Plain Language Summary
Introduction
Methods
Identification
Selection
Quality Determination and Data Extraction
Compliance with Ethics Guidelines
Results
Identification of Relevant Publications
Incidence and Prevalence
Reference | Study period | Study design | Patient population | C. diff diagnosis (n) | C. diff definition | Incidence | Prevalence | Risk factors for CDI |
---|---|---|---|---|---|---|---|---|
Akahoshi et al. [24] | November 2007–May 2014 | Retrospective, single-center, chart review n = 308 | HSCT (n = 102 autologous; n = 206 allogeneic) | n = 30 Occurring median 7 days (range 0–36) after HSCT conditioning | Diarrhea (≥ 3 loose stools/24 h) in first 100 days post HSCT Positive CD toxina or positive CD toxin plus GDHb | Cumulative incidence 6.2% in HSCT population (9.2% allogeneic; 1.0% autologous) | – | Allogeneic HSCT, total body irradiation, stem cell source, acute leukemia, duration of neutropenia – linked with increased risk for CDI Allogeneic HSCT: OR for CDI 18.6 (95% CI 2.48–139) p < 0.01; duration of neutropenia ≥ 17 days in first 30 days: OR for CDI 10.4 (95% CI 2.37–46 p < 0.01) |
Daida et al. [31] | July 2003–September 2012 | Retrospective case–control study via medical record review from a single center | Pediatric patients (aged 0–19 years) admitted to hospital with cancer (n = 51) Matching case controls were selected from patients without CDI admitted to hospital within 2 months (before/after) admission for patients with CDI (n = 94) | n = 51 | Test for C. diff toxins A/Ba Appearance of symptoms ≥ 3 days after admission | 51/189 = 26.98% | – | Multivariable analysis of risk factors for hospital-acquired CDI: younger age is a risk factor: age 0–3 years vs. age 4–6 years, OR 0.13 (95% CI 0.03–0.59); p = 0.008, and vs. age ≥ 7 years, OR 0.12 (95% CI 0.03–0.45); p = 0.002. Prolonged neutropenia is a risk factor: OR 1.11 (95% CI 1.03–1.20); p = 0.008. Use of ≥ 4 antibiotics in the 60 days from diagnosis or reference date: OR 3.55 (95% CI 1.40–9.04); p = 0.008 |
Furuichi et al. [32] | August 2012–March 2013 | Prospective, non-interventional cohort to assess rates of community-acquired C. diff colonization (single center) n = 346 children | Healthy neonates (n = 95) and pediatric patients at hospital admission (n = 251) | 0 (0%) C. diff (asymptomatic) colonization in neonates Pediatric population without underlying disease: C. diff colonization 21.6% and 9% toxin positive colonization; vs. with underlying disease 30.8% and 23.1% (colonization and toxin-positive colonization) | Cultured fecal samples positive for C. diff toxin A/Bb | Asymptomatic CDI 9% toxin-positive colonization in pediatric patients with no underlying disease 23.1% toxin-positive colonization in pediatrics with underlying disease | – | Risk factors for toxin-positive C. diff colonization: underlying disease (OR 4.17, 95% CI 1.15–15.04; p = 0.049); age 12–23 months (OR 4.19, 95% CI 1.52–11.52; p = 0.01); tube feeding (OR 24.28, 95% CI 4.70–125.34; p < 0.001); toxin-positive C. diff (OR 8.29, 95% CI 1.87–36.84; p = 0.005) |
Hashimoto et al. [26] | January 1996–November 2004 | Retrospective chart review (single center) n = 242 | Living-donor liver transplant recipients (adult) | Diarrhea 76/242; C. diff diarrhea 11/242 | C. diff diarrhea: ≥ 3 loose stools on ≥ 2 consecutive days and positive stool culture and assays for toxin Ac and GDHd | 11/242 C. diff diarrhea = 4.5%e | – | Male gender (OR 4.56; 95% CI 1.02–33.3, p = 0.05) and serum creatinine md/dL ≥ 1.5 (OR 16.0, 95% CI 3.85–68.3, p = 0.0003) predicted risk for C. diff diarrhea Intensity of antibiotic use did not predict for C. diff diarrhea |
Hata et al. [62] | November 2007–December 2012 | Phase 3, multicenter, open-label RCT (assessing antibiotic prophylaxis) n = 579 | Colorectal surgery (colorectal cancer patients; elective laparoscopic) | Rate of C. diff infection: oral–IV prophylaxis group 1/289; IV prophylaxis 3/290 | Positive test for C. diff toxins in patients developing enteritis/colitis/diarrhea (assay not described) | Incidence rate C. diff toxins in oral–IV and IV groups 0.3% and 1.0%, respectively (p = NS between oral–IV and IV prophylaxis groups) | 5.2 cases/1000 patientse | – |
