD rates in their intensive care units. During the same period, a significant increase occurred in the nonintensive care unit beds of hospitals with less than 250 beds.2 Moreover, rates of CDAD listed on US hospital discharges doubled from 31 cases per 100,000 population in 1996 to 61 per 100,000 in 2003, with a significant increase detected between 2000 and 2003.11
The EIN surveys provide both confirmation and perspective on these national data. First, 29% and 41% of respondents on the first and second surveys, respectively, reported increases in their CDAD caseloads. Overall, 219 (42%) of the 523 US-based respondents reported an increase in their CDAD caseloads on one or both surveys. Second, varying responses by geographic division and survey period suggest that some of the perceived increase reflected unreported or unrecognized outbreaks.
The EIN surveys also attest to a widespread increase in the severity of CDAD in the United States. Indeed, 29% and 46% of respondents to the first and second surveys, respectively, reported more cases of severe or fulminant CDAD. Overall, 229 (44%) of 523 US-based EIN members reported more cases of severe or fulminant CDAD on one or both surveys. These perceptions and the reported estimates also exhibited some geographic and temporal variation, again mirroring observations from different data sources.5-8 For example, in the Sherbrooke, Quebec, experience, complication rates increased from 7.1% in 1991 to 1992 to 18.2% in 2003.8 Mortality among CDAD cases was significantly higher, especially in the older age groups.12 In 12 Quebec hospitals, the directly attributable CDAD mortality in 2004 was 6.9% compared with 1.5% in 18 Canadian hospitals surveyed in 1997.6 In the Pittsburgh experience, 5.6% of 72 cases with nosocomial CDAD experienced severe complications or death in 1999 (0.15 cases/1000 discharges), whereas 8.8% of 419 cases (0.6 severe cases/1000 discharges) had these outcomes in 2000 to 2001 (P = 0.004).5 Others have reported similar increases in complication rates in recent years.3,4
Several investigators have offered convincing evidence that a new strain of C. difficile accounts for the changing epidemiologic features of CDAD.7,13-15 The epidemic strain in the Quebec and other North American outbreaks, characterized as toxinotype III, North American pulsed-field gel electrophoresis type 1, polymerase chain reaction ribotype 027, and restriction endonuclease analysis group BI, exhibits an 18-base pair deletion in a putative toxin down-regulator gene tcdC and carries the binary toxin gene cdtB.7,13,14 According to the results of an in vitro study, these strains produce 16 to 23 times the concentrations of toxins A and B produced by other common pulsed-field gel electrophoresis types,13,14 and they are consistently resistant to fluoroquinolone antibiotics, including gatifloxacin and moxifloxacin.6,7 Appearance of this strain in new geographic areas where fluoroquinolone selection pressures are high may account for recent increases in the incidence of CDAD in North America and higher national case rates. Increased severity of illness may be related to unusual virulence properties such as increased production of toxins A and B or the presence of binary toxin, but these properties were also found in non-outbreak toxinotype III US C. difficile isolates found in the 1980s and early 1990s.7 The results of the EIN survey suggest that this or possibly other virulent strains may be widely distributed throughout the United States. Because relatively few hospital laboratories perform stool cultures for C. difficile, surveillance to detect different strain types is not presently possible.
Although concerns about the adequacy of current therapies for CDAD have received renewed attention in 2005,16 the relationship of therapeutic difficulties to the binary toxin-producing strain is unclear. Between 2002 and 2003 to 2004, investigators in Quebec noted a doubling of the proportion of patients switched from metronidazole to vancomycin because of a poor response to the former.17 In 2003 to 2004, they also noted that the proportion of patients relapsing after a course of metronidazole therapy rose from 20.8% to 47.2%. Notwithstanding, controversy about the adequacy of metronidazole therapy has existed for many years,16 and other reports impugning its value have arisen from areas not involved in the current outbreaks. For example, Musher et al reported that, in a series of 207 patients treated with metronidazole in Houston, poor responses to therapy were found in 22% and relapse within 90 days in 28%.18
Results of the EIN survey suggest that concerns about the efficacy of metronidazole therapy are common and pervasive. Thirty-one percent and 48.2% of respondents on the first and second surveys, respectively, reported more CDAD cases refractory to therapy. On one or both surveys, 244 (47%) of 523 US-based respondents reported increased CDAD cases refractory to their usual therapeutic approach. An even higher proportion, 285 (54%) of 523 respondents, had encountered more recurrent cases. Regrettably, only 8 EIN members had submitted C. difficile isolates for susceptibility testing, and none commented on the results. Accordingly, the cause of the therapeutic difficulties encountered remains unknown.
