Yuan Pin Hung

Impact of toxigenic Clostridium difficile colonization and infection among hospitalized adults at a district hospital in southern Taiwan.

Background The impact of toxigenic Clostridium difficile colonization (tCDC) in hospitalized patients is not clear. Aim To study the significance of tCDC in hospitalized patients. Methods A prospective study in the medical wards of a regional hospital was performed from January to June 2011. Fecal samples collected from patients at the time of admission were tested for tcdB by real-time polymerase chain reaction (PCR) and cultured for C. difficile. The patients were followed up weekly or when they developed diarrhea during hospitalization. If C. difficile was isolated, tcdA and tcdB would be tested by multiplex PCR. The primary outcome was the development of C. difficile-associated diarrhea (CDAD). Findings Of 168 patients enrolled, females predominated (87, 51.8%), and the mean patient age was 75.4 years old. Approximately 70% of the patients were nursing home residents, and one third had a recent hospitalization within the prior three months. Twenty-eight (16.7%) patients had tCDC, including 16 (9.5%) patients with tCDC at the time of admission and 12 (7.2%) with tCDC during the follow-up period. With regard to the medications taken during hospitalization, the patients were more likely to have tCDC if they had received more than one class of antibiotics than if they had received monotherapy (odds ratio [OR] 6.67, 95% confidence interval [CI] 1.41–31.56, P = 0.01), particularly if they received a glycopeptide in combination with a cephalosporin or penicillin or a cephalosporin and a carbapenem. More patients with tCDC developed CDAD than those without tCDC (17.9%, 5/28 vs. 1.4%, 2/140, P = 0.002). Overall 7 (4.2%) of the 168 patients developed CDAD, and crude mortality rate of those with and without tCDC was similar (21.4%, 6/28 vs. 19.4%, 27/140, P = 0.79). Conclusion Recent use of glycopeptides and β-lactam antibiotics is associated with toxigenic C. difficile colonization, which is a risk factor for developing C. difficile-associated diarrhea.

Clostridium difficile Infection at a Medical Center in Southern Taiwan: Incidence, Clinical Features and Prognosis

BACKGROUND/PURPOSE An increase in incidence of Clostridium difficile infection (CDI) among Western countries has been noted in recent years. Epidemiological data of CDI are scarce in Taiwan. This study is intended to depict the clinical features of CDI at a medical center in Southern Taiwan. METHODS From January 1, 2007 to March 31, 2008, hospitalized patients with CDI (defined as the presence of gastrointestinal symptoms and fecal C. difficile toxin) were identified. Their medical records were reviewed for further evaluation. RESULTS A total of 86 cases of CDI were identified in the study period. The incidence was 42.6 cases per 100,000 patient-days, or 3.4 cases per 1,000 discharges, and was highest in intensive care units (110.6 cases per 100,000 patient-days). Variable incidence rates were noted in different wards, and prevalence was higher in the infectious ward. Diarrhea, fever, and abdominal distension were common in 82 (95.3%), 47 (54.7%), and 29 (33.7%) patients, respectively. Metronidazole was the initial therapeutic regimen for 83 (96.5%) patients. Prolonged diarrhea was noted in 31 (36.4%) patients, especially in those on hemodialysis therapy. Recurrence was noted in 7 (8.1%) patients. Fecal carriage of vancomycin-resistant Enterococcus colonization was found in three patients after therapy for CDI. All-cause mortality rate of patients with CDI at 30 days was 23.3%. CONCLUSION CDI is increasingly being recognized within the medical departments, and should be considered in hospitalized adults with diarrhea, fever, or abdominal distension alone, or in combination.

Risk factors and clinical impact of levofloxacin or cefazolin nonsusceptibility or ESBL production among uropathogens in adults with community-onset urinary tract infections.

