Evaluation of Vitek II™ (VT2), penicillin disk susceptibility testing (DST), and nitrocephin testing for detecting the presence of blaZ-postivity in bacteraemic isolates of methicillin-susceptible S. aureus (MSSA)
Margot A Rosenthal, Vanessa Poliquin, Carol Schneider
Department of Obstetrics, Gynecology and Reproductive Sciences, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada
Background: The incidence of syphilis among women of child-bearing age has risen exponentially in several jurisdictions in Canada. While national guidelines assist clinicians in managing the maternal infection during pregnancy, there is little evidence available to guide management in situations where the developing fetus is already severely affected by congenital infection.
Case Presentation: We present the case of a 29-year-old female, initially presenting to care as a syphilis contact. She presented to the nursing station pregnant, with significant stiffness and pain in her neck and upper back with difficulty ambulating over the last month, concerning for neurosyphilis. Ultrasound performed in our tertiary care facility demonstrated single live intrauterine pregnancy with a gestational age of 19+3 weeks. The fetus was hydropic with large pleural effusions, ascites, and markedly abnormal brain. The placenta was cystic and enlarged. Amniocentesis was done and eventually confirmed fetal syphilis infection on polymerase chain reaction. Penicillin G was initiated at 24 million units per day divided q4h per neurosyphilis dosing protocols. After 9 days of therapy, maternal status had mildly improved. Fetal status had worsened; the fetal abdomen was grossly distended from ascites and measuring 13 weeks ahead of gestational age. The prognosis was extremely poor, given the timing of infection during development and anomalous appearing fetal brain. The parents chose termination of pregnancy.
Conclusion: This case highlights the potential severity of fetal syphilis infection and underscores the paucity of available current literature to guide management of a severely affected fetus in the antepartum period. Only one case study was available to suggest IV penicillin treatment for fetal infection. Information sharing should be encouraged as this history disease becomes more commonplace and much of the previous clinical acumen may have been lost over time.
Bhreagh C Phipps1, Guillaume Poliquin1,2, Paul Van Caeseele1,3
1University of Manitoba, Winnipeg, Manitoba, Canada; 2National Microbiology Laboratory, Winnipeg, Manitoba, Canada; 3Cadham Provincial Laboratory, Winnipeg, Manitoba, Canada
Objectives: The interplay between varicella zoster virus (VZV) and herpes zoster (HZ) is complex. Latency of HZ is predominantly controlled by VZV-specific cell-mediated immunity (CMI). As VZV-specific CMI decreases over time, CMI falls below a certain threshold, allowing for re-activation of latent VZV as HZ. VZV vaccination has successfully reduced circulation of wild-type VZV. This positive outcome, however, could theoretically result in an increase in HZ rates by decreasing the amount of exogenous boosting that occurs through wild-type VZV exposure. We hypothesize an increase in incidence of HZ following VZV vaccination and that HZ is occurring earlier in life.
Methods: Our institution’s health database was queried for individuals diagnosed with HZ from 1996 to 2016. Rates of HZ relative to the introduction of VZV vaccination were analyzed using interrupted time-series analysis. The rate and age of occurrence of HZ pre-vaccination was compared with the rate of HZ post-vaccination. The 3-, 6-, 12-, and 24-month level effects were derived from the data.
Results: Our analysis showed a significant (p < 0.05) direct intervention effect of 0.109 increase in incidence per quarter of HZ post-VZV vaccination in the 35–49 years age group (SE 0.034; 95% CI –0.041 to 0.177). It also shows a significant (p < 0.05) increase in post-vaccination incidence rates of HZ per quarter in the 10–19, 20–34, 35–49, 50–64, 65–79, and 80+ years age groups.
Conclusions: We have observed an increase of HZ incidence post-VZV vaccination above the forecasted pre-VZV vaccination trend. HZ is occurring at a higher incidence earlier in life. Although the 0–9 years age group showed a decrease in incidence of HZ shortly after vaccination, this effect was not long-lasting. Additional research may be required to determine the need for additional interventions, such as varicella boosters and/or earlier herpes zoster vaccination, to reduce risk of HZ in the future.
