American Journal of Emergency Medicine 33 (2015) 238–243
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Prevalence of Clostridium difficile infection presenting to US EDs☆,☆☆ Aaron M. Smith, MD a, Brandon A. Wuerth, MD a, Timothy L. Wiemken, PhD, MPH b, Forest W. Arnold, DO, MSc b,⁎ a b
School of Medicine, University of Louisville, Louisville, KY, USA Division of Infectious Diseases, Department of Medicine, School of Medicine, University of Louisville, Louisville, KY, USA
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 7 November 2014 Accepted 20 November 2014
a b s t r a c t Objective: The objective of the study is to determine the prevalence of Clostridium difficile infection (CDI) presenting to emergency departments (EDs) in the United States. Secondary objectives included defining the burden of CDI. Methods: This is a retrospective, observational cohort study of 2006-2010 Nationwide Emergency Department Sample database of 980 US hospital EDs in 29 states. Prevalence, mortality rate, length of stay, hospital charges, and endemicity were measured. Results: A total of 474513 patients with CDI-related ED visits were identified. From 2006 to 2010, the prevalence of CDI increased from 26.2 to 33.1 per 100 000 population (P b .001). The number of CDI-related ED cases increased 26.1% (P b .001) over the study period: 18.6% from 2006 to 2007 (P b .001), 4.3% from 2007 to 2008 (P = .46), 1.8% from 2008 to 2009 (P = .73), and 0.13% from 2009 to 2010 (P = .95). Emergency department visits occurred more frequently with individuals 85 years or older (relative risk [RR], 13.74; P b .001), females (RR, 1.77; P b .001) and in the northeast United States (RR, 1.42; P b .001). From 2009 to 2010, the mortality rate decreased 17.9% (P = .01). Conclusions: The prevalence of CDI presenting to EDs increased each year from 2006 to 2010; however, the rate of increase slowed from each year to the next. The mortality rate increased from 2006 to 2009 and decreased significantly from 2009 to 2010. C difficile infection visits presenting to EDs occurred more frequently with older individuals, females, and in the northeast. © 2014 Elsevier Inc. All rights reserved.
1. Introduction 1.1. Background Infection due to Clostridium difficile now rivals methicillin-resistant Staphylococcus aureus as the most common etiology of health care– acquired infections in the United States [1]. It is a common cause of diarrhea traditionally known to be acquired only in the health care setting [2]. The more severe manifestation of C difficile infection (CDI) is pseudomembranous colitis, which can lead to toxic megacolon, colonic perforation, or death [3]. In fact, more deaths have been attributed to CDI than all other intestinal infections combined [4]. The prevalence of hospital-acquired CDI has been on the rise since the mid to late 1990s [5-8]. A Healthcare Cost and Utilization Project (HCUP) study using the Nationwide Inpatient Sample found that over
☆ Financial disclosures: This project had no sponsor and was not supported financially. The authors received no funding. The authors had no conflicts of interest. ☆☆ This manuscript was presented in part at IDWeek, San Diego, CA, October 17 to 21, 2012. ⁎ Corresponding author at: Division of Infectious Diseases, Department of Medicine, University of Louisville, University of Louisville Hospital Epidemiologist, 501 E. Broadway, Suite 120, Louisville, KY 40292. Tel.: +1 502 562 3473; fax: +1 502 852 1147. E-mail address: [email protected] (F.W. Arnold). http://dx.doi.org/10.1016/j.ajem.2014.11.023 0735-6757/© 2014 Elsevier Inc. All rights reserved.
the 8-year period from 1993 to 2001, the number of hospital stays with a principal diagnosis of CDI steadily increased 51% [6]. However, over the following 8-year period (2001-2009), the number of hospital stays with a principal diagnosis of CDI doubled. 1.2. Importance Despite historically being considered to occur strictly in the inpatient setting, recent studies have shown that 35% to 56% of CDI cases are actually diagnosed in outpatients and are classified as community onset, accordingly [9-12]. With the spread of CDI outside the hospital environment, it is important to track the changes of this pathogen presenting to emergency departments (EDs). Community-onset CDI (CO-CDI) has been shown to lack the classic risk factors for CDI—inpatient hospitalization of increased hospital stay and antimicrobial use [3,13]. Two studies conducted among outpatients with CDI found that, in the preceding month, only 27% to 33% had been hospitalized and only 63% to 70% had taken antimicrobials, indicating that these risk factors are associated with a lower proportion of cases than classic CDI [10,14]. The prevalence of CO-CDI is unclear, as it has only been assessed specifically in a limited number of studies with small numbers of cases [15-17]. Studies specifically looking at CO-CDI presenting to EDs
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
are also limited [18-21]. Thus, there is a need for a large comprehensive study to provide information including prevalence, pertaining to CDI presenting to US EDs, we term ED-CDI. 1.3. Goals of this investigation The primary objectives of this study were to determine the prevalence and trends of ED-CDI. Secondary objectives included defining the burden ED-CDI has on the health care system with regard to mortality, length of stay (LOS), and hospital charges.
239
2.2. Nationwide estimates The HCUP databases are based on statewide data collected by individual data organizations that provide it to AHRQ through the HCUP partnership. In total, this collection contains an approximately 20% sample of all US hospital-based EDs [23]. Based on the 20% sample, AHRQ calculated national estimates by weighting each discharge based upon hospital location, teaching status, bed size, and ownership control relative to that of all US hospitals. This made the estimates in the database more representative of all hospital-based EDs in the United States. 2.3. Selection of cases
2. Methods 2.1. Study design and setting This was a retrospective cohort study examining ED visits for CDI using discharge data from the publicly available Nationwide Emergency Department Sample (NEDS) database. It was developed as part of the HCUP by the Agency for Healthcare Research and Quality (AHRQ) [22]. The database contains between 26 and 29 million ED visits each year incorporating data from 28 states and 961 hospitals over the course of the 4-year study period, 2006 to 2010. Available variables in the limited public database include age, sex, insurance status, US region, hospital ownership, teaching status, and trauma center designation. Because NEDS only contains deidentified public data, this study received exemption from institutional review board approval. Data come from community hospitals in the HCUP that are defined as short term, nonfederal, general, and “other,” while excluding hospital units of institutions such as prisons. Pediatric, public, and academic medical hospitals are included in HCUP, whereas long-term care centers, rehabilitation, psychiatric, and chemical dependency hospitals are excluded. A list of the HCUP data partners that contributed to the database can be found at www.hcup-us.ahrq.gov/hcupdatapartners.jsp.
We queried the HCUP NEDS database using International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM), code 008.45 (intestinal infection due to C difficile) for 2006 to 2010. The HCUP has an online query system (www.HCUPnet.ahrq.org) called HCUPnet for limited instant access to HCUP's publically available databases. We obtained patient data with CDI listed as the principal diagnosis, indicating the patient was in the ED primarily due to CDI. 2.4. Methods of measurement An ED-CDI case was defined as an ED visit in the United States that received a principal diagnosis of CDI. The unit of analysis was each ED encounter (including the hospital stay if admitted). Regions of the United States were outlined by the Census Bureau (Fig. 1). Hospital charges were the amount that the hospital charged (not the amount reimbursed) for the entire hospital stay including the ED visit. The charges did not include physician fees. Charges were reported only for those cases when an ED visit resulted in an inpatient stay. Length of stay was defined as the number of nights a patient was hospitalized, not counting the time in the ED. Mortality was defined as death in the ED or as an inpatient for the hospitalization associated with the ED visit.
Fig. 1. US Census Bureau regions for data stratification [24].