BACKGROUND Gram-negative bacilli causing community-onset urinary tract infections (CoUTIs) are getting increasingly resistant to antimicrobial agents. Clinical significance and risk factors of the acquisition of antimicrobial-nonsusceptible pathogens are still under investigation. METHODS A prospective study was performed in the medical wards of two hospitals in southern Taiwan between August 2009 and January 2012. Patients were enrolled if they were aged >18, admitted through the emergency department, and had CoUTI due to Enterobacteriaceae isolates. RESULTS Overall 136 adults with CoUTI were enrolled. Their mean age was 67 years and females were predominant (68.4%). Comorbidities, such as diabetes mellitus (30.1%) and hypertension (54.4%), were common. Escherichia coli (111, 81.6%) was the predominant species, followed by Klebsiella pneumoniae (11, 8.1%), and Proteus mirabilis (7, 5.1%). Nine (8.0%) of E. coli isolates and 5 (45%) of K. pneumoniae isolates had extended-spectrum β-lactamase (ESBL) production. Out of 122 non-ESBL producing isolates, 35 (28.7%) and 31 (25.4%) were nonsusceptible to levofloxacin and cefazolin, respectively. In the multivariate analysis, several clinical characters were found to be independently associated with CoUTIs due to levofloxacin-nonsusceptible (i.e. males, recent hospitalization, underlying old stroke, diabetes mellitus, and altered consciousness, or absence of chills, pyuria, or tachycardia), cefazolin-nonsusceptible (i.e. males, recent hospitalization, underlying old stroke, absence of fever or chills), or ESBL-producing isolates (i.e. recent hospitalization or antimicrobial therapy). All patients survived and discharged. However, the patients with CoUTIs due to levofloxacin-nonsusceptible (16.1 vs. 7.5 days, p < 0.01), cefazolin-nonsusceptible (15.4 vs. 8.4 days, p < 0.01) or ESBL-producing (16.7 vs. 9.6 days; p < 0.01) pathogens had a longer hospitalization stay than those due to their susceptible comparators. CONCLUSION Several host factors were recognized to be independently associated with the acquisition of UTIs due to levofloxacin- or cefazolin- nonsusceptible, or ESBL-producing Gram-negative bacilli. The clinical impact of UTIs due to nonsusceptible uropathogens is that they result in the prolongation of hospital stays.

Clinical impact of Clostridium difficile colonization

Clostridium difficile can cause antibiotic-associated diarrhea in hospitalized patients. Asymptomatic colonization by C. difficile is common during the neonatal period and early infancy, ranging from 21% to 48%, and in childhood. The colonization rate of C. difficile in adult hospitalized patients shows geographic variation, ranging from 4.4% to 23.2%. Asymptomatic carriage in neonates caused no further disease in many studies, whereas adult patients colonized with toxigenic C. difficile were prone to the subsequent development of C. difficile-associated diarrhea (CDAD). However, the carriage of nontoxigenic C. difficile strains appears to prevent CDAD in hamsters and humans. Risk factors for C. difficile colonization include recent hospitalization, exposure to antimicrobial agents or gastric acid-suppressing drugs (such as proton-pump inhibitors and H2 blockers), a history of CDAD or cytomegalovirus infection, the presence of an underlying illness, receipt of immunosuppressants, the presence of antibodies against toxin B, and Toll-like receptor 4 polymorphisms. Asymptomatic C. difficile carriers are associated with significant skin and environmental contamination, similar to those with CDAD, and contact isolation and hand-washing practices should therefore be employed as infection control policies for the prevention of C. difficile spread. Treating patients with asymptomatic C. difficile colonization with metronidazole or vancomycin is not suggested by the currently available evidence. In conclusion, asymptomatic C. difficile colonization may lead to skin and environmental contamination by C. difficile, but more attention should be paid to the clinical impact of those with C. difficile colonization.

Risk factors for Clostridium difficile-associated diarrhea among hospitalized adults with fecal toxigenic C. difficile colonization.

BACKGROUND Patients with toxigenic Clostridium difficile colonization (tCDC) are at risk of developing C. difficile-associated diarrhea (CDAD). However, the risk factors of hospitalized patients with tCDC developing CDAD are not clear. METHODS We conducted an 18-month prospective study at a medical ward in a district hospital in southern Taiwan. Within 48 hours of admission, weekly stool samples from asymptomatic hospitalized patients were obtained to detect fecal CDC. A polymerase chain reaction for tcdB was performed to determine toxigenic isolates. CDAD was diagnosed if the patient had diarrhea and toxigenic C. difficile present in a stool sample. RESULTS A total 483 patients with stool samples were eligible for the study. Eighty-six (17.8%) patients had tCDC after screening, of whom 14 (16.3%) developed CDAD during follow-up. Among those with tCDC, patients with subsequent CDAD were more likely to have diabetes mellitus (p = 0.01) and to have received piperacillin-tazobactam (p = 0.04), or proton-pump inhibitors (PPIs; p = 0.04) than those without developing CDAD. The variables were statistically significant as determined by multivariate analysis. However, the 60-day crude mortality rates among tCDC patients with and without subsequent development of CDAD were similar. CONCLUSION Diabetes mellitus and recent receipt of piperacillin-tazobactam or PPIs are independent risk factors for the development of CDAD among hospitalized patients with tCDC.

