Clinical and Microbiological Risk Factors for Community-Associated Clostridioides difficile Infections

Young Ah Kim1 , Heejung Kim2 , Dokyun Kim2 , Changseung Liu3 , Seok Hoon Jeong2

1Department of Laboratory Medicine, National Health Insurance Service, Ilsan Hospital, Goyang
2Department of Laboratory Medicine and Research Institute of Bacterial Resistance, Yonsei University College of Medicine, Seoul
3Department of Laboratory Medicine, Gangneung Asan Hospital, University of Ulsan College of Medicine, Gangneung, Korea

Corresponding author :


Background: The incidence of community-associated (CA) Clostridioides difficile infection (CDI) has increased in Korea. In this study, we evaluated CA-CDI risk factors in terms of clinical features and ribotype considering its region-specific molecular epidemiology.
Methods: A retrospective case-control study was performed on two groups of CDI patients: 127 subjects with CA-CDI and 265 subjects with healthcare-associated (HA)-CDI. Risk factors for CA-CDI were evaluated in terms of clinical and microbiological features such as toxin type and ribotype.
Results: A comparison of the two groups of CDI patients revealed that inflammatory bowel disease, diarrhea, abdominal pain, and fever were more closely associated with CA-CDI. The toxin types and ribotypes of C. difficile were similar between the two groups. After adjusting for variables, no risk factors were identified for CA-CDI compared with HA-CDI.
Conclusion: Specific risk factors for CA-CDI were not identified in this study.


Clostridioides difficile infection, Community-associated, Risk factor


Clostridioides difficile causes infectious diarrhea with disease severity ranging from mild to severe [1]. Although the incidence and mortality rate of C. difficile infection (CDI) have increased dramatically worldwide since 2003 with the emergence of binary toxin-producing ribotype 027 strains [2], this type is not prevalent in Korea [3]. Although CDI has been regarded as a healthcare-associated (HA) disease entity, the incidence of community-associated (CA) cases has increased since 2011 [4]. This shift was observed in recent epidemiologic data from Korea, showing that CA-CDI accounted for 19.4% of all cases of CDI [3]. In this study, CA-CDI risk factors were evaluated in aspects of not only clinical features, but also ribotypes, considering region-specific molecular epidemiology. A retrospective case-control study was performed to compare patient characteristics, prognosis, and risk factors for CA-CDI.


Study population and definition

All patients who visited Ilsan Hospital or Gangnam Severance Hospital in 2018 who were diagnosed with CDI based on C. difficile culture were included in this study. We only included the first infection during the study period to avoid duplication. This retrospective case-control study was done with two groups: CA-CDI (n = 127) and HA-CDI (n = 265). CA case was defined if the case occurred within 48 hours of hospital admission and the patient had not been admitted to a healthcare facility in previous 12 weeks. Others were regarded as HA cases in this study.

Clinical features were obtained by reviewing electronic medical records. Variables included age, sex, associated disease, history (within 12 weeks) of antimicrobials, history (within 12 weeks) of chemotherapy, history (within 12 weeks) of proton pump inhibitor, sites of acquisition, CDI treatments, history of CDI (within 12 weeks), recurrence after eight weeks, death, toxin type, and ribotype of C. difficile.

Molecular study

Toxin production and molecular epidemiology were determined with polymerase chain reaction (PCR)-sequencing as described in a previous study [3]. For toxin A and B genes, primer pairs used were tcdA-F and tcdA-R for tcdA, NK104 and NK105 for tcdB, cdtA-pos and cdtA-rev for cdtA, and cdtB-pos and cdtB-rev for cdtB. PCR ribotyping was performed as previously described with primers CD1-CD1445 [3]. A comparison of PCR ribotyping patterns was performed visually with known standards (VPI 10463, UK078, 48489ATCC9689, ATCC43598, and ATCC70057). Ribotype patterns that differed by at least one band were assigned to different types. Multilocus sequence typing (MLST) was performed using a scheme previously described by Griffiths et al. [5], using seven housekeeping genes (adk, atpA, dxr, glyA, recA, soda, and tpi). PCR reactions for these seven loci were performed and amplicons were sequenced with forward and reverse primers. DNA sequences were submitted to MLST database ( to obtain sequence type.