240
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
2.5. Data analysis From the database, we recorded the number of ED visits, number of deaths, LOS, and hospital charges associated with CDI. The data were stratified by age group, sex, and geographical region. Population estimates were obtained from US census data and stratified similarly by age group, sex, and geographical region [25]. Using the extrapolated national ED visits from the NEDS database in the numerator and the population estimates from the US Census Bureau for the particular year of interest in the denominator, the prevalence of ED-CDI was calculated within the US population. Relative risks (RRs) were calculated to communicate the ED-CDI risk that one population (eg, female) had compared to another (eg, male). Odds ratios, on the other hand, were not used because the ED-CDI would be confusingly interpreted as a certain odds that a patient belongs to a certain population (eg, female). Similar to prevalence, we calculated the annual mortality rates using the number of deaths due to ED-CDI in the numerator and the total US Census Bureau population estimate for the year of interest in the denominator. To test whether apparent differences in the prevalence and mortality rates over the study period were significant, the SEs AHRQ provided (calculated using SUDAAN software) were entered into HCUP's online Z-test calculator to yield probability values [26]. To test whether trends in the number of cases and deaths over the study period were significant in proportion to all patients presenting to the ED, χ 2 test for trend was performed using R version 3.1. Extrapolated HCUP national estimates were used for these calculations as sample data are not available with HCUP. All other descriptive statistics were completed using Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA). SEs for LOS and hospital charges were those provided through HCUPnet. 3. Results From 2006 to 2010, there were 474513 cases of CDI presenting to US EDs with a prevalence of 31.2 cases per 100000 population over the entire period. The prevalence of ED-CDI increased 26.1% (P b .001) over the study period: 18.6% from 2006 to 2007 (P b .001), 4.3% from 2007 to 2008 (P = .46), 1.8% from 2008 to 2009 (P = .73), and 0.13% from 2009 to 2010 (P = .95). χ2 Test for trend in cases over the entire period yielded a P value of less than .001. The annual prevalence of ED-CDI is outlined in Table 1. The prevalence is also plotted by sex (Fig. 2A) and region (Fig. 2B). C difficile infection presenting to US EDs occurred more frequently among patient 85 years or older, females, and in the northeast. Individuals 85 years or older were found to have a risk 13.7 times that of individuals younger than 85 years (RR, 13.74; 95% confidence interval [CI], 13.64-13.83; P b .001), females were found to have a risk 1.8 times that of males (RR, 1.77; 95% CI, 1.76-1.78; P b .001), and patients in the northeast were found to have a risk 1.4 times that of individuals elsewhere (RR, 1.42; 95% CI, 1.41-1.43; P b .001). From 2006 to 2010, the mortality rate of ED-CDI was approximately 9 per 1 000 000. From 2006 to 2007, the mortality rate increased 34.1% (P b .001); from 2007 to 2008, it increased 1.9% (P = .81); from 2008 to 2009, it increased 3.6% (P = .65); and it decreased 17.9% (P = .01) from 2009 to 2010. χ 2 Test for trend in deaths over the entire period Table 1 The prevalence (cases/100000 population) by age among patients with CDI presenting to an ED in the United States from 2006 to 2010
All ED visits Age (y) 0-17 18-44 45-64 65-84 ≥85
2006
2007
2008
2009
2010
26.2
31.1
32.4
33.0
33.1
2.6 7.1 21.0 117.7 286.5
3.4 7.9 24.3 137.9 345.1
3.2 8.6 25.7 138.8 353.8
3.0 8.8 26.7 136.9 378.1
3.4 9.7 27.5 130.0 368.8
yielded a P value of .001. The CDI mortality rate from 2006 to 2010 was 10.76 deaths per million population. The CDI mortality rate is presented by age (Table 2) and by sex and region (Fig. 3A and B, respectively). The mortality trends mirrored the prevalence and were greater in those 85 years or older, females, and in the northeast. A total of 92.0% were admitted to the hospital, and 47.8% of those admitted were in the 65- to 84-year age group. The LOS after hospital admission remained constant through the 2006 to 2010 study period with a median LOS of 5 days (Table 2), mean of 7 days, and an SE of the mean of 0.1 days. The median ED and hospital charges of patients with ED-CDI (Table 3) were $21596 with a mean of $34576 and an SE of the mean of $1013. The median ED and hospital charges due to ED-CDI increased 26.9% over the study period: 8.6% from 2006 to 2007, 11.9% from 2007 to 2008, 2.4% from 2008 to 2009, and 2% from 2009 to 2010. Compared to all patients who visited the ED and were subsequently admitted, patients who had CDI and were seen and admitted from the ED had an LOS that was 2 days longer (P b .001) and charges that were $3200 more (P b .01).
4. Discussion This study, using the statistical power from a massive database, found that the prevalence and outcomes for patients with CDI presenting to EDs increased from 2006 to 2010. The prevalence for every demographic group measured (age, sex, and region) also increased. While the prevalence of ED-CDI cases increased each year, the rate of increase slowed from each year to the next. The mortality rate increased the most (34.1%) from 2006 to 2007 and increased for the next 2 years, although at much lower rates, then actually decreased from 2009 to 2010, which was found to be statistically significant. Hospital charges rose, whereas change in LOS was equivocal. It is important to know to what extent CDI is present in the community and presenting to EDs. This allows health care systems to know the resources that are required to manage patients who have it. To complicate infection control and prevention efforts, an overwhelming majority of patients with CDI presenting to EDs (92%) are eventually hospitalized, allowing for potential transmission to other inpatients. With a high mortality rate, the risk factors and severity of CO-CDI described in the present study need to be recognized by ED physicians and managed appropriately. In addition, proactive measures for contact isolation need to be considered regarding patients admitted from the ED with diarrhea. Only 1 previous study was identified directly calculating the incidence of CO-CDI, rather than the subset of community-associated CDI [17]. This study was a 1999 United Kingdom randomized, prospective case-control study consisting of urban (Leeds, UK) and semirural (Truro and Cornwall, UK) cohorts, each of 1000 patients, which calculated incidences of 29.5 and 20.2 per 100000 population, respectively [17]. A previous study looking at hospital costs showed an attributable cost of CDI to be around $2454 to $3240, a finding similar to ours; however, we examined hospital charges, not costs [27]. Another previous study coincided with ours, which found a mean LOS of 6.4 days [28]. Previous studies conducted in the United States looking specifically at ED-CDI all had a similar methodology, taking stool samples from patients presenting to the ED with diarrhea [18,20,21]. Two performed before 2005 in the pediatric population found 6.7% and 5.5% of the stool samples tested positive for CDI based on cytotoxicity assay [20,21]. The second study found that CDI positivity was similar in both cases and controls [21]. A study conducted in South Korea from 2006 to 2010 found that 36 (13.2%) of 272 diarrheic patients tested positive for C difficile and that rate increased year to year most notably 16.7% in 2010 [19]. That study also found a male-to-female ratio of 17:15 and a mean age of 55 years. A similar study conducted in California from 2001 to 2002 found 18 culture-positive cases of CDI from 56 stool samples and found a male-to-female ratio of 11:7 and a mean age of 53.7 years [18]. Females have been found to have higher
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
241
Fig. 2. A, The prevalence (cases/100000 population) by sex of CDI among patients presenting to an ED in the United States from 2006 to 2010. B, The prevalence (cases/100000 population) by region of CDI among patients presenting to an ED in the United States from 2006 to 2010.
