Association of fluoroquinolones or cephalosporin plus macrolide with Clostridioides difficile infection (CDI) after treatment for community-acquired pneumonia.

OBJECTIVE: Clostridioides difficile infection (CDI) is the most common cause of gastroenteritis, and community-acquired pneumonia (CAP) is the most common infection treated in hospitals. American Thoracic Society (ATS)/Infectious Diseases Society of America (IDSA) CAP guidelines recommend empiric therapy with a respiratory fluoroquinolone or cephalosporin plus macrolide combination, but the CDI risk of these regimens is unknown. We examined the association between each antibiotic regimen and the development of hospital-onset CDI. METHODS: We conducted a retrospective cohort study using data from 638 US hospitals contributing administrative including 177 also contributing microbiologic data to Premier, Inc. We included adults admitted with pneumonia and discharged from July 2010 through June 2015 with a pneumonia diagnosis code who received ≥3 days of either empiric regimen. Hospital-onset CDI was defined by a diagnosis code not present on admission and positive laboratory test on day 4 or later or readmission for CDI. Mixed propensity-weighted multiple logistic regression was used to estimate the associations of CDI with antibiotic regimens. RESULTS: Our sample included 58,060 patients treated with either cephalosporin plus macrolide (36,796 patients) or a fluoroquinolone alone (21,264 patients) and with microbiological data; 127 (0.35%) patients who received cephalosporin plus macrolide and 65 (0.31%) who received a fluoroquinolone developed CDI. After adjustment for patient demographics, comorbidities, risk factors for antimicrobial resistance, and hospital characteristics, CDI risks were similar for fluoroquinolones versus cephalosporin plus macrolide (odds ratio [OR], 0.98; 95% confidence interval [CI], 0.70-1.38). CONCLUSION: Among patients with CAP at US hospitals, CDI was uncommon, occurring in ∼0.33% of patients. We did not detect a significant association between the choice of empiric guideline recommended antibiotic therapy and the development of CDI.

Association of antibiotic use and acute kidney injury in patients hospitalized with community-acquired pneumonia.

BACKGROUND: Acute kidney injury (AKI) is common among hospitalized patients with community-acquired pneumonia (CAP). We aimed to estimate and compare the risk of AKI for various antibiotic combinations in adults hospitalized for CAP. METHODS: We conducted a retrospective cohort study of the Premier Healthcare Database containing all admissions for 660 US hospitals from 2010 to 2015. We included adults aged ≥18 years hospitalized with CAP and considered 6 different antibiotic combinations based on continuous use in the first 3 hospital days. The primary outcome was incident AKI, defined by ICD-9 codes 584.5-584-9. We evaluated associations of AKI with in-hospital mortality and length-of-stay. We excluded patients who were admitted directly to the intensive care unit, had AKI codes present on admission or had dialysis in the first 2 days. We used generalized linear mixed models with the hospital as a random effect and covariate adjustment for patient demographics, comorbidities, other treatments on day 0/1, and hospital characteristics. RESULTS: The total sample included 449,535 patients, 3.15% of whom developed AKI. All other regimens but fluoroquinolones exhibited higher AKI odds than 3rd generation cephalosporin with or without macrolide. The combination of piperacillin/tazobactam and vancomycin with or without other antibiotics was associated with the highest AKI odds (OR = 1.89; 95% CI: 1.73-2.06). Patients with incident AKI had an increased odds of hospital mortality (OR = 6.37; 95% CI: 6.07-6.69) and longer length-of-stay (mean multiplier = 1.84; 95% CI: 1.82, 1.86). CONCLUSION: Compared to 3rd generation cephalosporin with or without macrolide, piperacillin/tazobactam, vancomycin, and their combination were associated with higher odds of developing AKI, which in turn were associated with worse clinical outcomes.

Repeat stool testing to diagnose Clostridium difficile infection using enzyme immunoassay does not increase diagnostic yield.