Hikone et al. [20] | August 2011–September 2013 | Retrospective chart review of in- and outpatient samples (single center) n = 2193 samples tested for C. diff toxin | In- and outpatient samples tested for C. diff | 107 specimens positive for C. diff toxin; 76 cases of healthcare-facility onset CDI | Positive C. diff toxin testb | Incidence rate 0.8 cases/10,000 patient-days 30-day and 90-day mortality rates: 7.9% and 14.5%, respectively | – | Risk factors for recurrent CDI: malignant disease (OR 7.98; 95% CI 1.22–52.2; p = 0.03); history of ICU hospitalization (OR 49.9; 95% CI 1.01–2470; p = 0.049) |
Honda et al. [18] | September 2010–August 2012 | Retrospective chart review (single tertiary care center) n = 22,863 adult patients; and 1537 C. diff tests in 851 patients | Cases of CDI in a non-outbreak setting | 126 cases diagnosed with CDI (86.5% were healthcare-facility onset CDI) | Diarrhea and positive toxin assayb or presence of pseudomembranous colitisf Healthcare-facility onset CDI: symptom onset > 3 days from admission Community-onset CDI: symptom onset prior to or within 3 days of admission | Healthcare-facility onset CDI: 3.11 cases/10,000 patient-days Community-onset CDI: 0.2 cases/10,000 patient-days for CDI attributable to the study hospital 30-day all-cause mortality in CDI cohort: 15.1% | 126/22,863 = 5.5 cases/1000 patientse | – |
Hosokawa et al. [25] | January 2007–December 2008 | Retrospective cohort (single center) n = 201 patients | Allogeneic HSCT patients (135 unrelated cord blood; 39 unrelated bone marrow and 27 related peripheral blood stem cell) | 167/201 patients tested for C. diff 17/201 diagnosed C. diff diarrhea C. diff diarrhea in: 11/135 (9%) unrelated cord blood recipients; 2/39 (6%) unrelated BMT recipients; 4/27 (16%) related PBST recipients | C. diff diarrhea: > 3 loose stools/24 h for 2 consecutive days and positive ELISA for C. diff toxin A | Cumulative incidence of C. diff diarrhea 9% at post-transplant day 100 | – | Total body irradiation associated with reduced risk of C. diff diarrhea C. diff diarrhea was not a cause of any death; no recurrence of C. diff diarrhea after treatment |
Iwamoto et al. [27] | Two periods: March 2004–February 2006 and April 2008–December 2008 | Prospective observational cohort (single center) n = 1226 | Rheumatology inpatients | 54 cases of healthcare associated infection of which 2 were C. diff infection (1 patient in each study period) | Healthcare-associated infection: developing > 3 days after admission | 2/1226 in rheumatology patients (0.16%)e | – | – |
Iwashima et al. [44] | April 2005–March 2008 | Retrospective cohort study assessing genotypic features of isolates and clinical characteristics of CDI (single center) n = 610 submitted specimens | Patients with stools found positive for C. diff culture (n = 106; of which 35 excluded as asymptomatic carriers and n = 14 excluded for non-toxigenic strains) | 71 C. diff infection cases assessed | PCR assessment of toxin A and B; ribotyping CDI: diarrhea or colitis with positive test for C. diff toxin B and no other enteropathogenic microorganisms Recurrent CDI: recurring within 2 months of previous episode | Incidence of CDIs with binary toxin-positive strains 5.6% (noted in patients with non-severe CDI) | Prevalence < 5 CDI cases/month | – |
Kaneko et al. [30] | January 2006–April 2009 | Retrospective cohort investigating for CDI during active phase of inflammatory bowel disease (single center) n = 137 | Active ulcerative colitis | 55/137 (40.1%) tested samples were CDI positive | Presence of toxin A antigeng in gut lavage | 40.1% in a sample tested for possible CDI | – | – |
Kobayashi et al. [21] | April 2012–September 2013 | Retrospective cohort study based on chart review at four teaching hospitals in Japan | Patients aged ≥ 14 years with hospital-onset CDI | n = 160 with hospital-onset CDI | According to SHEA/IDSA 2010 guidelines, based on positive CD toxin EIA.b Hospital-onset CDI: hospitalized for condition other than CDI for ≥ 2 days | 1.04 cases per 10,000 patient-days; 1.61 cases per 1000 admissions | – | – |