prevalence of both community-acquired and hospital-acquired CDI than males, coinciding with our findings of a female predominance [29]. The present study analyzed a much larger, more current national ED discharge database to determine the magnitude of and epidemiologic factors associated with CDI presenting to US EDs. It also estimated the national prevalence, mortality rate, and resource burden of CDI presenting to EDs, while tracking these characteristics over time to identify trends. No major change in diagnostic methods or treatment modalities of CDI occurred during the study period. The first commercially available real-time polymerase chain reaction assay was approved by the Food and Drug Administration in December 2008; however, C difficile enzyme Table 2 Mortality rate and LOS among patients with CDI presenting to an ED in the United States from 2006 to 2010 LOS after hospital admission (median days)
Death during hospital stay after ED visit (mortality rate/million population)
2006 2007 2008 2009 2010 2006 All ED visits Age (y) 0-17 18-44 45-64 65-84 ≥85
5
5
5
5
5
3 3 5 6 6
3 4 5 6 6
3 4 5 6 6
3 4 5 6 6
3 4 4 5 6
8.6
2007
2008
2009
2010
11.5
11.7
12.1
9.9
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 2.6 3.7 3.6 3.0 42.5 60.7 54.8 56.6 45.3 195.8 229.7 242.4 262.5 222.5
immunoassay of toxin in stool was likely the most widely used method of diagnosis due to its rapid response and lower cost during the study period [30]. According to an article in the College of American Pathologists online newsletter in November 2010, the author states that more than 90% of C difficile testing was still done by enzyme immunoassay [31]. Oral vancomycin and metronidazole have been the drugs of choice for CDI, whereas immunotherapy and fecal bacteriotherapy are rare treatments reserved for relapsing patients. Fidaxomicin had not yet been approved for use during the study period. Other medicines, such as probiotics, toxin-binding agents, and rifaximin, are complimentary agents rather than alternatives. Because the diagnostic measures and treatments primarily used across the United States were essentially unchanged during our study period, increases found in CDI presenting to EDs were likely due to a variable other than how CDI was diagnosed or treated, such as education, awareness, prevalence of antibiotic exposure, and availability of testing. One possible explanation of the increasing prevalence and severity of ED-CDI is a change in the pathogen itself. Studies have shown the emergence of a more virulent and resistant, toxigenic strain called BI/ NAP1/027, with reports of it even in the community [8,32-38]. This strain has been reported in the United States, Canada, Australia, and Asia as well as in European countries such as the UK, Belgium, France, and the Netherlands [8,32,34,35,37]. We found the largest increase in prevalence to occur in the northeastern region of the United States; reasons for this are unclear but may be associated with an epicenter of the BI/NAP/027 strain located in Montreal, Quebec, Canada [8].
242
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
Fig. 3. A, Mortality rates (deaths per million population) by sex of CDI among patients presenting to an ED in the United States from 2006 to 2010. B, Mortality rates (deaths per million population) by region of CDI among patients presenting to an ED in the United States from 2006 to 2010.
Additional hypervirulent strains, NAP7 and NAP8/078/V, have also been increasing in prevalence in the United States, Canada, and Europe over recent years [39-41]. These cases have been highly associated with CO-CDI. Other potential factors contributing to the increasing prevalence of ED-CDI presenting to EDs include increased testing due to
Table 3 Emergency department and hospital charges (median charge in dollars [$]) among patients with CDI in the United States from 2006 to 2010 2006
All ED visits Age (y) 0-17 18-44 45-64 65-84 ≥85 Sex Male Female Region Northeast Midwest South West
2007
2008
2009
2010
US $
US $
US $
US $
US $
18 426
20 012
22 396
22 932
23 386
10 194 15 338 17 192 20 030 18 588
12 216 16 595 19 137 21 193 20 848
13 343 18 136 21 339 24 088 22 809
13 591 18 811 21 456 24 568 24 097
13 770 19 370 22 539 24 959 24 780
18 282 18 505
19 886 20 098
22 710 22 166
22 987 22 898
23 535 23 270
21 617 14 146 17 528 25 088
21 577 15 868 19 436 30 122
24 139 17 106 21 478 33 791
23 547 18 505 22 565 36 823
24 189 19 778 22 375 32 755
increased physician awareness and asymptomatic carriage in the community [13,17,42].
4.1. Limitations The present study was limited by the use of ICD-9-CM coding as the case identifier, potentially allowing misclassification into our results. We estimate that this level of misclassification was minimal due to several studies that found ICD-9-CM code 008.45 to be an accurate way to identify CDI [24,43,44]. Individuals who presented to an ED and were discharged with a secondary diagnosis of CDI were not included in this study, as we only included those cases with CDI listed as the primary diagnosis. We calculated that primary diagnoses of CDI only accounted for 37% of all listed diagnoses of CDI. Solely using visits with CDI listed as the principal diagnosis may have caused an underestimation of the prevalence and mortality of our results (a potential error of omission; α error). Unique patient identifiers were not available, and the unit of analysis for HCUP data was the hospital encounter (ie, the ED visit or the ED visit in addition to the hospital stay), not each individual patient. Thus, multiple visits by the same individual may have been counted more than once. This would have caused an overestimation of the prevalence and mortality of our results (a potential error of commission; β error). In addition, deaths that occurred during a hospitalization with a primary diagnosis of CDI do not necessarily mean that CDI was the cause of the death, which our results assumed. However, using
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
only those cases with a primary diagnosis of CDI, we believe, decreased the severity of this limitation. Lastly, using the publically available database limited the variables in which we could analyze and limited our ability to calculate adjusted risk ratios.
[20]
[21]
5. Conclusion [22]
The prevalence of CDI presenting to US EDs increased in the United States from 2006 to 2010, albeit, at a slowing rate from year to year. The mortality rate also increased over the study period but did not reach significance. Emergency department visits occurred more frequently in older individuals, in females, and in the northeast United States. Acknowledgments The authors are grateful to L. Clifford McDonald, MD, FACP, FSHEA, for previewing the manuscript. No compensation was provided.
[23]
[24]
[25]
[26]