Risk Factors of Fecal Toxigenic or Non-Toxigenic Clostridium difficile Colonization: Impact of Toll-Like Receptor Polymorphisms and Prior Antibiotic Exposure

Background This study is to investigate the significance and risk factors of fecal toxigenic (tCdC) or non-toxigenic Clostridium difficile colonization (ntCdC) among hospitalized patients. Methods Adults admitted to medical wards in a district hospital between January 2011 and June 2012 were enrolled, and those with a history of colectomy, C. difficile fecal colonization or infection or receipt of either metronidazole or oral vancomycin within 3 months, were excluded. Stools collected within 48 hours after admission and every week during hospitalization were cultured for C. difficile. Findings Among the 441 enrolled patients, 84 (20.0%) had CdC at initial screening, including 58 (13.2%) with tCdC and 26 (6.8%) with ntCdC. Among patients with initial negative fecal screening for CdC, it took an average of 70.6 days or 66.5 days to develop tCdC or ntCdC during the study period. Finally 78 (17.7%) had tCdC and 34 (7.7%) had ntCdC. During the follow-up period, the patients with tCdC had a higher risk of CDAD (11/79, 14.1%) than those without CdC (3/328, 0.9%) and those with ntCdC (0/34, 0%) (P<0.001). In multivariate analysis, the TLR4 rs1927914 polymorphism (GG genotype) (odds ratio [OR] 4.4, 95% confidence interval [CI] 1.6–11.8, P = 0.003) and recent cefepime therapy (OR 5.3, 95% CI 2.1–13.2, P<0.001) were independently associated with tCdC, whereas recent cefuroxime (OR 11.7, 95% CI 2.3–60.2, P = 0.003) and glycopeptide therapy (OR 10.9, CI: 2.1–57.2, P = 0.005) associated with ntCdC. Conclusion The incidence of CDAD is highest in patients with tCdC and lowest in patients with ntCdC, and the TLR4 rs1927914 polymorphism GG genotype and recent cefepime therapy were independently associated with tCdC.

Incidence of Aeromonas bacteremia in Southern Taiwan: Vibrio and Salmonella bacteremia as comparators

The aim of the investigation was to describe the incidence of Aeromonas bacteremia in a city with a population of about 1.87 million inhabitants, located in southern Taiwan, between 2008 and 2010. Such data were compared with the incidences of Vibrio and Salmonella bacteremia in the same period and the incidence of Aeromonas bacteremia in other countries in the literature. The data revealed the average annual incidences of bacteremia due to Aeromonas, Vibrio, and Salmonella species were 76, 38, and 103 cases/million inhabitants, respectively. The incidence of Aeromonas bacteremia was higher than those in Western countries.

Cytomegalovirus disease in nonimmunocompromised, human immunodeficiency virus-negative adults with chronic kidney disease

BACKGROUND/PURPOSE(S) To identify the clinical characteristics of cytomegalovirus (CMV) disease in chronic kidney disease (CKD) patients. METHODS Patients from two sources were reviewed: (1) a retrospective study of hospitalized patients admitted between January 1990 and February 2009 was performed at a tertiary hospital in Taiwan; (2) the English literature from 1990 to 2009 was reviewed for additional cases, and adults with CKD and histopathologically documented cytomegalovirus disease were included. RESULTS Seven CKD patients from our hospital and seven from the literature were included. Nine (64.3%) patients were males, and the mean age was 66 years. Histopathologically proven CMV disease was present in the gastrointestinal (GI) tract of 13 (92.9%) and in the skin of one (7.1%) patient. GI symptoms included bleeding (78.6%), abdominal pain (35.7%), and diarrhea (28.6%).The most common comorbidities were diabetes mellitus (7, 50%) and hypertension (8, 57.1%). Thirteen patients had CMV GI disease. The endoscopic gross features of the GI tract lesions included single or multiple ulcers and a large polypoid or uneven surface mass. Of the seven cases with available data, a low body mass index (22.3 ± 1.3 kg/m(2)) and hypoalbuminemia (25 ± 7.0 g/L) were noted. Twelve patients had received ganciclovir or valganciclovir therapy. Five (35.7%) patients died, and the death of two patients was directly related to bowel perforation caused by CMV colitis. CONCLUSION CMV disease may occur in CKD patients without the presence of overt immunodeficiency. The gastrointestinal tract is the most common site of involvement. Clinicians should be aware of this possibility in CKD patients who have GI symptoms.