Statistical analysis

A continuous variable such as age was analyzed using the Mann-Whitney U test. Chi-squared test was used for comparative analysis of categorical variables to determine independent risk factors. Odds ratio and 95% confidence interval values were calculated for binomial variables. Variables with P values of less than 0.1 in univariate analyses were included in a multivariate logistic regression analysis model to determine independent risk factors. Statistical significance was defined at P < 0.05. SPSS 23.0 software (IBM Corp., Armonk, NY, USA) was used for univariate analyses and multivariate analyses. This study was approved by our institutional review board as required by the hospital policy (IRB No. NHIMC-2020-05-015).


Comparison of CA-CDI and HA-CDI

When two groups of CDI were compared, in.ammatory bowel disease (6.3% in CA-CDI vs. 0.4% in HA-CDI, P = 0.0070), diarrhea (66.1% in CA-CDI vs. 46.0% in HA-CDI, P = 0.0002), abdominal pain (22.8% in CA-CDI vs. 10.9% in HA-CDI, P = 0.0023), and fever (20.5% in CA-CDI vs. 12.5% in HA-CDI, P = 0.0394) occurred more in the CA-CDI group (Table 1).


Table 1. Comparison between CA-CDI and HA-CDI groups


CA-CDI (n = 127)

HA-CDI (n = 265)


Age (yr)




Sex, male

51 (40.2)

120 (45.3)


Charlson comorbidity index




Associated disease

Biliary tract disease

3 (2.4)

8 (3.0)



22 (17.3)

59 (22.3)



18 (14.2)

57 (21.5)


Heart failure

5 (3.9)

8 (3.0)


Chronic respiratory disease

5 (3.9)

22 (8.3)


Chronic renal disease

24 (18.9)

42 (15.9)


Diabetes mellitus

23 (18.1)

38 (14.3)


Cerebrovascular disease

7 (5.5)

34 (12.8)


Alcohol disorder

5 (3.9)

4 (1.5)




4 (1.5)



4 (3.2)

9 (3.4)


Esophageal disorder

1 (0.8)

8 (3.0)


Nutrition deficiency

1 (0.8)

4 (1.5)


Inflammatory bowel disease

8 (6.3)

1 (0.4)


Gastric ulcer

2 (1.6)

10 (3.8)


History of antimicrobial use


104 (81.9)

244 (92.1)



28 (22.1)

73 (27.6)


Narrow-spectrum cephalosporin

12 (9.5)

37 (14.0)


Extended-spectrum cephalosporin

31 (24.4)

69 (26.0)



12 (9.5)

60 (22.6)



9 (7.1)

46 (17.4)



16 (12.6)

58 (21.9)



4 (3.2)

39 (14.7)




5 (1.9)


History of PPI use

13 (10.2)

34 (12.8)


History of chemotherapy

13 (10.2)

28 (10.6)


CDI-associated symptom


84 (66.1)

122 (46.0)


Abdominal pain

29 (22.8)

29 (10.9)


Fever (> 38°C)

26 (20.5)

33 (12.5)




112 (88.9)

224 (84.5)



4 (3.2)

15 (5.7)


ICU admission

3 (2.4)

19 (7.2)


Crude mortality

14 (11.0)

44 (16.6)


C. difficile toxin


115 (90.6)

232 (87.6)


B only

8 (6.3)

22 (8.3)



4 (3.2)

11 (4.2)


Ribotypes of C. difficile

AB24 (ST129)

3 (2.4)

8 (3.0)


AB25 (ST102)

5 (3.9)

8 (3.0)


Ribotype 001

6 (4.7)

16 (6.0)