A number of limitations in the EIN surveys require acknowledgment. First, EIN members may not adequately represent IDCs in the United States, and the response rates for the 2 surveys were only 54% and 44%. Notwithstanding, 523 IDCs practicing in all 50 states did respond to one or both surveys. Second, survey responses may have included data from more than one individual in the same practice. This seems unlikely because many EIN members have indicated that a designated person responds for the group. Third, the survey did not attempt to define terms such as "severe or fulminant disease," "CDAD more refractory to therapy," or "more recurrent disease." Fourth, numbers reported by EIN members were not validated, and data from previous years are not available for comparison with current findings. Finally, members may have made rough estimates on the day that they completed their surveys; hence, the potential role of recall bias cannot be discounted.
Despite these limitations, the results of the EIN surveys complement those obtained from outbreak reports and other surveillance systems. Taken as a whole, they are consistent with widespread dissemination of at least one virulent strain into many practice settings. They raise questions about current approaches to surveillance, diagnosis, therapy, and prevention. Additional studies are warranted to better characterize current problems with C. difficile to improve management of CDAD.
1. Bartlett JG. Antibiotic-associated diarrhea. N Engl J Med. 2002;346:334-339.
2. Archibald LK, Banerjee SN, Jarvis WR. Secular trends in hospital-acquired Clostridium difficile disease in the US, 1987-2001. J Infect Dis. 2004;189:1585-1589.
3. Morris AM, Jobe BA, Stoney M, et al. Clostridium difficile colitis: an increasingly aggressive disease? Arch Surg. 2002;137:1096-1100.
4. Dallal RM, Harbrect BG, Boujoukas AJ, et al. Fulminant Clostridium difficile: an underappreciated and increasing cause of death and complications. Ann Surg. 2002;235:363-372.
5. Muto CA, 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. 2005;26:273-280.
6. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi-institutional outbreak of Clostridium difficile-associated disease with high morbidity and mortality. N Engl J Med. 2005;353:2442-2449.
7. McDonald LC, Killgore GE, Thompson A, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med. 2005;353:2433-2441.
8. Pépin J, Valiquette L, Alary M-E, et al. Clostridium difficile-associated diarrhea in a division of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ. 2004;171:466-472.
9. Excecutive Committee of the Infectious Diseases Society of America Emerging Infections Network. The Emerging Infections Network: a new venture for the Infectious Diseases Society of America. Clin Infect Dis. 1997;25:34-36.
11. McDonald LC, Banerjee S, Jernigan DB. Increasing incidence of Clostridium difficile-associated disease in U.S. acute care hospitals, 1993-2001 (Abstract 67). In: Proceedings of the 14th Annual Scientific Meeting of the Society for Healthcare Epidemiology of America (Philadelphia, PA). Alexandria, VA: The Society for Healthcare Epidemiology of America; 2004:32.
12. Pépin J, Valiquette L, Cossette B. Mortality attributable to nosocomial Clostridium difficile-associated disease during an epidemic caused by a hypervirulent strain in Quebec. CMAJ. 2005;173:1037-1042.
13. 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. 2005;366:1079-1084.
14. Norén T. Outbreak from a high-toxin intruder: Clostridium difficile. Lancet. 2005;366:1053-1054.
15. 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. 2005;40:265-272.
16. Gerding DN. Metronidazole for Clostridium difficile-associated disease: is it okay for mom? Clin Infect Dis. 2005;40:1598-1600.
17. Pépin J, Alary M-E, Valiquette L, et al. Increasing risk of relapse after treatment of Clostridium difficile colitis in Quebec, Canada. Clin Infect Dis. 2005;40:1591-1597.
18. Musher DM, Aslam S, Logan N, et al. Relatively poor outcome after treatment of Clostridium difficile colitis with metronidazole. Clin Infect Dis. 2005;40:1586-1590.
© 2006 Lippincott Williams & Wilkins, Inc.