Fazia Tadount1,2, Zineb Laghdir1,2, Caroline Quach1,2,3,4
1Research Institute—CHU Sainte-Justine, Montréal, Québec, Canada; 2Infection Prevention & Control, CHU Sainte-Justine, Montréal, Québec, Canada; 3Department of Microbiology, Infectious Diseases, and Immunology, Faculty of Medicine, University of Montréal, Montréal, Québec, Canada; 4Department of Clinical Laboratory Medicine, CHU Sainte-Justine, Montréal, Québec, Canada
Background: Neuraminidase inhibitors (NAIs) are likely to be part of the rapid response and control in situations of influenza pandemics and institutional outbreaks. Health care workers (HCWs) are at risk for acquiring and transmitting influenza infections due to their direct contact with patients. Furthermore, absenteeism due to influenza-like illness (ILI) among HCWs may cause a substantial disruption in health care services during a pandemic. Although several systematic reviews (SR) have assessed the use of NAIs, none has assessed NAI use in HCWs to date. Therefore, we conducted an SR to identify and appraise the current evidence on the use of NAIs in HCWs in the context of an influenza pandemic.
Methods: We searched five electronic databases for observational studies and clinical trials that assessed the use of NAIs in HCWs. Two reviewers independently screened all references, then proceeded to data extraction and quality assessment using the Cochrane tools for included studies.
Results: We screened 5,256 unique citations by title/abstract and 130 references were deemed eligible for full-text screening. Eight studies were included in the review; seven assessed the effectiveness of NAIs in preventing influenza in HCWs and one assessed their effectiveness in preventing nosocomial transmission.
Conclusions: Data on NAI use in HCWs are scarce. Identified studies were at moderate to serious risk of bias, and many were underpowered. NAI prophylaxis may provide limited benefit in preventing influenza in HCWs, but more studies are required to determine the effectiveness of NAIs in HCWs, and the best strategy for their use in health care settings.
William Stokes1, John Lam1, Michael Parkins1, Daniel B Gregson1,2
1University of Calgary, Calgary, Alberta, Canada; 2Alberta Precision Laboratories, Calgary, Alberta, Canada
Background: S. aureus bacteriuria (SABU) has increasingly been recognized as an important clinical entity due to its potential association with S. aureus bacteremia (SAB). Unfortunately, fewer than 50% of patients identified with SABU in the Calgary Health Zone (CHZ) have blood cultures (BC) collected. This is a quality improvement project to determine if the inclusion of a passive, synchronous alert for clinicians to consider additional investigations in patients identified with SABU in the CHZ would yield higher compliance with BC collection.
Methods: A retrospective cohort study of hospitalized and emergency room patients with SABU was conducted in 2018–2019. Our study examined two time frames: 3 months before and after insertion of the alert on all SABU reports. Detailed chart reviews were performed to determine the proportion of BC collected in response to SABU. In addition, a standardized, qualitative feedback survey about the alert was distributed to a subset of front-line health care providers.
Results: A total of 64 and 67 SABU episodes were identified before and after alert insertion, respectively. BC ordered trended toward an increase in the 3 months following implementation of the alert (58.2% versus 45.3%, p = 0.14) and were more likely to be performed within 48 hours of SABU detection (76.9% versus 58.6%, p = 0.003). Of the 55 front-line health care providers surveyed (56% internal medicine residents, 25% family physicians, 19% hospitalists), 87.5% deemed the alert worthwhile to include on SABU reports, and 33.3% felt that the alert changed their clinical practice in assessing and managing SABU.
Conclusions: The initial analysis of this automated lab-generated warning suggests that clinicians were more likely to comply with guideline-directed investigations for SAB. This intervention has the potential to reduce the time to patient diagnosis and improve outcomes and warrants further study.
Arif Ismail1, William Stokes2, Byron M Berenger2,3, L Robbin Lindsay4, Raymond Tellier3, Andrew Johnson2
1University of Alberta, Edmonton, Alberta, Canada; 2University of Calgary, Calgary, Alberta, Canada; 3Public Health Laboratory (ProvLab), Calgary, Alberta, Canada; 4National Microbiology Laboratory, Winnipeg, Manitoba, Canada
Objective: Encephalitis presenting in persons returning from regions endemic to multiple flaviviruses poses a diagnostic challenge. This is due in part to serologic cross-reactivity, short periods of viremia, and potential co-infection with other encephalitis viruses. We present a case that demonstrates an approach to this diagnostic challenge.
Case Presentation: A 72-year-old female presented with 4 days of fever, headache, and acutely altered mental status after a 6-month visit to rural northwest India. She sought no pre-travel advice or vaccination. Her temperature and vital signs were normal. Glasgow Coma Scale was 10 (E3V2M5) in the absence of meningismus, focal neurologic deficit, or rash. Extensive imaging of the head and neck was normal. Dexamethasone and meningitic doses of vancomycin, ceftriaxone, ampicillin, and acyclovir were discontinued after 5 days following negative cerebrospinal fluid cultures and polymerase chain reaction. Her mental status recovered to baseline by day 18 without discernable long-term neurologic sequelae.