Ribotype 002

12 (9.5)

17 (6.4)


Ribotype 012

8 (6.3)

14 (5.3)


Ribotype 014/020

17 (13.4)

43 (16.2)


Ribotype 017

5 (3.4)

18 (6.8)


Ribotype 018

29 (22.8)

58 (21.9)


Ribotype 046

4 (3.2)

18 (6.8)


Ribotype 070

3 (2.4)

4 (1.5)


Ribotype 106

8 (6.3)

14 (5.3)






Data are presented in number (%) or mean±standard deviation; Bold format indicates statistical significance.
*Others included AB11, AB15, AB21, AB23, AB27, AB30, AB32, AB33, AB37, AB38, AB39, AB43, AB45, AB47, AB59, AB62, AB72, AB84, AB85, AB86, AB89, C29, C3, C31, R020, R023, R027, R078, R081, R087, R088, R103, R137, R159, R161, R163, and R369.
Abbreviations: CDI, C. difficile infection; CA, community-associated; HA, healthcare-associated; PPI, proton pump inhibitor; ICU, intensive care unit; cdt, binary toxin; ST, sequence type.


However, older age (66.9±18.9 years in CA-CDI vs. 72.1±13.5 years in HA-CDI, P = 0.0064), cerebrovascular disease (5.5% in CA-CDI vs. 12.8% in HA-CDI, P = 0.0314), past history of any antimicrobial use (81.9% in CA-CDI vs. 92.1% in HA-CDI, P = 0.0036), inhibitor combination use (9.5% in CA-CDI vs. 22.6% in HA-CDI, P = 0.0022), carbapenem use (7.1% in CA-CDI vs. 17.4% in HA-CDI, P = 0.0080), .uoroquinolone use (12.6% in CA-CDI vs. 21.9% in HA-CDI, P = 0.0298), and teicoplanin use (3.2% in CA-CDI vs. 14.7% in HA-CDI, P = 0.0019) were more frequent in the HA-CDI group (Table 1). Toxin types and ribotypes of C. difficile were similar to each other between the two groups.

The risk factors of CA-CDI over HA-CDI

After variables such as age, underlying diseases (pneumonia, cerebrovascular disease, in.ammatory bowel disease), past antimicrobial use (inhibitor combination, carbapenem, fluoroquinolone, teicoplanin), CDI-related symptoms (diarrhea, abdominal pain, fever), and intensive care unit admission were adjusted for, no risk factor for CA-CDI over HA-CDI was found.


Transmission of C. difficile could be plausibly sustained by asymptomatically colonized persons in the community or exposure to animal reservoirs [6]. Under-reporting and systematic misclassification might also underplay the role of community transmission because the potentially long incubation period can make patients display symptoms for the first time in a healthcare facility [7]. According to the present study, the infection should be classified as being acquired prior to admission if symptoms begin within five days of admission. However, we used the commonly recommended two-day cut-off [8].

Although specific risk factors associated with CA-CDI were not found in multivariate analysis, in.ammatory bowel disease and CDI-related symptoms (diarrhea, abdominal pain, and fever) were more commonly found in the CA-CDI group. One study has shown that the CDI-CA group tends to be younger and healthier than the HA-CDI group [7]. It has been suggested that those with CDI-CA might be at a higher risk for recurrence than those with HA-CDI [7]. In this study, we could not find a difference in recurrence rate or recovery between the two groups. However, age was younger in the CA-CDI group, consistent with the previous study [7].

The increase of CDI occurring among persons without recent hospitalizations or stays in a long-term care facility could be another challenge to national efforts for reducing CDI with infection prevention and antibiotic stewardship [9]. Great use of outpatient antimicrobials is a well-known contributing factor of CA-CDI [10], but the past antimicrobial use was not a significant risk factor for CA-CDI over HA-CDI after adjustment in this study. The limitation of study is that antimicrobial use was evaluated only according to the electronic medical record findings and deep interview need to be included not to miss the antimicrobial use in other clinics. Although antimicrobial prescription has decreased after the Korean government has implemented a series of healthcare policies, most (72%) of total orders are administered in clinics [11]. Although changing prescribing behaviors can be challenging, we need to force guidelines to optimize antimicrobial therapy in outpatient settings.