Komatsu et al. [22] | June 2008–December 2013 | Single-center RCT n = 379 Looking at efficacy of perioperative synbiotics to prevent infectious complications (particularly surgical site infection) | Colorectal surgery (laparoscopic) | 0/168 cases C. diff infection in synbiotics group and 1/194 in control group Author reports use of synbiotics suppressed increases in potentially pathogenic C. diff (detected in 4% before surgery both groups; detected in 4% 7 days after surgery in synbiotic group vs. 13% control; p = 0.05 vs. day before surgery | Gut microbiota assessed by YIF-SCAN and PCR analyses | 1/379 colorectal surgery patients (0.3%)e | – | – |
Mizui et al. [37] | February 2010– February 2011 | Retrospective study of risk factors for C. diff diarrhea (single center) n = 2716 patients given an injectable antibiotic Study also assessed impact of probiotics | Inpatients given antibiotics | 29 had C. diff diarrhea (2687 had non-C. diff diarrhea) Risk factors investigated between groups re: use of antibiotics ≥ 8 days; enteral nutrition; IV hyperalimentation; fasting, proton pump inhibitors H2 blockers; serum albumin ≤ 2.9 g/dL | C. diff diarrhea; tests not defined | – | – | Risk factors for C. diff diarrhea were: antibiotic use ≥ 8 days (OR 4.071; 95% CICI 1.333–12.430; p = 0.014), IV hyperalimentation (OR 3.414; 95% CI 1.469–7.934; p = 0.004), PPIs (OR 3.224; 95% CI 1.421–7.315; p = 0.005), H2 blockers (OR 2.376; 95% CI 1.047–5.391; p = 0.039) |
Mori and Aoki [19] | January 2010– December 2014 | Retrospective case–control, epidemiological, single-center study assessing risk factors for CDI | Outpatients (1,914,011 patient-years examined) CDI cases Age- and sex-matched controls (C. diff toxin- and culture-negative) | 26 patients had community-acquired CDI | Community-acquired CDI: outpatient presentation with diarrhea, stool culture positive C. diff toxin assaya,h | Incidence for community-acquired CDI 1.4/100,000 patient-years | – | 84.6% of patients with community acquired CDI had prior exposure to antibiotics Patients with community-acquired CDI more likely to have had prior antibiotics (OR 8.12; 95% CI 2.43—26.98) |
Ogami et al. [38] | 4-year period (dates not given) | Single-center, retrospective hospital cohort n = 463 | Inpatients with antimicrobial associated diarrhea | 95/463 cases (20.5%) were CDI | CDI manifesting as antimicrobial-associated diarrhea (≥ 3 stools/day > 48 h after-ward admission) and stool toxin positive (A and/or B)a | – | – | Increased ward use of antimicrobials clindamycin (OR 1.739; 95% CI 1.050–2.881; p = 0.032) and piperacillin (OR 1.598; 95% CI 1.006–2.539; p = 0.047) increased risk of CDI |
Oshima et al. [40] | Published studies 1990–2016 | Systematic review and meta-analysis of 67 published studies | Adults and pediatric (≤ 18 years) patients receiving PPI who developed acute-onset diarrhea. Also, control group | n = 17,217 in the test group; n = 286,018 in the control group Recurrent CDI occurred in n = 1279; n = 5459 in the control group | Laboratory confirmation of C. diff or clinical definition. No further detail provided | – | – | PPI use increased risk for initial CDI episode (random effects model, overall OR 2.34, 95% CI 1.94–2.82; p < 0.00001) Age-stratified subgroup analyses: significant associations between PPI use and initial CDI episode in adults (OR 2.30, 95% CI 1.89–2.80; p < 0.00001) and pediatrics (OR 3.00, 95% CI 1.44–6.23; p < 0.00001) |
Roughead et al. [36] | 2008–2013 (insurance database) 1996–2014 (hospital dataset) | Retrospective data from worker insurance database and a hospital in-/outpatient dataset from a single center 1.2 million patient records examined and sequence symmetry analysis used to assess PPI use as risk factor for CDI | n = 310 patients received PPIs and oral vancomycin (proxy indicator for CDI) | – | – | – | – | Positive association between PPI use and CDI (adjusted sequence ratio for insurance dataset 5.40; 95% CI 2.73–8.75 and for hospital dataset 3.21; 95% CI 2.12–4.55) |
Sadahiro et al. [63] | May 2008–October 2011 | Prospective, single-center RCT comparing oral antibiotics and probiotics pre surgery to prevent infection n = 310 | Colon cancer | No change in detection of C. diff toxin across three treatment groups (probiotics; antibiotics; control (no probiotics or antibiotics)) | Assessment of C. diff toxin (A and B) in stool samples by RIDASCREEN | Rates of CDI increased post-operatively in all groups (probiotic group, from 2.0% to 7.0%; antibiotic group, 5.1% to 9.1%; control group, 2.1% to 10.5%) | – | – |