Results: Acute and convalescent serology demonstrated Dengue virus (DENV) antibody seroconversion, a greater than four-fold rise in DENV plaque reduction neutralization test (PRNT) titres from 1:80 to >1:640, and DENV viremia, consistent with acute DENV infection in a previously exposed host. In contrast, positive but static Japanese encephalitis virus (JEV) titres and absence of JEV viremia suggested prior JEV infection or vaccination rather than acute infection. West Nile virus (WNV) and Zika virus IgG antibodies also returned positive, consistent with past WNV infection based on high avidity and cross-reactive Zika antibodies based on a negative PRNT.
Conclusions: This case highlights the inherent difficulty of clinically distinguishing between the various presentations of flavivirus encephalitis and the diagnostic uncertainty arising from serologic cross-reactivity. Avidity assays and serial PRNT demonstrating greater than four-fold rise in titres or seroconversion can help with diagnosis, though neither is perfect.
Amro Qaddoura1, Megan McQuiston2, Greg Tyrrell3,4, Matthew Croxen3,4, Vincent Li3, Rhonda Demarco5, Suzanne Pinfield5, Ruziyya Ramazanova5, Karen Hope6, Edith-Rose Cairns6, Judy MacDonald6,7, Jia Hu6, Oscar Larios1,5, Joseph Kim1,5, Bayan Missaghi1,5, Joseph Vayalumkal5,8, Valeri Marsten2, Jennifer Soucie2, Robert Doug Wilson2, Colin Birch2, John Conly1,5
1Department of Medicine, University of Calgary and AHS-Calgary, Calgary, Alberta, Canada; 2Department of Obstetrics and Gynecology, University of Calgary and AHS-Calgary, Calgary, Alberta, Canada; 3Public Health Laboratories (ProvLab), Alberta Precision Laboratories, AHS-Edmonton, Edmonton, Alberta, Canada; 4Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, Alberta, Canada; 5Infection Prevention and Control, AHS-Calgary, Calgary, Alberta, Canada; 6Communicable Disease Control, Public Health, AHS-Calgary, Calgary, Alberta, Canada; 7Department of Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada; 8Department of Pediatrics, University of Calgary and AHS-Calgary, Calgary, Alberta, Canada
Objectives: Invasive group A streptococcal (iGAS) puerperal sepsis (PS) is a recognized complication that can have devastating consequences. We retrospectively analyzed PS cases in a large Canadian health region with epidemiologic and molecular techniques.
Methods: PS cases from parturition to 6 weeks post-partum between 2013 and 2018 were identified as confirmed (severe invasive disease with GAS isolated from a sterile site) or probable (GAS isolated from a non-sterile site in the absence of other causes) and conjointly adjudicated by public health/infection control physicians as hospital/delivery- or community-acquired based on medical record review. Available isolates from patients and health care workers (HCW) had emm typing, sic gene, and/or whole genome sequencing performed. The use of masking and personal protective equipment (PPE) was determined from delivery records.
Results: Eighteen cases were identified (1.6 cases/10,000 live births) with 16 (88.9%) adjudicated as hospital/delivery-acquired. There was 1 maternal death, and there were 8 intensive care unit admissions and 6 emergency hysterectomies. There were 10 emm types in total. Two temporo-spatially related cases had emm1, sic gene, and core single-nucleotide variant (SNV) analysismatched isolates to an isolate identified in an unmasked delivering HCW. Two additional temporo-spatially related cases who had the same delivery team had emm11and were found to match with core SNV analysis. Masks were not used in 13 (72.2%) cases, with information about PPE unavailable for the remaining 5 cases. We developed and launched health care policy interventions using this data, including universal masking and point-of-care PPE for all delivery personnel, which were officially accepted by our health region in February 2020.
Conclusions: PS secondary to hospital/delivery-acquired iGAS is a potentially preventable infection. Our epidemiologic and molecular data, which found that most cases of iGAS were hospital/delivery-acquired, underscores the need to carefully adjudicate all cases of PS and implement policy interventions for appropriate use of PPE for all vaginal deliveries. Our data drove policy changes and interventions within our health region.