References [27] [1] Miller BA, Chen LF, Sexton DJ, Anderson DJ. Comparison of the burdens of hospitalonset, healthcare facility-associated Clostridium difficile infection and of healthcareassociated infection due to methicillin-resistant Staphylococcus aureus in community hospitals. Infect Control Hosp Epidemiol 2011;32:387–90. [2] Kelly CP, Pothoulakis C, LaMont JT. Clostridium difficile colitis. N Engl J Med 1994;330: 257–62. [3] Centers for Disease Control & Prevention. Severe Clostridium difficile–associated disease in populations previously at low risk—four states, 2005. MMWR Morb Mortal Wkly Rep 2005;54:1201–5. [4] Redelings MD, Sorvillo F, Mascola L. Increase in Clostridium difficile–related mortality rates, United States, 1999-2004. Emerg Infect Dis 2007;13:1417–9. [5] Chandler RE, Hedberg K, Cieslak PR. Clostridium difficile–associated disease in Oregon: increasing incidence and hospital-level risk factors. Infect Control Hosp Epidemiol 2007;28:116–22. [6] Elixhauser A, Jhung M. Clostridium difficile–associated disease in U.S. hospitals, 1993-2005: Statistical Brief #50. Rockville (MD): Healthcare Cost and Utilization Project (HCUP) Statistical Briefs; 2006. [7] McDonald LC, Owings M, Jernigan DB. Clostridium difficile infection in patients discharged from US short-stay hospitals, 1996-2003. Emerg Infect Dis 2006;12:409–15. [8] Pepin J, Valiquette L, Alary ME, Villemure P, Pelletier A, Forget K, et al. Clostridium difficile–associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004;171:466–72. [9] Dumyati G, Stevens V, Hannett GE, Thompson AD, Long C, MacCannell D, et al. Community-associated Clostridium difficile infections, Monroe County, New York, USA. Emerg Infect Dis 2012;18:392–400. [10] Kuntz JL, Johnson ES, Raebel MA, Petrik AF, Yang X, Thorp M, et al. Epidemiology and healthcare costs of incident Clostridium difficile infections identified in the outpatient healthcare setting. Infect Control Hosp Epidemiol 2012;33:1031–8. [11] Kutty PK, Woods CW, Sena AC, Benoit SR, Naggie S, Frederick J, et al. Risk factors for and estimated incidence of community-associated Clostridium difficile infection, North Carolina, USA. Emerg Infect Dis 2010;16:197–204. [12] Naggie S, Frederick J, Pien BC, Miller BA, Provenzale DT, Goldberg KC, et al. Community-associated Clostridium difficile infection: experience of a veteran affairs medical center in southeastern USA. Infection 2010;38:297–300. [13] Bauer MP, Goorhuis A, Koster T, Numan-Ruberg SC, Hagen EC, Debast SB, et al. Community-onset Clostridium difficile–associated diarrhoea not associated with antibiotic usage—two case reports with review of the changing epidemiology of Clostridium difficile–associated diarrhoea. Neth J Med 2008;66:207–11. [14] Hirshon JM, Thompson AD, Limbago B, McDonald LC, Bonkosky M, Heimer R, et al. Clostridium difficile infection in outpatients, Maryland and Connecticut, USA, 20022007. Emerg Infect Dis 2011;17:1946–9. [15] Fellmeth G, Yarlagadda S, Iyer S. Epidemiology of community-onset Clostridium difficile infection in a community in the South of England. J Infect Public Health 2010;3:118–23. [16] Stergachis A, Levy DG, McFarland LV, Van Vorst K, Graham DJ, Johnson ES, et al. Antibiotics and Clostridium difficile diarrhea in the ambulatory care setting. Clin Ther 2000;22:91–102. [17] Wilcox MH, Mooney L, Bendall R, Settle CD, Fawley WN. A case-control study of community-associated Clostridium difficile infection. J Antimicrob Chemother 2008; 62:388–96. [18] Martirosian G, Szczesny A, Silva J. Clostridium difficile in emergency room. Anaerobe 2005;11:258–61. [19] Shin BM, Moon SJ, Kim YS, Shin WC, Yoo HM. Characterization of cases of Clostridium difficile infection (CDI) presenting at an emergency room: molecular and clinical fea-
[28]
[29]
[30] [31] [32]
[33]
[34]
[35]
[36]
[37] [38]
[39]
[40]
[41]
[42]
[43]
[44]
243
tures differentiate community-onset hospital-associated and community-associated CDI in a tertiary care hospital. J Clin Microbiol 2011;49:2161–5. Klein EJ, Boster DR, Stapp JR, Wells JG, Qin X, Clausen CR, et al. Diarrhea etiology in a children's hospital emergency department: a prospective cohort study. Clin Infect Dis 2006;43:807–13. Denno DN, Shaikh N, Stapp JR, Qin X, Hutter CM, Hoffman V, et al. Diarrhea etiology in a pediatric emergency department: a case control study. Clin Infect Dis 2012;55: 897–904. HCUPnet, Healthcare Cost and Utilization Project (HCUP). Nationwide Emergency Department Sample (NEDS). Rockville, MD: Agency for Healthcare Research and Quality (AHRQ); 2006–2010. Population Estimates, Historical Data. United States Census Bureau Website. http:// www.census.gov/popest/data/national/totals/2013/index.html. [Accessed May 15, 2013. Updated December 30, 2013]. Jhung MA, Thompson AD, Killgore GE, Zukowski WE, Songer G, Warny M, et al. Toxinotype V Clostridium difficile in humans and food animals. Emerg Infect Dis 2008;14:1039–45. Healthcare Cost and Utilization Project (HCUP). Healthcare Cost and Utilization Project (HCUP), Nationwide Emergency Department Sample (NEDS) and Agency for Healthcare Research and Quality (AHRQ). Website. Rockville, MD http://www. hcup-us.ahrq.gov/nedsoverview.jsp. [Accessed September 16, 2014. Updated June 24, 2014]. Z-Test calculator. Healthcare Cost and Utilization Project (HCUP) hyperlink to Z-Test Calculator Website. http://hcupnet.ahrq.gov/ZTestCalc.jsp?Now=1389033854065&JS=Y. [Accessed May 15, 2013]. Dubberke ER, Reske KA, Olsen MA, McDonald LC, Fraser VJ. Short and long-term attributable costs of Clostridium difficile–associated disease in nonsurgical inpatients. Clin Infect Dis 2008;46:497–504. O'Brien JA, Lahue BJ, Caro JJ, Davidson DM. The emerging infectious challenge of Clostridium difficile–associated disease in Massachusetts hospitals: clinical and economic consequences. Infect Control Hosp Epidemiol 2007;28:1219–27. Khanna S, Pardi DS, Aronson SL, Kammer PP, Baddour LM. The epidemiology of community-acquired Clostridium difficile infection: a population-based study. Am J Gastroenterol 2012;107:89–95. Dubberke ER, Reske KA, McDonald LC, Fraser VJ. ICD-9 codes and surveillance for Clostridium difficile–associated disease. Emerg Infect Dis 2006;12:1576–9. Scheurer DB, Hicks LS, Cook EF, Schnipper JL. Accuracy of ICD-9 coding for Clostridium difficile infections: a retrospective cohort. Epidemiol Infect 2007;135:1010–3. Schmiedeskamp M, Harpe S, Polk R, Oinonen M, Pakyz A. Use of International Classification of Diseases, Ninth Revision, Clinical Modification codes and medication use data to identify nosocomial Clostridium difficile infection. Infect Control Hosp Epidemiol 2009;30:1070–6. Kvach EJ, Ferguson D, Riska PF, Landry ML. Comparison of BD GeneOhm Cdiff realtime PCR assay with a two-step algorithm and a toxin A/B enzyme-linked immunosorbent assay for diagnosis of toxigenic Clostridium difficile infection. J Clin Microbiol 2010;48:109–14. Merrigan M. C. diff molecular testing with less time and effort. College of American Pathologists Website. http://www.cap.org/apps/cap.portaL?_nfpb=true&cntvwrPtlt_ actionOverride=%2Fportlets%2FcontentViewer%2Fshow&_windowLabel= cntvwrPtlt&cntvwrPtlt%7BactionForm.contentReference%7D=cap_today% 2F1110%2F1110g_c_diff.html&_state=maximized&_pageLabel=cntvwr. [Accessed April 26, 2013. Updated November 2010]. Clements AC, Magalhaes RJ, Tatem AJ, Paterson DL, Riley TV. Clostridium difficile PCR ribotype 027: assessing the risks of further worldwide spread. Lancet Infect Dis 2010;10:395–404. Huang H, Weintraub A, Fang H, Nord CE. Community acquired Clostridium difficile infection due to a moxifloxacin susceptible ribotype 027 strain. Scand J Infect Dis 2009;41:158–9. Kuijper EJ, Coignard B, Tull P. Emergence of Clostridium difficile–associated disease in North America and Europe. Clin Microbiol Infect 2006;12(Suppl. 6):2–18. McDonald LC, Killgore GE, Thompson A, Owens Jr RC, Kazakova SV, Sambol SP, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005; 353:2433–41. Quesada-Gomez C, Vargas P, Lopez-Urena D, Gamboa-Coronado Mdel M, RodriguezCavallini E. Community-acquired Clostridium difficile NAP1/027-associated diarrhea in an eighteen month old child. Anaerobe 2012;18:581–3. Richards M, Knox J, Elliott B, Mackin K, Lyras D, Waring LJ, et al. Severe infection with Clostridium difficile PCR ribotype 027 acquired in Melbourne, Australia. Med J Aust 2011;194:369–71. Warny M, Pepin J, Fang A, Killgore G, Thompson A, Brazier J, et al. Toxin production by an emerging strain of Clostridium difficile associated with outbreaks of severe disease in North America and Europe. Lancet 2005;366:1079–84. Goorhuis A, Bakker D, Corver J, Debast SB, Harmanus C, Notermans DW, et al. Emergence of Clostridium difficile infection due to a new hypervirulent strain, polymerase chain reaction ribotype 078. Clin Infect Dis 2008;47:1162–70. Mulvey MR, Boyd DA, Gravel D, Hutchinson J, Kelly S, McGeer A, et al. Hypervirulent Clostridium difficile strains in hospitalized patients, Canada. Emerg Infect Dis 2010; 16:678–81. Khanna S, Pardi DS, Aronson SL, Kammer PP, Baddour LM. Outcomes in communityacquired Clostridium difficile infection. Aliment Pharmacol Ther 2012;35:613–8.