Proton-Pump Inhibitor Exposure Aggravates Clostridium difficile–Associated Colitis: Evidence From a Mouse Model

BACKGROUND Clostridium difficile is currently the leading cause of infectious diarrhea in hospitalized patients. In addition to the infection due to toxigenic C. difficile in the gastrointestinal tract of susceptible hosts, other predisposing factors for C. difficile infection (CDI) are identified, including advanced age, a prolonged hospital stay, and use of acid-suppressive drugs. Of note, exposure to gastric acid-reducing agents, such as H2 blockers and proton pump inhibitors (PPIs), remains a controversial risk factor, and has been associated with CDI in some studies but not in others. A mouse model of antibiotic-associated clostridial colitis was established to examine the role of PPIs for CDI. MATERIALS AND METHODS A mouse model of antibiotic-associated clostridial colitis was set up. NF-κB reporter mice were used to address the in vivo spatial and temporal inflammatory patterns of C. difficile-associated colitis. Serum levels of lipopolysaccharide and dextran-FITC were measured to reflect the barrier permeability of affected intestines. RESULTS Mice with CDI that were exposed to PPI exhibited greater losses of stool consistency and body and cecal weights than those that were not exposed to PPI. Further, more neutrophilic infiltrations, epithelial damage, and inflammatory cytokine expression were noted in colon specimens of the mice with PPI exposure. More-evident inflammatory responses were detected by in vivo imaging of NF-κB reporter mice with CDI that were exposed to PPI. Gut barrier permeability was increased to a greater extent, as reflected by higher serum levels of lipopolysaccharide and dextran-FITC in mice with CDI that were exposed to PPI. CONCLUSIONS Our mouse model demonstrates that PPI exposure increases the severity of intestinal inflammation in mice with C. difficile-associated colitis.

Propensity score-matched analysis comparing the therapeutic efficacies of cefazolin and extended-spectrum cephalosporins as appropriate empirical therapy in adults with community-onset Escherichia coli, Klebsiella spp. and Proteus mirabilis bacteraemia

In this study, the therapeutic efficacy of cefazolin was compared with that of extended-spectrum cephalosporins (ESCs) (cefotaxime, ceftriaxone and ceftazidime) as appropriate empirical therapy in adults with community-onset monomicrobial bacteraemia caused by Escherichia coli, Klebsiella spp. or Proteus mirabilis (EKP). Compared with cefazolin-treated patients (n = 135), significantly higher proportions of patients in the ESC treatment group (n = 456) had critical illness at bacteraemia onset (Pitt bacteraemia score ≥4) and fatal co-morbidities (McCabe classification). Of the 591 patients, 121 from each group were matched using propensity score matching (PSM) based on the following independent predictors of 28-day mortality: fatal co-morbidities (McCabe classification); Pitt bacteraemia score ≥4 at bacteraemia onset; initial syndrome of septic shock; and bacteraemia due to pneumonia. After appropriate PSM, no significant differences were observed in the early clinical failure rate (10.7% vs. 7.4%; P = 0.37), the proportion of critical illness (Pitt bacteraemia score ≥4) (0% vs. 0%; P = 1.00) and defervescence (52.6% vs. 42.6%; P = 0.13) on Day 3 between the cefazolin and ESC treatment groups. Similarly, no significant differences were observed in the mean of time to defervescence (4.1 days vs. 4.9 days; P = 0.15), late clinical failure rate (18.2% vs. 10.7%; P = 0.10) and 28-day crude mortality rate (0.8% vs. 3.3%; P = 0.37) between the two groups. These data suggest that the efficacy of cefazolin is similar to that of ESCs when used as appropriate empirical antimicrobial treatment for community-onset EKP bacteraemia.