Amro Qaddoura1, Chris Lata2, Ye Shen3,4, Brenda Chow4, Kevin Fonseca5, Jenine Leal4,6, Jan Storek1,7, Mona Shafey1,8, Andrew Daly1,8, John Conly1,4
1Department of Medicine, University of Calgary and AHS-Calgary, Calgary, Alberta, Canada; 2Division of Infectious Diseases, Department of Medicine, Cape Breton Regional Hospital, Dalhousie University, Sydney, NS, Canada; 3ProvSurv, AHS-Calgary, Calgary, Alberta, Canada; 4Infection Prevention and Control, AHS-Calgary, Calgary, Alberta, Canada; 5ProvLab, AHS-Calgary, Calgary, Alberta, Canada; 6Department of Community Health Sciences, University of Calgary, Calgary, Alberta, Canada; 7Department of Oncology, University of Calgary and AHS-Calgary, Calgary, Alberta, Canada; 8Division of Hematology, Department of Medicine, University of Calgary and AHS-Calgary, Calgary, Alberta, Canada
Objectives: Clostridioides difficile infection (CDI) is a significant complication of hematopoietic stem cell transplantation (HSCT). We sought to study the risk factors and incidence for CDI in the first 100 days after HSCT and its association with patient outcomes including acute graft-versus-host-disease (aGvHD).
Methods: We conducted a single-centre retrospective cohort study of patients receiving allogeneic and autologous HSCT (allo-HSCT/auto-HSCT) between 2011 and 2016. We excluded patients with prior transplants, cord-blood HSCT, HIV infection, and use of immunosuppressive medications. We defined CDI using symptom-based and laboratory criteria with a validated, standardized protocol. We analyzed data using chi-square, Fisher’s exact test, ANOVA, and multinomial logistic regression for aGvHD. All analyses were performed in R or SPSS V26.
Results: Of 333 allo-HSCT and 421 auto-HSCT patients, 61 (18.3%) and 71 (16.9%) patients developed CDI within 100 days post-transplant, respectively. In auto-HSCT patients, CDI was significantly more common in patients with Hodgkin’s and non-Hodgkin’s lymphoma (p = 0.037), and it tended to increase with study years (p = 0.056). In allo-HSCT, CDI was more common in patients with acute myelogenous and lymphoblastic leukemia (p = 0.091). Sex, age, cytomegalovirus mismatch, donor type, and hospital length of stay were not significantly associated with CDI for allo-HSCT. There was no significant association between CDI and all-cause mortality, but it directly contributed to death in 3 allo-HSCT and 1 auto-HSCT patient. CDI was not significantly associated with aGvHD or aGvHD severity (p = 0.49 and p = 0.25, respectively). In adjusted analyses for allo-HSCT, none of the variables were independently associated with aGvHD, though we did not address the exact date of onset of aGvHD as a variable (only within first 100 days).
Conclusions: CDI is common in patients with HSCT, andsurveillance for it should continue beyond the immediate post-transplant period. More studies are required to delineate risk factors for aGvHD, and whether CDI is associated with an increased risk.
Cheyanne Boehm1, Christopher Doig2,3, Justin Z Chen4,5, Wendy I Sligl6, Sean Bagshaw7, John Conly8
1Department of Pharmacy Services, Alberta Health Services, Calgary, Alberta, Canada; 2Department of Critical Care Medicine, Alberta Health Services, Calgary, Alberta, Canada; 3University of Calgary, Calgary, Alberta, Canada; 4Division of Infectious Diseases, Department of Medicine, University of Alberta, Edmonton, Alberta, Canada; 5University of Alberta Hospital, Alberta Health Services, Edmonton, Alberta, Canada; 6Department of Critical Care Medicine and Division of Infectious Diseases, Alberta Health Services, Edmonton, Alberta, Canada; 7Department of Critical Care Medicine, Department of Medicine, University of Alberta and Alberta Health Services, Edmonton, Alberta, Canada; 8Department of Medicine, University of Calgary and Alberta Health Services, Calgary, Alberta, Canada
West Nile virus neuroinvasive disease (WNV-NID) is associated with significant morbidity and mortality. Given that WMV-NID is clinically and microbiologically challenging to diagnose, adding procalcitonin (PCT) as a diagnostic tool may be useful. We present four cases of WNV-NID with serum PCT measurements given the dearth of data in this setting.
Serial PCT measurements (bioMérieux, St. Laurent, PQ, Canada) were examined in patients with confirmed WNV-NID. Patients were identified from within a large provincial sepsis optimization study enrolling critically ill patients with sepsis. Daily PCT levels were measured in these patients for up to 7 days. PCT values of >0.5 ng/mL were suggestive of bacterial infection.