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Prevalence of Clostridium difficile infection presenting to US EDs☆,☆☆ Aaron M. Smith, MD a, Brandon A. Wuerth, MD a, Timothy L. Wiemken, PhD, MPH b, Forest W. Arnold, DO, MSc b,⁎ a b
School of Medicine, University of Louisville, Louisville, KY, USA Division of Infectious Diseases, Department of Medicine, School of Medicine, University of Louisville, Louisville, KY, USA
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 7 November 2014 Accepted 20 November 2014
a b s t r a c t Objective: The objective of the study is to determine the prevalence of Clostridium difficile infection (CDI) presenting to emergency departments (EDs) in the United States. Secondary objectives included defining the burden of CDI. Methods: This is a retrospective, observational cohort study of 2006-2010 Nationwide Emergency Department Sample database of 980 US hospital EDs in 29 states. Prevalence, mortality rate, length of stay, hospital charges, and endemicity were measured. Results: A total of 474513 patients with CDI-related ED visits were identified. From 2006 to 2010, the prevalence of CDI increased from 26.2 to 33.1 per 100 000 population (P b .001). The number of CDI-related ED cases increased 26.1% (P b .001) over the study period: 18.6% from 2006 to 2007 (P b .001), 4.3% from 2007 to 2008 (P = .46), 1.8% from 2008 to 2009 (P = .73), and 0.13% from 2009 to 2010 (P = .95). Emergency department visits occurred more frequently with individuals 85 years or older (relative risk [RR], 13.74; P b .001), females (RR, 1.77; P b .001) and in the northeast United States (RR, 1.42; P b .001). From 2009 to 2010, the mortality rate decreased 17.9% (P = .01). Conclusions: The prevalence of CDI presenting to EDs increased each year from 2006 to 2010; however, the rate of increase slowed from each year to the next. The mortality rate increased from 2006 to 2009 and decreased significantly from 2009 to 2010. C difficile infection visits presenting to EDs occurred more frequently with older individuals, females, and in the northeast. © 2014 Elsevier Inc. All rights reserved.
1. Introduction 1.1. Background Infection due to Clostridium difficile now rivals methicillin-resistant Staphylococcus aureus as the most common etiology of health care– acquired infections in the United States [1]. It is a common cause of diarrhea traditionally known to be acquired only in the health care setting [2]. The more severe manifestation of C difficile infection (CDI) is pseudomembranous colitis, which can lead to toxic megacolon, colonic perforation, or death [3]. In fact, more deaths have been attributed to CDI than all other intestinal infections combined [4]. The prevalence of hospital-acquired CDI has been on the rise since the mid to late 1990s [5-8]. A Healthcare Cost and Utilization Project (HCUP) study using the Nationwide Inpatient Sample found that over
☆ Financial disclosures: This project had no sponsor and was not supported financially. The authors received no funding. The authors had no conflicts of interest. ☆☆ This manuscript was presented in part at IDWeek, San Diego, CA, October 17 to 21, 2012. ⁎ Corresponding author at: Division of Infectious Diseases, Department of Medicine, University of Louisville, University of Louisville Hospital Epidemiologist, 501 E. Broadway, Suite 120, Louisville, KY 40292. Tel.: +1 502 562 3473; fax: +1 502 852 1147. E-mail address: [email protected] (F.W. Arnold). http://dx.doi.org/10.1016/j.ajem.2014.11.023 0735-6757/© 2014 Elsevier Inc. All rights reserved.
the 8-year period from 1993 to 2001, the number of hospital stays with a principal diagnosis of CDI steadily increased 51% [6]. However, over the following 8-year period (2001-2009), the number of hospital stays with a principal diagnosis of CDI doubled. 1.2. Importance Despite historically being considered to occur strictly in the inpatient setting, recent studies have shown that 35% to 56% of CDI cases are actually diagnosed in outpatients and are classified as community onset, accordingly [9-12]. With the spread of CDI outside the hospital environment, it is important to track the changes of this pathogen presenting to emergency departments (EDs). Community-onset CDI (CO-CDI) has been shown to lack the classic risk factors for CDI—inpatient hospitalization of increased hospital stay and antimicrobial use [3,13]. Two studies conducted among outpatients with CDI found that, in the preceding month, only 27% to 33% had been hospitalized and only 63% to 70% had taken antimicrobials, indicating that these risk factors are associated with a lower proportion of cases than classic CDI [10,14]. The prevalence of CO-CDI is unclear, as it has only been assessed specifically in a limited number of studies with small numbers of cases [15-17]. Studies specifically looking at CO-CDI presenting to EDs
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
are also limited [18-21]. Thus, there is a need for a large comprehensive study to provide information including prevalence, pertaining to CDI presenting to US EDs, we term ED-CDI. 1.3. Goals of this investigation The primary objectives of this study were to determine the prevalence and trends of ED-CDI. Secondary objectives included defining the burden ED-CDI has on the health care system with regard to mortality, length of stay (LOS), and hospital charges.
239
2.2. Nationwide estimates The HCUP databases are based on statewide data collected by individual data organizations that provide it to AHRQ through the HCUP partnership. In total, this collection contains an approximately 20% sample of all US hospital-based EDs [23]. Based on the 20% sample, AHRQ calculated national estimates by weighting each discharge based upon hospital location, teaching status, bed size, and ownership control relative to that of all US hospitals. This made the estimates in the database more representative of all hospital-based EDs in the United States. 2.3. Selection of cases
2. Methods 2.1. Study design and setting This was a retrospective cohort study examining ED visits for CDI using discharge data from the publicly available Nationwide Emergency Department Sample (NEDS) database. It was developed as part of the HCUP by the Agency for Healthcare Research and Quality (AHRQ) [22]. The database contains between 26 and 29 million ED visits each year incorporating data from 28 states and 961 hospitals over the course of the 4-year study period, 2006 to 2010. Available variables in the limited public database include age, sex, insurance status, US region, hospital ownership, teaching status, and trauma center designation. Because NEDS only contains deidentified public data, this study received exemption from institutional review board approval. Data come from community hospitals in the HCUP that are defined as short term, nonfederal, general, and “other,” while excluding hospital units of institutions such as prisons. Pediatric, public, and academic medical hospitals are included in HCUP, whereas long-term care centers, rehabilitation, psychiatric, and chemical dependency hospitals are excluded. A list of the HCUP data partners that contributed to the database can be found at www.hcup-us.ahrq.gov/hcupdatapartners.jsp.
We queried the HCUP NEDS database using International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM), code 008.45 (intestinal infection due to C difficile) for 2006 to 2010. The HCUP has an online query system (www.HCUPnet.ahrq.org) called HCUPnet for limited instant access to HCUP's publically available databases. We obtained patient data with CDI listed as the principal diagnosis, indicating the patient was in the ED primarily due to CDI. 2.4. Methods of measurement An ED-CDI case was defined as an ED visit in the United States that received a principal diagnosis of CDI. The unit of analysis was each ED encounter (including the hospital stay if admitted). Regions of the United States were outlined by the Census Bureau (Fig. 1). Hospital charges were the amount that the hospital charged (not the amount reimbursed) for the entire hospital stay including the ED visit. The charges did not include physician fees. Charges were reported only for those cases when an ED visit resulted in an inpatient stay. Length of stay was defined as the number of nights a patient was hospitalized, not counting the time in the ED. Mortality was defined as death in the ED or as an inpatient for the hospitalization associated with the ED visit.
Fig. 1. US Census Bureau regions for data stratification [24].