Sasabuchi et al. [28] | July 2010–March 2013 | Retrospective cohort study using the Japanese Diagnosis Procedure Combination database (multicenter) n = 15,651 receiving prophylaxis n = 15,651 controls | Severe sepsis and receiving stress ulcer prophylaxis within 2 days of hospital admission; propensity-matched controls did not receive prophylaxis | In propensity-matched cohort, 215 and 204 cases of CDI in the stress ulcer prophylaxis and control groups, respectively | Not specified, but ICD-10 codes used for other definitions. CDI coded as ‘complication’ in medical records during hospitalization | 1.4% in stress ulcer prophylaxis group and 1.3% in control (p = 0.588) | – | – |
Suzuki et al. [23] | April 2010–March 2012 | Single-center prospective cohort pre and post intensive infection control measures n = 80 | Hospitalized patients | – | Based on medical records and healthcare resource use New-onset nosocomial C. diff-associated disease | C. diff-associated disease reduced from 0.47 cases/1000 inpatient days to 0.11 (p < 0.001) after intensive infection control team interventions | – | – |
Takahashi et al. [39] | November 2010– October 2011 | Multicenter case–control and cohort study n = 1026 CDI n = 878 controls | National Hospital Organization cohort Assessed for newly diagnosed CDI and matched controls (no CDI) | 93.9% of CDI cases developed within 48 h of hospital admission | GI symptoms, clinical suspicion of CDI and positive C. diff toxinsa,h,i from stool or C. diff isolation from stool cultures, or both | – | – | Risk factors for CDI development: disruption of feeding/parenteral and enteral feeding; first- and second-generation cephem antibiotics (OR 1.44; 95% CI 1.10–1.87), third- and fourth-generation cephem antibiotics (OR 1.86; 95% CI 1.48–2.33), carbapenem antibiotics (OR 1.87; 95% CI 1.44–2.42) Comorbidities more common in patients with CDI Analysis of 924 cases noted 11% mortality within 30 days of CDI onset Use of vancomycin reduce mortality (OR 0.43; 95% CI 0.25–0.75) PPIs and penicillin did not increase risk for CDI |
Watanabe et al. [64] | January–June 2005 | Multicenter, retrospective cohort n = 294 fecal samples submitted for C. diff testing | Hospitalized patients | 79/294 (5.5 cases/1000 beds monthly) were C. diff toxin A+ | C. diff toxin testc | 5.5 cases/1000 beds monthly, assessed for C. diff were found to be C. diff toxin A+ | – | – |
Yasunaga et al. [29] | 2007–2010 | Retrospective database review: analysis of factors affecting C. diff- associated disease and outcomes of C. diff diarrhea after GI surgery Japanese Diagnosis Procedure Combination inpatient database (multicenter) n = 143,652 | Inpatients/GI surgical patients | 409 cases of C. diff diarrhea (0.28%) Higher rates in colorectal surgery (0.37%) vs. gastrectomy (0.21%) and esophagectomy (0.25%) (p < 0.001) | CD enterocolitic ICD-10 code | Rate 0.28% Risk factors included: older age; higher Charlson comorbidity index; longer pre-operative LOS; non-academic center care In-hospital mortality higher in C. diff diarrhea than in non-C. diff diarrhea (3.4% vs. 1.6%: OR 1.83; 95% CI 1.07–3.13, p = 0.027) | 409/143,652 = 2.8/1000 patientse | Risk factors included: older age; higher Charlson comorbidity index; longer pre-operative LOS; non-academic center care In-hospital mortality higher in C. diff diarrhea (3.4% vs. 1.6% in non-C. diff diarrhea: OR 1.83; 95% CI 1.07–3.13; p = 0.027) LOS attributable to post-operative C. diff diarrhea 12.4 days (95% CI 9.7–15.0; p < 0.001) |
Risk Factors
Specific Strains Responsible for CDI
Reference | Study period | Study design | Patient population | C. diff definition | Test methods | Isolates and strains | Binary toxin |
---|---|---|---|---|---|---|---|
Collins et al. [3] | Pre-2013 | Narrative review and meta-analysis of 42 cohort studies on C. diff in Asian countries | Various | Various | In Japan: tcdA and tcdB characterization, PFGE, PCR ribotyping and slpA typing | Predominance of ribotype smz (018) in past decade Other common ribotypes: 014, 002, 001 | Not specifically mentioned in review of Japanese papers |