Case one presented with a meningoencephalitis syndrome, with WNV infection confirmed on day 1. The PCT value was 0.09 ng/mL, which peaked at 0.23 ng/mL. Case two had viral meningitis and possible aspiration pneumonitis, with confirmed WNV infection on day 1. The PCT peaked at 0.34 ng/mL on admission. Case three presented with a presumed viral meningoencephalitis and a confirmed urinary tract infection. The initial PCT was 0.59 ng/mL and decreased to 0.15 ng/mL. On day 6, WNV infection was confirmed. Case four presented with a suspected pyelonephritis and pneumonia. The admission PCT was 23.18 ng/mL and decreased to 1.35 ng/mL. A meningoencephalitis syndrome later emerged and WNV infection was diagnosed on day 9.
In three patients presenting with sepsis and WNV-NID, PCT levels were low at initial presentation. The fourth case presented as a non-viral respiratory infection with declining PCT values and was subsequently diagnosed with WNV meningoencephalitis. These cases reflect low PCTs, consistent with viral etiology, unless accompanied by a documented bacterial infection. To our knowledge, these cases are the first to report serial PCT evaluations in confirmed cases of WNV-NID.
Cheryl Pei Zhen Foo1, Catherine Gayle Sutcliffe2, L Robbin Lindsay3
1Memorial University of Newfoundland and Labrador, St. John’s, Newfoundland, Canada; 2Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States; 3Zoonotic Diseases and Special Pathogens, National Microbiology Laboratory, Public Health Agency of Canada, Winnipeg, Manitoba, Canada
Objectives: The expanding geographic range of Ixodes scapularis ticks poses an emerging public health risk. We assessed the risk of exposure to Borrelia burgdorferi, Anaplasma phagocytophilum, Babesia microti and Borrelia miyamotoi in Manitobaby determining the geographic distribution of infected tick submissions to the Public Health Agency of Canada’s tick surveillance program and the minimum infection rate (MIR) of submitted ticks from 1995 to 2017.
Methods: All submitted black-legged ticks were tested by polymerase chain reaction for B. burgdorferi since 1995, A. phagocytophilum since 2006, and B. microti and B. miyamotoi since 2013. MIR of ticks acquired from Manitoba was calculated for each pathogen and analyzed by linear regression with year as the independent variable. The global positioning system coordinates of positive tick submissions were plotted by year of collection.
Results: There were 3,218 tick submissions acquired from Manitoba. The MIR per 1,000 ticks for B. burgdorferi, A. phagocytophilum, B. microti and B. miyamotoi were 141.5 (95% CI 130.8 to 152.1), 44.9 (95% CI 38.3 to 51.6), 10.9 (95% CI 6.8 to 15.2), and 15.3 (95% CI 6.7 to 23.9), respectively. Since 2005, the MIR of B. burgdorferi has increased by 10.2 per 1,000 ticks per year (95% CI 4.2 to 16.3). MIRs of A. phagocytophilum, B. microti, and B. miyamotoi had no statistically significant change over the sampling period. The geographic distribution of B. burgdorferi–positive tick submissions also expanded from 1995 to 2017.
Conclusion: B. burgdorferi and A. phagocytophilum are emerging tick-borne pathogens. The MIR for both these pathogens increased over time in black-legged ticks acquired in Manitoba. New regions in Manitoba are at risk due to the expanding geographic range of B. burgdorferi–infected ticks. Surveillance should focus on areas surrounding known at-risk regions to capture emergence of infected ticks in new areas.
Maggie OY Wong, Kieran Shah, Kennard Tan, Kevin Afra
Fraser Health, Surrey, British Columbia, Canada
Objectives: Infections from carbapenemase-producing organisms (CPO) have increased steadily in British Columbia. This study reviews characteristics of patients with CPO bacteremia and describes their treatment, outcomes, and factors associated with mortality over the past 7 years.
Methods: Patients with positive blood cultures containing organisms carrying NDM, KPC, and/or OXA48 genes were identified for retrospective chart review. Multivariate logistic regression was used to identify variables associated with 30-day mortality.
Results: Twenty-nine patients had CPO bacteremia in 9 hospitals from January 2013 to August 2019. The mean age was 69 years and average Charlson comorbidity index was 6. Seven patients (24%) had cancer. Pitt bacteremia score (PBS) was ≥4 in 30% of patients. The main source of bacteremia was urinary tract (38%), followed by intra-abdominal (24%) and catheter-related infections (14%).