240
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
2.5. Data analysis From the database, we recorded the number of ED visits, number of deaths, LOS, and hospital charges associated with CDI. The data were stratified by age group, sex, and geographical region. Population estimates were obtained from US census data and stratified similarly by age group, sex, and geographical region [25]. Using the extrapolated national ED visits from the NEDS database in the numerator and the population estimates from the US Census Bureau for the particular year of interest in the denominator, the prevalence of ED-CDI was calculated within the US population. Relative risks (RRs) were calculated to communicate the ED-CDI risk that one population (eg, female) had compared to another (eg, male). Odds ratios, on the other hand, were not used because the ED-CDI would be confusingly interpreted as a certain odds that a patient belongs to a certain population (eg, female). Similar to prevalence, we calculated the annual mortality rates using the number of deaths due to ED-CDI in the numerator and the total US Census Bureau population estimate for the year of interest in the denominator. To test whether apparent differences in the prevalence and mortality rates over the study period were significant, the SEs AHRQ provided (calculated using SUDAAN software) were entered into HCUP's online Z-test calculator to yield probability values [26]. To test whether trends in the number of cases and deaths over the study period were significant in proportion to all patients presenting to the ED, χ 2 test for trend was performed using R version 3.1. Extrapolated HCUP national estimates were used for these calculations as sample data are not available with HCUP. All other descriptive statistics were completed using Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA). SEs for LOS and hospital charges were those provided through HCUPnet. 3. Results From 2006 to 2010, there were 474513 cases of CDI presenting to US EDs with a prevalence of 31.2 cases per 100000 population over the entire period. The prevalence of ED-CDI increased 26.1% (P b .001) over the study period: 18.6% from 2006 to 2007 (P b .001), 4.3% from 2007 to 2008 (P = .46), 1.8% from 2008 to 2009 (P = .73), and 0.13% from 2009 to 2010 (P = .95). χ2 Test for trend in cases over the entire period yielded a P value of less than .001. The annual prevalence of ED-CDI is outlined in Table 1. The prevalence is also plotted by sex (Fig. 2A) and region (Fig. 2B). C difficile infection presenting to US EDs occurred more frequently among patient 85 years or older, females, and in the northeast. Individuals 85 years or older were found to have a risk 13.7 times that of individuals younger than 85 years (RR, 13.74; 95% confidence interval [CI], 13.64-13.83; P b .001), females were found to have a risk 1.8 times that of males (RR, 1.77; 95% CI, 1.76-1.78; P b .001), and patients in the northeast were found to have a risk 1.4 times that of individuals elsewhere (RR, 1.42; 95% CI, 1.41-1.43; P b .001). From 2006 to 2010, the mortality rate of ED-CDI was approximately 9 per 1 000 000. From 2006 to 2007, the mortality rate increased 34.1% (P b .001); from 2007 to 2008, it increased 1.9% (P = .81); from 2008 to 2009, it increased 3.6% (P = .65); and it decreased 17.9% (P = .01) from 2009 to 2010. χ 2 Test for trend in deaths over the entire period Table 1 The prevalence (cases/100000 population) by age among patients with CDI presenting to an ED in the United States from 2006 to 2010
All ED visits Age (y) 0-17 18-44 45-64 65-84 ≥85
2006
2007
2008
2009
2010
26.2
31.1
32.4
33.0
33.1
2.6 7.1 21.0 117.7 286.5
3.4 7.9 24.3 137.9 345.1
3.2 8.6 25.7 138.8 353.8
3.0 8.8 26.7 136.9 378.1
3.4 9.7 27.5 130.0 368.8
yielded a P value of .001. The CDI mortality rate from 2006 to 2010 was 10.76 deaths per million population. The CDI mortality rate is presented by age (Table 2) and by sex and region (Fig. 3A and B, respectively). The mortality trends mirrored the prevalence and were greater in those 85 years or older, females, and in the northeast. A total of 92.0% were admitted to the hospital, and 47.8% of those admitted were in the 65- to 84-year age group. The LOS after hospital admission remained constant through the 2006 to 2010 study period with a median LOS of 5 days (Table 2), mean of 7 days, and an SE of the mean of 0.1 days. The median ED and hospital charges of patients with ED-CDI (Table 3) were $21596 with a mean of $34576 and an SE of the mean of $1013. The median ED and hospital charges due to ED-CDI increased 26.9% over the study period: 8.6% from 2006 to 2007, 11.9% from 2007 to 2008, 2.4% from 2008 to 2009, and 2% from 2009 to 2010. Compared to all patients who visited the ED and were subsequently admitted, patients who had CDI and were seen and admitted from the ED had an LOS that was 2 days longer (P b .001) and charges that were $3200 more (P b .01).
4. Discussion This study, using the statistical power from a massive database, found that the prevalence and outcomes for patients with CDI presenting to EDs increased from 2006 to 2010. The prevalence for every demographic group measured (age, sex, and region) also increased. While the prevalence of ED-CDI cases increased each year, the rate of increase slowed from each year to the next. The mortality rate increased the most (34.1%) from 2006 to 2007 and increased for the next 2 years, although at much lower rates, then actually decreased from 2009 to 2010, which was found to be statistically significant. Hospital charges rose, whereas change in LOS was equivocal. It is important to know to what extent CDI is present in the community and presenting to EDs. This allows health care systems to know the resources that are required to manage patients who have it. To complicate infection control and prevention efforts, an overwhelming majority of patients with CDI presenting to EDs (92%) are eventually hospitalized, allowing for potential transmission to other inpatients. With a high mortality rate, the risk factors and severity of CO-CDI described in the present study need to be recognized by ED physicians and managed appropriately. In addition, proactive measures for contact isolation need to be considered regarding patients admitted from the ED with diarrhea. Only 1 previous study was identified directly calculating the incidence of CO-CDI, rather than the subset of community-associated CDI [17]. This study was a 1999 United Kingdom randomized, prospective case-control study consisting of urban (Leeds, UK) and semirural (Truro and Cornwall, UK) cohorts, each of 1000 patients, which calculated incidences of 29.5 and 20.2 per 100000 population, respectively [17]. A previous study looking at hospital costs showed an attributable cost of CDI to be around $2454 to $3240, a finding similar to ours; however, we examined hospital charges, not costs [27]. Another previous study coincided with ours, which found a mean LOS of 6.4 days [28]. Previous studies conducted in the United States looking specifically at ED-CDI all had a similar methodology, taking stool samples from patients presenting to the ED with diarrhea [18,20,21]. Two performed before 2005 in the pediatric population found 6.7% and 5.5% of the stool samples tested positive for CDI based on cytotoxicity assay [20,21]. The second study found that CDI positivity was similar in both cases and controls [21]. A study conducted in South Korea from 2006 to 2010 found that 36 (13.2%) of 272 diarrheic patients tested positive for C difficile and that rate increased year to year most notably 16.7% in 2010 [19]. That study also found a male-to-female ratio of 17:15 and a mean age of 55 years. A similar study conducted in California from 2001 to 2002 found 18 culture-positive cases of CDI from 56 stool samples and found a male-to-female ratio of 11:7 and a mean age of 53.7 years [18]. Females have been found to have higher
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
241
Fig. 2. A, The prevalence (cases/100000 population) by sex of CDI among patients presenting to an ED in the United States from 2006 to 2010. B, The prevalence (cases/100000 population) by region of CDI among patients presenting to an ED in the United States from 2006 to 2010.