Iwashima et al. [44] | April 2005–March 2008 | Retrospective cohort study assessing genotypic features of isolates and clinical characteristics of CDI (single center) n = 610 submitted specimens | Patients with stools found positive for C. diff culture (n = 106; of which 35 excluded as asymptomatic carriers and n = 14 excluded for non-toxigenic strains) 71 CDI cases assessed | CDI defined as: diarrhea or colitis with toxin B positive C. diff and no other enteropathogenic microorganisms | PCR assessment of toxins A and B and ribotyping | Isolates A+B+CDT+: 4/71 A+B+CDT−: 58/71 A−B+CDT−: 9/71 Ribotype A+B+CDT+: 2 were j52; 1 was nc07109; 1 was km0403 A+B+CDT−: 19 were smz; 14 were yok; 13 were hr; 12 other A−B+CDT−: 6 were trf; 2 were fr; 1 was sgf No predominant ribotype spreading; the dominant types were smz, yok and hr (hr = equivalent to ribotype 014) No ribotypes 027 and 078 found in the study Duration of CDI longer in yok group (p < 0.05) | Incidence of CDIs with binary toxin-positive strains 5.6% (noted in non-severe CDI) |
Kato et al. [43] | February 2004–April 2004 | Single-center study to validate efficacy of slpA sequence typing 28 samples positive for toxin A from 17 patients with C. diff diarrhea | C. diff diarrhea: 22/28 samples positive by stool culture | Not specifically defined (see testing methods) | Detection of toxin A using UNIQUICK slpA sequence typing | All samples except hj2-2 isolate were A+B+CDT− (positive for toxin A and toxin B but negative for binary toxin) smz-1 (n = 10) and smz-2 (n = 6) accounted for 73% strains Strain pattern suggested nosocomial infection Yok-1, yok-2, t25–1, hr-1 and hj2-2 identified in at least 1 patient each | Binary toxin assessed |
Kato et al. [49] | 2003–2007 | Multicenter study typing C. diff isolates by slpA sequencing | 160 stool samples from symptomatic patients (hospitalized with a diagnosis of antibiotic-associated diarrhea or colitis) | Not specifically defined (see testing methods) 90 stool samples were typed of which 77 were positive by culture for C. diff | Stool culture: PCR for toxins A and B, and CDT PCR ribotyping and slpA sequence typing | Smz sequence type was dominant and detected by culture and/or typing in 61/99 stool samples positive for toxic culture and/or direct slpA sequencing (smz in 62%; smz-01, smz-02, smz-04) One isolate type gc8 corresponded with PCR ribotype 027 BI/NAP1/027); no PCR ribotype 078 found Direct typing from DNA extracted from stool samples: 77/90 were positive for C. diff and typing results agreed with isolated strain typing slpA subtypes smz-01, -02 and -04 found in 51/86 (59%) of stool samples that were tcdB-positive C.diff cultured and in 67% of stool where direct typing could be obtained | Of 87 isolates, 75 (86%) were A+B+ and 12 (14%) were A−B+; 3 A+B+ isolates were positive for PCR detecting the binary toxin gene (A+B+CDT+) |
Kawada et al. [47] | October 2009–January 2010 | Single-center study evaluating a single kit for rapid detection of GDH and toxin A/B in feces (as diagnosis of C. diff infection) | 60 specimens from 60 patients with antibiotic-associated diarrhea | C. diff culture was reference method C. diff evaluated 28 inpatients diagnosed as having CDI | Evaluation of C DIFF QUIK CHEK COMPLETE® vs. GDH detection by ImmunoCard and toxin A/B detection by TOX A/B | The kit had GDH sensitivity 100%; specificity 93.3%; negative predictive value 100% Kit had Toxin A/B sensitivity 78.6%, specificity 96.9% compared with toxigenic culture (culture B positive) The 22/23 specimens that were dual positive for GDH and toxin A/B were culture positive Dual negatives by the kit were C. diff culture negative | Not reported |
Kikkawa et al. [65] | January–June 2005 | Multicenter study looking at prevalence of A−/B+ strains in fecal samples submitted for C. diff tests | C. diff isolated in 159/332 specimens | As per test methods | Culture PCR analysis of toxigenic typing Genotyping by PCR, ribotyping and PFGE | 332 sample; C. diff isolated from 159: 137 strains (41% examined specimens and 86% of isolated C. diff) were A+B+; 10 (3% and 6%) were A−B+ and 12 (4% and 8%) were A−B− Therefore 10 (6.3%) of 159 C. diff strains were A−B+ All 10 A−/B+ strains had identical pattern by PCR ribotyping | Not reported |