The most common species identified were Klebsiella pneumoniae (52%) and Escherichia coli (34%); NDM was the predominant gene in all isolates. Prior to positive blood cultures, 59% of patients were CPO colonized.
Eighteen out of 27 patients (67%) received active empiric therapy. Two patients were deemed palliative within 48 hours of CPO bacteremia. Directed therapy included 7 monotherapy, 2 non-carbapenem combination therapy, and 18 carbapenem combination therapy. Eleven patients were admitted to the ICU. Thirteen patients (45%) died within 30 days of CPO bacteremia. PBS and source control within 5 days were included in the multivariate model; only PBS was associated with mortality.
Conclusion: This is the largest Canadian study on CPO bacteremia. Unlike other North American cities where KPC is the most common gene, NDM is the predominant gene in our geography. PBS less than 4 was associated with lower mortality. Qualitatively, malignancy and ICU admission appeared to be associated with worse outcomes.
Heather J Adam1,2, Melanie R Baxter1, Alyssa R Golden1, Philippe RS Lagacé-Wiens1,2, Andrew Walkty1,2, James A Karlowsky1,2, George G Zhanel1
1Department of Medical Microbiology and Infectious Diseases, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada; 2Shared Health, Winnipeg, Manitoba, Canada
Objectives: Lefamulin, a novel oral and IV pleuromutilin, was approved by the FDA in August 2019 for the treatment of community-acquired bacterial pneumonia (CABP) and is awaiting approval by Health Canada. This study assessed the in vitro activity of lefamulin versus Streptococcus pneumoniae (SPN) respiratory and blood isolates obtained across Canada by the CANWARD surveillance study from 2015 to 2018.
Methods: Antimicrobial susceptibility testing was performed on 314 respiratory isolates and 167 blood isolates using the Clinical and Laboratory Standards Institute (CLSI) broth microdilution method (2019). Lefamulin MICs were interpreted using FDA interpretive criteria (https://www.fda.gov/drugs/development-approval-process-drugs; ≤0.5 µg/mL susceptible) and MICs to other agents were interpreted using CLSI criteria (2019).
Results: Table A011 describes the in vitro activity of lefamulin versus various SPN phenotypes along with select comparators.
|MIC90 (µg/mL) / %S|
|All (n = 481)||0.12/100||4/74.8||0.25/99.8|
|Pen-R (n = 21)||0.12/100||>32/19.0||0.25/100|
|Cefurox-R (n = 31)||0.12/100||>32/16.1||0.25/100|
|Clari-R (n = 110)||0.25/100||>32/0||0.25/99.1|
|Doxy-R (n = 67)||0.25/100||>32/16.4||0.25/98.5|
|TMP/SMX-R (n = 39)||0.12/100||>32/43.6||0.25/100|
|MDR* (n = 10)||0.12/100||>32/0||0.5/100|
|Resp SPN (n = 314)||0.25/100||32/73.6||0.25/99.7|
|Blood SPN (n = 167)||0.12/100||4/77.2||0.25/100|
* MDR = multi-drug resistant to penicillin, clarithromycin, and TMP/SMX
Conclusion: Lefamulin demonstrated potent in vitro activity versus both respiratory and bacteremic isolates of SPN. One hundred percent of SPN were susceptible to lefamulin (MIC ≤0.5 µg/mL), including isolates resistant to penicillins, cephalosporins, clarithromycin, doxycycline, and TMP/SMX as well as MDR isolates.
Heather J Adam1,2, Melanie R Baxter1, Neil Irfan3, Sergio Borgia4, Rosemary Zvonar5, Robert Ariano6, Michel Savoie7, Philippe RS Lagacé-Wiens1,2, James A Karlowsky1, George G Zhanel1
1Department of Medical Microbiology and Infectious Diseases, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada; 2Shared Health, Winnipeg, Manitoba, Canada; 3Hamilton Health Sciences Centre, Hamilton, Ontario, Canada; 4Brampton Civic Hospital, Brampton, Ontario, Canada; 5Ottawa Hospital, Ottawa, Ontario, Canada; 6St. Boniface General Hospital, Winnipeg, Manitoba, Canada; 7Hôpital Maisonneuve-Rosemont, Montréal, Québec, Canada
Objectives: The CLEAR registry was established in 2019 to capture usage data on newly released intravenous antimicrobials in Canada, including ceftolozane-tazobactam. Infectious diseases physicians/pharmacists across the country were invited to voluntarily submit completed questionnaires detailing the antimicrobial’s use, rationale for use, and patient outcomes.