prevalence of both community-acquired and hospital-acquired CDI than males, coinciding with our findings of a female predominance [29]. The present study analyzed a much larger, more current national ED discharge database to determine the magnitude of and epidemiologic factors associated with CDI presenting to US EDs. It also estimated the national prevalence, mortality rate, and resource burden of CDI presenting to EDs, while tracking these characteristics over time to identify trends. No major change in diagnostic methods or treatment modalities of CDI occurred during the study period. The first commercially available real-time polymerase chain reaction assay was approved by the Food and Drug Administration in December 2008; however, C difficile enzyme Table 2 Mortality rate and LOS among patients with CDI presenting to an ED in the United States from 2006 to 2010 LOS after hospital admission (median days)
Death during hospital stay after ED visit (mortality rate/million population)
2006 2007 2008 2009 2010 2006 All ED visits Age (y) 0-17 18-44 45-64 65-84 ≥85
5
5
5
5
5
3 3 5 6 6
3 4 5 6 6
3 4 5 6 6
3 4 5 6 6
3 4 4 5 6
8.6
2007
2008
2009
2010
11.5
11.7
12.1
9.9
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 2.6 3.7 3.6 3.0 42.5 60.7 54.8 56.6 45.3 195.8 229.7 242.4 262.5 222.5
immunoassay of toxin in stool was likely the most widely used method of diagnosis due to its rapid response and lower cost during the study period [30]. According to an article in the College of American Pathologists online newsletter in November 2010, the author states that more than 90% of C difficile testing was still done by enzyme immunoassay [31]. Oral vancomycin and metronidazole have been the drugs of choice for CDI, whereas immunotherapy and fecal bacteriotherapy are rare treatments reserved for relapsing patients. Fidaxomicin had not yet been approved for use during the study period. Other medicines, such as probiotics, toxin-binding agents, and rifaximin, are complimentary agents rather than alternatives. Because the diagnostic measures and treatments primarily used across the United States were essentially unchanged during our study period, increases found in CDI presenting to EDs were likely due to a variable other than how CDI was diagnosed or treated, such as education, awareness, prevalence of antibiotic exposure, and availability of testing. One possible explanation of the increasing prevalence and severity of ED-CDI is a change in the pathogen itself. Studies have shown the emergence of a more virulent and resistant, toxigenic strain called BI/ NAP1/027, with reports of it even in the community [8,32-38]. This strain has been reported in the United States, Canada, Australia, and Asia as well as in European countries such as the UK, Belgium, France, and the Netherlands [8,32,34,35,37]. We found the largest increase in prevalence to occur in the northeastern region of the United States; reasons for this are unclear but may be associated with an epicenter of the BI/NAP/027 strain located in Montreal, Quebec, Canada [8].
242
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
Fig. 3. A, Mortality rates (deaths per million population) by sex of CDI among patients presenting to an ED in the United States from 2006 to 2010. B, Mortality rates (deaths per million population) by region of CDI among patients presenting to an ED in the United States from 2006 to 2010.
Additional hypervirulent strains, NAP7 and NAP8/078/V, have also been increasing in prevalence in the United States, Canada, and Europe over recent years [39-41]. These cases have been highly associated with CO-CDI. Other potential factors contributing to the increasing prevalence of ED-CDI presenting to EDs include increased testing due to
Table 3 Emergency department and hospital charges (median charge in dollars [$]) among patients with CDI in the United States from 2006 to 2010 2006
All ED visits Age (y) 0-17 18-44 45-64 65-84 ≥85 Sex Male Female Region Northeast Midwest South West
2007
2008
2009
2010
US $
US $
US $
US $
US $
18 426
20 012
22 396
22 932
23 386
10 194 15 338 17 192 20 030 18 588
12 216 16 595 19 137 21 193 20 848
13 343 18 136 21 339 24 088 22 809
13 591 18 811 21 456 24 568 24 097
13 770 19 370 22 539 24 959 24 780
18 282 18 505
19 886 20 098
22 710 22 166
22 987 22 898
23 535 23 270
21 617 14 146 17 528 25 088
21 577 15 868 19 436 30 122
24 139 17 106 21 478 33 791
23 547 18 505 22 565 36 823
24 189 19 778 22 375 32 755
increased physician awareness and asymptomatic carriage in the community [13,17,42].
4.1. Limitations The present study was limited by the use of ICD-9-CM coding as the case identifier, potentially allowing misclassification into our results. We estimate that this level of misclassification was minimal due to several studies that found ICD-9-CM code 008.45 to be an accurate way to identify CDI [24,43,44]. Individuals who presented to an ED and were discharged with a secondary diagnosis of CDI were not included in this study, as we only included those cases with CDI listed as the primary diagnosis. We calculated that primary diagnoses of CDI only accounted for 37% of all listed diagnoses of CDI. Solely using visits with CDI listed as the principal diagnosis may have caused an underestimation of the prevalence and mortality of our results (a potential error of omission; α error). Unique patient identifiers were not available, and the unit of analysis for HCUP data was the hospital encounter (ie, the ED visit or the ED visit in addition to the hospital stay), not each individual patient. Thus, multiple visits by the same individual may have been counted more than once. This would have caused an overestimation of the prevalence and mortality of our results (a potential error of commission; β error). In addition, deaths that occurred during a hospitalization with a primary diagnosis of CDI do not necessarily mean that CDI was the cause of the death, which our results assumed. However, using
A.M. Smith et al. / American Journal of Emergency Medicine 33 (2015) 238–243
only those cases with a primary diagnosis of CDI, we believe, decreased the severity of this limitation. Lastly, using the publically available database limited the variables in which we could analyze and limited our ability to calculate adjusted risk ratios.
[20]
[21]
5. Conclusion [22]
The prevalence of CDI presenting to US EDs increased in the United States from 2006 to 2010, albeit, at a slowing rate from year to year. The mortality rate also increased over the study period but did not reach significance. Emergency department visits occurred more frequently in older individuals, in females, and in the northeast United States. Acknowledgments The authors are grateful to L. Clifford McDonald, MD, FACP, FSHEA, for previewing the manuscript. No compensation was provided.
[23]
[24]
[25]
[26]