Kobayashi et al. [45] | April 2008–March 2009 | Single-center retrospective study to test/validate the 3-day rule for ordering a C. diff toxin test in Japanese patients | 1597 stool cultures from 992 patients; 880 CD toxin tests performed in 529 patients 83 species from 81 specimens considered enteric pathogens | As per test methods | CD toxin by TOX A/B QUIK CHEK® | Rate of positive stool culture in different patient groups: 14.2% outpatients; 3.6% inpatient ≤ 3 days; 1.3% inpatients ≥ 4 days Respective CD toxin positive test rates: 1.9% outpatients; 7.1% inpatient ≤ 3 days; 8.5% inpatients ≥ 4 days The study validates the 3-day rule: the rule can be used to estimate the pre-test probability of a stool microbiological test | Not reported |
Kunishima et al. [66] | February 2003–February 2006 | Single-center study of antimicrobial susceptibility of C. diff isolates | Studied 157 C. diff isolates from patients with diarrhea and probable CDI | – | Antimicrobial sensitivity of isolates: broth microdilution method to determine MICs of 15 drugs | Found no strains resistant to either metronidazole or vancomycin | Not reported |
Kuwata et al. [46] | April 2012–March 2013 | Single-center study of molecular epidemiology and antimicrobial sensitivity of C. diff isolates | C. diff isolates (n = 130) | – | Toxin genotypes; MLST and eBURST analysis Results compared with 9 strains previously analyzed by PCR ribotyping Strains identified by C DIFF QUIK CHEK COMPLETE®; multiplex PCR for toxigenic type | 95 toxigenic strains (73%), including 7 A−B+CDT− and 3 A+B+CDT+ (23 sequence types) 35 (27%) non-toxigenic strains (12 sequence types) Sequence type (ST)17 was most common (21.8%) MLST and eBURST showed 139 strains belonged to 7 groups and singletons; most A+B+CDT− (89/91, 98%) were classed into group 1 MLST and eBURST suggest most A+B+CDT− strains (including ST17, ST2, ST8) may be derived from ST28 | This study reported a prevalence of A−B+CDT− (5%) and A+B+CDT+ (2%), which is considered low compared with MLST studies in China and Spain |
Mikamo et al. [52] | May 2012–May 2015 | Phase 3, multicenter (35 in Japan), double-blind RCT, n = 93 CDI diagnosed by EIA (97%) and stool culture (3%) | Adults (≥ 18 years) prescribed SOC antibiotics for CDI with planned duration 10–14 days Inpatients, n = 86; ≥ 65 years, n = 85 | Diarrhea (≥ 3 loose stools/24 h) + positive stool test for toxigenic C. diff | Cell culture cytotoxicity assays, stool culture with toxigenic strain typing, stool culture with toxin detection from C. diff isolates or commercially available assays (ELISA/PCR with ≥ 94% specificity) | 54 strains identified from culture. PCR ribotypes were 052 (28%), 018 (19%), 002 (15%), 369 (9%), 159 (6%), 005 (4%), 173 (4%), 012 (2%), 014 (2%), 043 (2%), 056 (2%), 103 (2%), 212 (2%), 235 (2%), 254 (2%), 632 (2%). 052 isolated from 11 of 35 sites and 018 isolated from 9 of 35 sites | – |
Mori et al. [42] | 12-month period in 2010 | Single-center retrospective analysis of stool culture database to study extent/ reasons for incorrect diagnosis of CDI | n = 975 stool culture samples | Definitions: toxigenic C. diff, C. diff with any toxin gene; CDI, diarrhea plus a toxigenic C. diff isolate | PCR assay of toxin gene A, B and binary PCR ribotyping Incidence of healthcare-facility onset CDI (within 48 h) estimated at 1.6 cases/10,000 patient-days | The prevalence rate of toxigenic C. diff in all stool cultures was 13% (127/975) 177 C. diff isolates detected of which 127 were toxigenic: 124 (70%) A+B+; 3 (1.7%) A−B+ The most common ribotype was 369 (21.6%), with 018 (10.8%); 014/020 and 002 were 9.9% each Clinically important isolates such as 027 and 078 were not identified 58 (45.7%) with toxigenic C. diff had unformed stool; incidence of CDI was 1.6/10,000 patient-days But of these 58 cases, 40 were not diagnosed in routine testing due to lack of clinical suspicion (24.1%) or a negative C. diff toxin assay result (44.8%) | Among A+B+, 12/177 (6.8%) were CDT+ |
Oka et al. [67] | 2002–2005 | Two-center study of molecular characterization of C. diff isolates from single, relapse and recurrent cases | n = 73 clinical isolates of C. diff (n = 20 isolates from 20 single infections; n = 53 isolates from 20 recurrent cases) | As test methods | PFGE and PCR ribotyping, and PCR toxin detection | 11 ribotypes Of 73 strains studied, 67 strains (91.8%) A+B+; 2 were toxin A−, B+ [B+] (2.7%); 4 (5.4%) were A−B− 80% of relapses were caused by the same strain as the first infection; 20% were due to a different strain | – |