Methods: The REDCap web application, designed to manage online surveys and databases for research studies, was used to develop the CLEAR ceftolozane-tazobactam questionnaire and make it accessible to Canadian practitioners. Full survey details are available at https://is.gd/CLEARceftolozanetazobactam.
Results: As of December 2019, 11 surveys had been completed: 7 from Ontario, 2 from Québec, 1 from New Brunswick, and 1 from Manitoba. Ceftolozane-tazobactam was used to treat Pseudomonas aeruginosa in all cases. Three were for the treatment of hospital-associated pneumonia, 2 each for ventilator-associated pneumonia and bone and joint infections, and 1 each for community-associated pneumonia, complicated intra-abdominal infection, and complicated skin/skin structure infection. Ceftolozane-tazobactam was used in 10 of 11 cases because of resistance to other antibiotics; 1 case was due to failure of the previous antibiotic regimen. Ceftolozane-tazobactam was used in combination with other agents in 7 of 11 cases [intravenous only aminoglycosides (3), carbapenems (1), fluroroquinolones (2), and ceftriaxone (1)]. ntimicrobial susceptibility testing was performed for 8 of 11 cases. Treatment of >10 days was the most common therapy course (7/11). The most common dose was 2 g (1 g) Q8H (4/11) and the majority (7/11) were delivered as short infusions (15 min–1 h). Four of 11 had presumed or confirmed eradication of the pathogen and 6 of 11 had clinical resolution or improvement. No adverse events were reported for 8 of 11 cases, while 1 case reported gastrointestinal effects and 2 were listed as unknown.
Conclusion: In Canada, ceftolozane-tazobactam is used to treat a variety of infections caused by P. aeruginosa when the organism is resistant to other agents.
Katherin Badke1,2, Kaitlin Warrener1,2, Jessica Moe3,4,5, Amneet Aulakh1,2,6, Lillian Cao2, Lillian PC Chen2, Shelly ZQ Lu2, Jennifer Grant5,6,7,8
1Department of Pharmaceutical Sciences, Vancouver General Hospital, Vancouver, British Columbia, Canada; 2Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada; 3Department of Emergency Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 4Department of Emergency Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 5Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 6Antimicrobial Stewardship Programme: Innovation, Research, Education, and Safety (ASPIRES), Vancouver Coastal Health, Vancouver, British Columbia, Canada; 7Division of Infectious Disease, Vancouver General Hospital, Vancouver, British Columbia, Canada; 8Department of Medical Microbiology, Vancouver General Hospital, Vancouver, British Columbia, Canada
Objectives: Recent literature has demonstrated that atypical coverage for community-acquired pneumonia (CAP) is not necessary for all patients and does not improve CAP 90-day mortality rates in non-ICU patients. We assessed the proportion of patients who received atypical coverage for CAP in the emergency department (ED) at Vancouver General Hospital. We assessed appropriateness based on CRB-65 score.
Methods: We evaluated the use of antibiotics through retrospective chart review of 200 randomly selected immunocompetent patients who were diagnosed with CAP in the ED in 2018. Atypical coverage was defined as the use of macrolides, respiratory fluoroquinolones, or doxycycline. Atypical coverage was considered inappropriate for those with a CRB-65 ≤2; however, doxycycline only counted as inappropriate atypical coverage when combined with β-lactams. Atypical coverage was considered appropriate in patients with a CRB-65 ≥3.
Results: Mean age of patients was 70 years; 39% were female, 58% were admitted to hospital, and 4% were admitted to the ICU. Six percent of patients died. Atypical coverage was prescribed for 67% of patients in the ED. Fifty-two percent of non-admitted patients received atypical coverage while in the ED or on discharge from the ED. The initial analysis of all patients with a CRB-65 ≤2, found that 64% received atypical coverage. Secondary analysis was done on those with a CRB-65 ≤2, modifying the analysis using additional factors that could influence prescribing and increase atypical coverage. The secondary analysis added the following factors: admission to ICU, use of sepsis pre-printed order, qSOFA ≥2, radiologist report suggesting atypical infection, β-lactam allergy, and presentation to the ED while receiving β-lactam antibiotics. In the secondary analysis, 62% of patients with a CRB-65 ≤2 and no additional factors that would warrant atypical coverage received atypical coverage.
Conclusion: Over half of patients with no clear indication for atypical coverage received atypical coverage in the ED.