References [27] [1] Miller BA, Chen LF, Sexton DJ, Anderson DJ. Comparison of the burdens of hospitalonset, healthcare facility-associated Clostridium difficile infection and of healthcareassociated infection due to methicillin-resistant Staphylococcus aureus in community hospitals. Infect Control Hosp Epidemiol 2011;32:387–90. [2] Kelly CP, Pothoulakis C, LaMont JT. Clostridium difficile colitis. N Engl J Med 1994;330: 257–62. [3] Centers for Disease Control & Prevention. Severe Clostridium difficile–associated disease in populations previously at low risk—four states, 2005. MMWR Morb Mortal Wkly Rep 2005;54:1201–5. [4] Redelings MD, Sorvillo F, Mascola L. Increase in Clostridium difficile–related mortality rates, United States, 1999-2004. Emerg Infect Dis 2007;13:1417–9. [5] Chandler RE, Hedberg K, Cieslak PR. Clostridium difficile–associated disease in Oregon: increasing incidence and hospital-level risk factors. Infect Control Hosp Epidemiol 2007;28:116–22. [6] Elixhauser A, Jhung M. Clostridium difficile–associated disease in U.S. hospitals, 1993-2005: Statistical Brief #50. Rockville (MD): Healthcare Cost and Utilization Project (HCUP) Statistical Briefs; 2006. [7] McDonald LC, Owings M, Jernigan DB. Clostridium difficile infection in patients discharged from US short-stay hospitals, 1996-2003. Emerg Infect Dis 2006;12:409–15. [8] Pepin J, Valiquette L, Alary ME, Villemure P, Pelletier A, Forget K, et al. Clostridium difficile–associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004;171:466–72. [9] Dumyati G, Stevens V, Hannett GE, Thompson AD, Long C, MacCannell D, et al. Community-associated Clostridium difficile infections, Monroe County, New York, USA. Emerg Infect Dis 2012;18:392–400. [10] Kuntz JL, Johnson ES, Raebel MA, Petrik AF, Yang X, Thorp M, et al. Epidemiology and healthcare costs of incident Clostridium difficile infections identified in the outpatient healthcare setting. Infect Control Hosp Epidemiol 2012;33:1031–8. [11] Kutty PK, Woods CW, Sena AC, Benoit SR, Naggie S, Frederick J, et al. Risk factors for and estimated incidence of community-associated Clostridium difficile infection, North Carolina, USA. Emerg Infect Dis 2010;16:197–204. [12] Naggie S, Frederick J, Pien BC, Miller BA, Provenzale DT, Goldberg KC, et al. Community-associated Clostridium difficile infection: experience of a veteran affairs medical center in southeastern USA. Infection 2010;38:297–300. [13] Bauer MP, Goorhuis A, Koster T, Numan-Ruberg SC, Hagen EC, Debast SB, et al. Community-onset Clostridium difficile–associated diarrhoea not associated with antibiotic usage—two case reports with review of the changing epidemiology of Clostridium difficile–associated diarrhoea. Neth J Med 2008;66:207–11. [14] Hirshon JM, Thompson AD, Limbago B, McDonald LC, Bonkosky M, Heimer R, et al. Clostridium difficile infection in outpatients, Maryland and Connecticut, USA, 20022007. Emerg Infect Dis 2011;17:1946–9. [15] Fellmeth G, Yarlagadda S, Iyer S. Epidemiology of community-onset Clostridium difficile infection in a community in the South of England. J Infect Public Health 2010;3:118–23. [16] Stergachis A, Levy DG, McFarland LV, Van Vorst K, Graham DJ, Johnson ES, et al. Antibiotics and Clostridium difficile diarrhea in the ambulatory care setting. Clin Ther 2000;22:91–102. [17] Wilcox MH, Mooney L, Bendall R, Settle CD, Fawley WN. A case-control study of community-associated Clostridium difficile infection. J Antimicrob Chemother 2008; 62:388–96. [18] Martirosian G, Szczesny A, Silva J. Clostridium difficile in emergency room. Anaerobe 2005;11:258–61. [19] Shin BM, Moon SJ, Kim YS, Shin WC, Yoo HM. Characterization of cases of Clostridium difficile infection (CDI) presenting at an emergency room: molecular and clinical fea-
[28]
[29]
[30] [31] [32]
[33]
[34]
[35]
[36]
[37] [38]
[39]
[40]
[41]
[42]
[43]
[44]
243
tures differentiate community-onset hospital-associated and community-associated CDI in a tertiary care hospital. J Clin Microbiol 2011;49:2161–5. Klein EJ, Boster DR, Stapp JR, Wells JG, Qin X, Clausen CR, et al. Diarrhea etiology in a children's hospital emergency department: a prospective cohort study. Clin Infect Dis 2006;43:807–13. Denno DN, Shaikh N, Stapp JR, Qin X, Hutter CM, Hoffman V, et al. Diarrhea etiology in a pediatric emergency department: a case control study. Clin Infect Dis 2012;55: 897–904. HCUPnet, Healthcare Cost and Utilization Project (HCUP). Nationwide Emergency Department Sample (NEDS). Rockville, MD: Agency for Healthcare Research and Quality (AHRQ); 2006–2010. Population Estimates, Historical Data. United States Census Bureau Website. http:// www.census.gov/popest/data/national/totals/2013/index.html. [Accessed May 15, 2013. Updated December 30, 2013]. Jhung MA, Thompson AD, Killgore GE, Zukowski WE, Songer G, Warny M, et al. Toxinotype V Clostridium difficile in humans and food animals. Emerg Infect Dis 2008;14:1039–45. Healthcare Cost and Utilization Project (HCUP). Healthcare Cost and Utilization Project (HCUP), Nationwide Emergency Department Sample (NEDS) and Agency for Healthcare Research and Quality (AHRQ). Website. Rockville, MD http://www. hcup-us.ahrq.gov/nedsoverview.jsp. [Accessed September 16, 2014. Updated June 24, 2014]. Z-Test calculator. Healthcare Cost and Utilization Project (HCUP) hyperlink to Z-Test Calculator Website. http://hcupnet.ahrq.gov/ZTestCalc.jsp?Now=1389033854065&JS=Y. [Accessed May 15, 2013]. Dubberke ER, Reske KA, Olsen MA, McDonald LC, Fraser VJ. Short and long-term attributable costs of Clostridium difficile–associated disease in nonsurgical inpatients. Clin Infect Dis 2008;46:497–504. O'Brien JA, Lahue BJ, Caro JJ, Davidson DM. The emerging infectious challenge of Clostridium difficile–associated disease in Massachusetts hospitals: clinical and economic consequences. Infect Control Hosp Epidemiol 2007;28:1219–27. Khanna S, Pardi DS, Aronson SL, Kammer PP, Baddour LM. The epidemiology of community-acquired Clostridium difficile infection: a population-based study. Am J Gastroenterol 2012;107:89–95. Dubberke ER, Reske KA, McDonald LC, Fraser VJ. ICD-9 codes and surveillance for Clostridium difficile–associated disease. Emerg Infect Dis 2006;12:1576–9. Scheurer DB, Hicks LS, Cook EF, Schnipper JL. Accuracy of ICD-9 coding for Clostridium difficile infections: a retrospective cohort. Epidemiol Infect 2007;135:1010–3. Schmiedeskamp M, Harpe S, Polk R, Oinonen M, Pakyz A. Use of International Classification of Diseases, Ninth Revision, Clinical Modification codes and medication use data to identify nosocomial Clostridium difficile infection. Infect Control Hosp Epidemiol 2009;30:1070–6. Kvach EJ, Ferguson D, Riska PF, Landry ML. Comparison of BD GeneOhm Cdiff realtime PCR assay with a two-step algorithm and a toxin A/B enzyme-linked immunosorbent assay for diagnosis of toxigenic Clostridium difficile infection. J Clin Microbiol 2010;48:109–14. Merrigan M. C. diff molecular testing with less time and effort. College of American Pathologists Website. http://www.cap.org/apps/cap.portaL?_nfpb=true&cntvwrPtlt_ actionOverride=%2Fportlets%2FcontentViewer%2Fshow&_windowLabel= cntvwrPtlt&cntvwrPtlt%7BactionForm.contentReference%7D=cap_today% 2F1110%2F1110g_c_diff.html&_state=maximized&_pageLabel=cntvwr. [Accessed April 26, 2013. Updated November 2010]. Clements AC, Magalhaes RJ, Tatem AJ, Paterson DL, Riley TV. Clostridium difficile PCR ribotype 027: assessing the risks of further worldwide spread. Lancet Infect Dis 2010;10:395–404. Huang H, Weintraub A, Fang H, Nord CE. Community acquired Clostridium difficile infection due to a moxifloxacin susceptible ribotype 027 strain. Scand J Infect Dis 2009;41:158–9. Kuijper EJ, Coignard B, Tull P. Emergence of Clostridium difficile–associated disease in North America and Europe. Clin Microbiol Infect 2006;12(Suppl. 6):2–18. McDonald LC, Killgore GE, Thompson A, Owens Jr RC, Kazakova SV, Sambol SP, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005; 353:2433–41. Quesada-Gomez C, Vargas P, Lopez-Urena D, Gamboa-Coronado Mdel M, RodriguezCavallini E. Community-acquired Clostridium difficile NAP1/027-associated diarrhea in an eighteen month old child. Anaerobe 2012;18:581–3. Richards M, Knox J, Elliott B, Mackin K, Lyras D, Waring LJ, et al. Severe infection with Clostridium difficile PCR ribotype 027 acquired in Melbourne, Australia. Med J Aust 2011;194:369–71. Warny M, Pepin J, Fang A, Killgore G, Thompson A, Brazier J, et al. Toxin production by an emerging strain of Clostridium difficile associated with outbreaks of severe disease in North America and Europe. Lancet 2005;366:1079–84. Goorhuis A, Bakker D, Corver J, Debast SB, Harmanus C, Notermans DW, et al. Emergence of Clostridium difficile infection due to a new hypervirulent strain, polymerase chain reaction ribotype 078. Clin Infect Dis 2008;47:1162–70. Mulvey MR, Boyd DA, Gravel D, Hutchinson J, Kelly S, McGeer A, et al. Hypervirulent Clostridium difficile strains in hospitalized patients, Canada. Emerg Infect Dis 2010; 16:678–81. Khanna S, Pardi DS, Aronson SL, Kammer PP, Baddour LM. Outcomes in communityacquired Clostridium difficile infection. Aliment Pharmacol Ther 2012;35:613–8.