Sawabe et al. [50] | November 1999–October 2004 | Molecular analysis of C. diff isolates linked with diarrhea or colitis at a single center | n = 148 isolates | As test methods | PCR and PFGE ribotyping Toxin (A, B and CDT) determined by PCR | 26 PCR ribotypes among 148 isolates Shift from predominant ribotype a (15/33; 45% in 2000) to ribotype f (identical to smz) (18/28; 64% in 2004) PFGE allowed further sub-classification: f isolates were of 4 types and 11 subtypes Only one ribotype 027 recovered | 110/148 (74%) A+B+CDT−; 33/148 (22%) A−B+CDT−; 5/148 (3%) A+B+CDT+ |
Senoh et al. [51] | April 2011–March 2013 for non-outbreak 2010 and 2009 outbreak data | Multicenter study to assess C. diff isolates in Japan | n = 120 C. diff isolates during a non-outbreak season; n = 18 and n = 21 isolates from hospitals during outbreaks | As test methods | Toxin detection and typing by PCR | 120 outbreak isolates: 80% were A+B+CDT−, 15.8% were A−B+CDT−; 4.2% A+B+CDT+ PCR-ribotype smz (A+B+CDT−) accounted for 34.2% isolates All A−B+CDT− isolates were PCR ribotype trf Non-outbreak isolates: Japan ribotypes smz (018) and ysmz 39.2%, and Japan ribotype trf 15.8% Types smz/ysmz also predominated in outbreaks 5 binary toxin-positive isolates (only 1 was 027 and 1 was 078) All trf isolates were A−B+ (new ribotype 369) High rates of resistance to antimicrobials observed in the 018 isolates | See ‘Isolates and strains’ |
Shimizu et al. [41] | April 2013–March 2014 | Study to evaluate differences in disease severity score according to toxigenic culture testing and GDH/EIA testing (single center) | n = 334 fecal samples from patients with diarrhea | Severe CDI defined as pseudomembranous colitis on endoscopy, admission to ICU or any two of age > 60 years, temperature > 38.3 °C, serum albumin < 2.5 g/dL, white cell count > 15,000 cells/mm3 | Simultaneous detection of GHD and toxins A/B by C DIFF QUIK CHEK COMPLETE® | 252 GDH-negative/EIA toxin-negative (i.e. no CDI) 82 GDH-positive, of which 25 were EIA-positive (CDI) and 57 EIA toxin-negative (equivocal cases) When toxins were detected in the initial screening test (GDH-positive/EIA toxin-positive), cases were more severe than in those only identified after toxigenic culture | – |
Yuhashi et al. [68] | Retrospective assessment of cases tested for C. diff diarrhea (single center) n = 68 | As test methods | EIA testing for C. diff diarrhea (C DIFF QUIK CHEK COMPLETE®) Patients grouped as toxin-positive stool; toxin-negative/toxin-positive isolate; dual toxin negative (stool and isolate) | 39 toxin-positive; 14 toxin-positive isolate group; and a dual toxin-negative stool and isolate group n = 15. All cases confirmed to be GDH positive by EIA Toxin-negative stool specimens associated with shorter diarrhea duration | – |
Recurrence
Reference | Patient population | Treatment of initial CDI | Definition of recurrence | Recurrence | Recurrence risk factors |
---|---|---|---|---|---|
Akahoshi et al. [24] | HSCT (n = 102 autologous; n = 206 allogeneic); n = 30 with CDI | Oral metronidazole (500 mg three-times daily, 10–14 days) | New episode of diarrhea and positive toxin EIA within 365 days after first episode of CDI | 1/30 (3.3%) within 100 days after HSCT | NR |
Daida et al. [31] | Pediatric patients (aged 0–19 years) admitted to hospital with cancer | Oral metronidazole (30 mg/kg) for > 10 days until resolution of symptoms and neutrophil recovery to > 500/μL | Presence of CDI 2 weeks after resolution of primary CDI symptoms | 13/51 (26%) | Statistical tests not performed. Recurrence more common in younger age (0–3 years; 9/13, 69%) than older children (19/38, 50%) |
Hashimoto et al. [26] | Retrospective chart review (single center) of 242 living donor liver transplant recipients (adults) n = 11 with C. diff diarrhea | Oral vancomycin (n = 8; dose not given) or conservative management (no detail given) | No definition given, but patients assessed from hospital admission to 3 months after transplant | 3/11 (27.3%); 2/8 (25%) in patients who received vancomycin | NR |