BACKGROUND & AIMS: Clostridium difficile infection (CDI) is a hospital-acquired infection with increasing incidence and severity. The most frequently used test to diagnose CDI is an enzyme immunoassay (EIA) for toxins A and B in stool samples. It is common to test 2 or more stool samples, based on the assumption that this detects CDI with greater sensitivity than analysis of 1 sample. We investigated whether repeat stool testing significantly improves the diagnostic yield for CDI. METHODS: We performed a retrospective analysis of hospitalized patients who were tested for CDI using EIA. From year 2005 to 2008, 39,402 stool samples from 17,971 patients with 29,373 diarrhea episodes were tested. Transition probabilities were calculated based on results from repeated tests. RESULTS: A total of 2692 diarrheal episodes (9.17%) were diagnosed with CDI. Based on results of 3 consecutive tests, 2675 (99.36%) were diagnosed with CDI. The first stool sample tested produced positive results for 90.7% of cases. When samples were tested consecutively, for the second and third time, an additional 6.6% and 2% patients had positive test results, respectively. If the first test result was negative, the probability of the second test result being positive was 2.7%. If the first 2 test results were negative, the probability of the third test result being positive was 2.3%. CONCLUSIONS: In patients who had multiple stool samples tested for CDI by EIA, almost 91% were accurately diagnosed based on the results of a single stool sample alone. Subsequent testing yielded a positive result in only 8.6% of patients. We therefore recommend that repeat testing not be done on a routine basis because it does not significantly improve diagnostic yield.

Fusobacterium nucleatum: a rare cause of pyogenic liver abscess.

While pyogenic liver abscesses are uncommon, they are associated with significant mortality and morbidity. Most pyogenic liver abscesses are polymicrobial and are caused by enteric bacteria and anaerobes. Rarely, mono-microbial infections may occur, typically in immunocompromised individuals. We report the unusual case of a 69 year-old immunocompetent female who developed a pyogenic liver abscess due to Fusobacterium nucleatum infection, likely from a dental source. Poor oropharyngeal hygiene seems to have a major role in infection from this organism and therefore F. nucleatum should be considered as a differential for causes of pyogenic liver abscess in such patients. Drainage of the abscess and antibiotic therapy are the mainstays of therapy.

Repeat stool testing for Clostridium difficile using enzyme immunoassay in patients with inflammatory bowel disease increases diagnostic yield.

BACKGROUND: The incidence and severity of Clostridium difficile infection (CDI) in patients with inflammatory bowel disease (IBD) is increasing. CDI is diagnosed by toxin enzyme immunoassay (EIA) or real-time polymerase chain reaction (PCR) performed on stool samples. An earlier study evaluating EIA in IBD patients with CDI suggested that more than one stool sample be tested to increase diagnostic yield. We investigated whether repeat stool testing improves diagnostic accuracy for CDI in hospitalized IBD patients compared to hospitalized patients with CDI and no IBD. METHODS: We performed retrospective data analysis from January 2005-May 2011 on 63,086 hospitalized patients who were tested for CDI using EIA or PCR. Of these, 2579 patients had IBD. Transition probabilities were calculated based on results from repeated tests. RESULTS: Inclusive of all inpatients tested for CDI, 56,583 were tested using toxin EIA and 6503 were tested using PCR. In patients with no IBD, the first stool sample tested was positive in 90% and 94% with EIA and PCR respectively. In IBD patients tested using EIA, 101 were diagnosed with CDI. The first stool sample tested was positive in 81% of patients. Successive second and third stool samples yielded additional 14% and 5% CDI positive IBD patients. CONCLUSIONS: Approximately one in five IBD patients with CDI required repeat testing to yield a positive result with EIA. There are minimal diagnostic gains of repeat testing by EIA or PCR in patients without IBD. We recommend repeat stool testing for CDI when using EIA to increase diagnostic yield in IBD patients.