배경: 지역사회 관련(community-associated, CA) Clostridioides difficile 감염(CDI)이 국내에서 증가하고 있다. 본 연구에서는 CA-CDI의 위험 요인을 임상적 특징과 ribotype을 포함한 지역 특이 분자역학을 고려하여 평가하고자 한다.
방법: 후향적 환자-대조군 연구로 환자군 CA-CDI 127명과 대조군 의료 관련 CDI 265명을 비교하였다. CA-CDI 관련 위험 인자는 임상적 특징과 독소형 및 리보형과 같은 미생물학적인 특성을 포함하여 분석하였다.
결과: 염증성 장질환, 설사, 복통, 발열은 CA-CDI와 더 관련이 있었다. 독소형과 ribotype은 두 그룹 사이에서 서로 유사했다. 변수를 조정한 후에는 CA-CDI와 연관된 의미있는 위험 인자가 없었다.
결론: CA-CDI와 관련된 특정 위험 인자는 본 연구에서 확인되지 않았다.


No potential con.icts of interest relevant to this article were reported.


We would like to thank So Ra Yoon (Ph.D., Research Institute of National Health Insurance Ilsan Hospital) for providing help on statistical analyses.


This research was supported by a research grant (2017E4400202) from Korea Centers for Disease Control and Prevention.


1. Bartlett JG. Clostridium difficile: history of its role as an enteric pathogen and the current state of knowledge about the organism. Clin Infect Dis 1994;18:S265-72.  

2. Wamy M, Pe’pin 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;366:1079­84. 

3. Byun JH, Kim H, Kim JL, Kim D, Jeong SH, Shin JH, et al.€A nationwide study of molecular epidemiology and antimicrobial susceptibility of Clostridioides difficile in South Korea. Anaerobe 2019;60:102106.  

4. Evans ME, Simbartl LA, Kralovic SM, Jain R, Roselle GA. Clostridium difficile infections in Veterans Health Administration acute care facilities. Infect Control Hosp Epidemiol 2014; 5:1037­42.  

5. Griffiths D, Fawley W, Kachrimanidou M, Bowden R, Crook DW, Fung R, et al.€Multilocus sequence typing of Clostridium difficile. J Clin Microbiol 2010;48:770-8.   

6. McLure A, Clements ACA, Kirk M, Glass K. Modelling diverse sources of Clostridium difficile in the community: importance of animals, infants and asymptomatic carriers. Epidemiol Infect 2019;147:e152, 1­9.   

7. McLure A, Clements ACA, Kirk M, Glass K. Clostridium difficile classification overestimates hospital-acquired infections. J Hosp Infect 2018;99:453-60.  

8. McDonald LC, Coignard B, Dubberke E, Song X, Horan T, Kutty PK, et al.€Recommendations for surveillance of Clostridium difficile-associated disease. Infect Control Hosp Epidemiol 2007;28:140-5.  

9. McDonald LC, Gerding DN, Johnson S, Bakken JS, Carroll KC, Coffin SE, et al.€Clinical practice guidelines for Clostridium difficile infection in adults and children: 2017 update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis 2018;66:e1-48.   

10. Guh AY, Adkins SH, Li Q, Bulens SN, Farley MM, Smith Z, et al.€Risk factors for community-associated Clostridium difficile infection in adults: a case-control study. Open Forum Infect Dis 2017;4:ofx171.   

11. Kim YA, Park YS, Youk T, Lee H, Lee K. Changes in antimicrobial usage patterns in Korea: 12-year analysis based on database of the National Health Insurance Service-National Sample Cohort. Sci Rep 2018;8:12210.