Hikone et al. [20] | In- and outpatient samples tested for C. diff (n = 2193 samples) n = 76 with healthcare-associated CDI | Oral metronidazole or vancomycin (doses not given) for median 14 days (range 6–52 days) | New episode of CDI within 8 weeks from the previous episode; diagnosis based on presence of diarrhea and positive toxin EIA | 14/76 (18.4%) | Univariate analysis: no risk factors identified Multivariate analysis: malignant disease (OR 7.98; 95% CI 1.22–52.2; p = 0.03) and ICU hospitalization (OR 49.9; 95% CI 1.01–2470; p = 0.049) |
Honda et al. [18] | CDI cases in a non-outbreak setting n = 126 with CDI (86.5% were healthcare-facility onset CDI) | Oral metronidazole (500 mg three-times daily), oral vancomycin (125 mg or 500 mg four-times daily), combination oral metronidazole (500 mg three-times daily) plus vancomycin (125 mg or 500 mg four-times daily), combination oral metronidazole plus rectal vancomycin, combination oral and rectal vancomycin, or no treatment (stop unnecessary antimicrobials) | New episode of diarrhea and positive toxin assay within 30 days since last date of completing therapy for first CDI episode | 8/126 (6%) | NR |
Hosokawa et al. [25] | Allogeneic HSCT patients (135 unrelated CBT; 39 unrelated BMT and 27 related PBSCT) n = 17 with C. diff diarrhea | Oral metronidazole or oral vancomycin (dosage and duration not given) | New episode of diarrhea and positive toxin test within 8 weeks after improvement of first properly treated episode | 0 | NR |
Iwashima et al. [44] | CDI cases in a hospital setting n = 71 consecutive patients with CDI | No formal regimens specified; 3 patients received treatment prior to recurrent CDI. Vancomycin mentioned (total dose range 6–12 g given for range of 1–28 days) | New CDI episode within 2 months after recovery from previous CDI episode | 9/71 (12.7%) | NR |
Kobayashi et al. [21] | Patients aged ≥ 14 years with hospital-onset CDI, n = 160 | Metronidazole (n = 88, 55%), vancomycin (n = 52, 33%), metronidazole + vancomycin (n = 10, 6.3%), no antimicrobial (n = 10, 6.3%). Doses and duration of therapy not specified | According to SHEA/IDSA 2010 guidelines [69] – re-emergence of CDI symptoms, according to infectious disease physician judgment, ≤ 4 weeks after completion of treatment for initial CDI episode | 23/160 (14%) | Recurrence did not differ according to severe/non-severe CDI or according to whether treatment adhered/did not adhere to clinical guidelines |
Mori et al. [42] | Stool culture database n = 58 cases with CDI | Vancomycin or metronidazole (dosage and duration not given) | No definition given, but recurrence recorded within 60 days following symptom onset | 3/58 (5.2%) | NR |
Oshima et al. [40] | Adults and pediatric (≤ 18 years) patients receiving PPI who developed acute-onset diarrhea (n = 17,217). Also, control group (n = 286,018) Recurrent CDI (reported in 9 studies) occurred in n = 1279; n = 5459 in the control group | – | Not given in this systematic review and meta-analysis of published studies, but based on recurrence, as reported in published studies. CDI presence based on laboratory confirmation of C. diff or clinical definition | – | PPI use increased risk for recurrent CDI (pooled OR 1.73, 95% CI 1.39–2.15; p = 0.02) |
Shimizu et al. 2015 [41] | Patients in a hospital setting with diarrhea (n = 334 fecal samples) n = 28 patients with severe CDI | Metronidazole, vancomycin (dosage and duration not given) or no treatment | No definition given and duration of assessment not detailed | Overall: 7/28 (25%); 5/16 (31.3%) in patients with GDH-positive/EIA toxin-positive test and 2/12 (16.7%) in patients with initial GDH-positive/EIA toxin-negative test, but who had confirmed positive toxigenic culture | No difference in incidence of recurrence between the two groups (p = 0.662) |
Takahashi et al. [39] | National Hospital Organization cohort Assessed for newly diagnosed CDI (n = 878 patients) and matched controls (no CDI) | Metronidazole, vancomycin (dosage and duration not given) or no treatment | No definition given, but recurrence assessed and recorded within 30 days of initial CDI episode | 34/714 (4.8%) among patients treated for CDI | NR |