Calvin KF Lo1,2, Dominik Mertz3,4,5, Deborah Yamamura1, Mark Loeb1,4,5
1Hamilton Regional Laboratory Program, Hamilton, Ontario, Canada; 2Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 3Department of Medicine, McMaster University, Hamilton, Ontario, Canada; 4Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, Ontario, Canada; 5Michael G. DeGroote Institute for Infectious Diseases Research, McMaster University, Hamilton, Ontario, Canada
Objectives: Limited data exists on how new microbiology diagnostic methods (e.g., matrix assisted laser desorption ionization-time of flight [MALDI-TOF] mass spectroscopy) impact clinical practice. We assessed whether MALDI-TOF, through improving identification time, reduces time to directed antibiotic coverage compared with standard methods (e.g., VITEK 2) across two acute care hospitals in an academic centre in Ontario.
Methods: Retrospective microbiology and pharmacy data for patients with gram-negative bacteremia from January 2016 to December 2017 were collected. Blood cultures who underwent MALDI-TOF diagnostic testing were compared with those with standard identification. Differences were assessed in identification time, days to initiating directed therapy, and discontinue inappropriate antibiotics based on identification results.
Results: A total of 377 eligible cultures were analyzed with 48.5% (183/377) of cultures processed through MALDI. MALDI significantly reduced pathogen identification times from 48.91 to 34.58 hours (mean difference 14.33 h, 95% CI 11.15 to 17.51, p < 0.001).
Among 377 cultures, 228 cultures were eligible for analyzing time differences to prescribe directed antibiotic therapy based on specimen identifications. MALDI had significantly reduced time to initiate post-identification antibiotic therapy, with mean difference of 16.37 hours, 95% CI 10.05 to 22.69 (50.34 ± 21.21 h versus VITEK, 66.71 ± 27.12 h; p < 0.001). Among those, 79.7% (181 cultures) were considered appropriate therapy.
Meanwhile, 252 eligible cultures were analyzed for time to discontinue previous therapy. MALDI cultures had significantly shorter discontinuation times (58.21 h versus VITEK 68.39 h; mean difference 10.19 h; 95% CI 3.32 to 17.06, p = 0.004). Ninety-seven cultures had clear classifications for discontinuing prescriptions, with 97.9% (95/97) being appropriate.
Conclusions: MALDI significantly reduced identification times, from 48.91 to 34.58 hours (mean difference 14.33 h; 95% CI 11.15 to 17.51, p < 0.001). Significant time reductions both for initiating prescription of directed antibiotic therapy and discontinuing previous antibiotic therapy were also achieved through use of MALDI.
Margot A Rosenthal1, Guillaume Poliquin2, Andrew Walkty3, Vanessa Poliquin1
1Department of Obstetrics, Gynecology and Reproductive Sciences, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada; 2Department of Pediatrics, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada; 3Department of Medical Microbiology and Infectious Diseases, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada
Objectives: Group A Streptococcus (GAS) is well-known as a causative agent of puerperal sepsis. Little evidence is available in the literature to guide management of GAS vaginal colonization identified in the antenatal period.
Methods: Laboratory databases at two tertiary care hospitals identified patients whose third trimester rectovaginal swabs for group B Streptococcus (GBS) grew GAS between 2011 and 2018. Retrospective chart review was undertaken to describe the prenatal, intrapartum, postpartum, and neonatal course of these patients.
Results: Of 31,502 rectovaginal swabs collected during the study period, GAS was isolated from 64 (0.2%). The findings from 10 (10/64) of these patients are presented here. All patients had GAS identified on routine GBS rectovaginal swab in third trimester, prior to their presentation in labour. None of the patients received treatment at the time of diagnosis. One patient received penicillin G in labour as GAS prophylaxis, a second had a co-infection with GBS and received penicillin G for GBS prophylaxis, and a third received no intrapartum antibiotics but was discharged home on oral penicillin. One patient (1/10) who did not receive antibiotics developed puerperal sepsis and septic pelvic thrombophlebitis requiring ICU stay and prolonged admission. Her neonate developed a superficial skin infection which was positive for GAS. The remaining 6 patients received no peripartum antibiotics and went on to have uncomplicated labours without neonatal complication. Interestingly, within our group, only half of obstetric providers noted the positive GAS swab in the prenatal record.
Conclusions: Management of pregnant women with GAS identified on a rectovaginal swab appears to be variable among obstetricians. In preliminary analysis, we identified puerperal sepsis in 1 of 10 cases of GAS colonization identified in the antepartum period, which is significantly higher than the average obstetrical population.
Ariana Saatchi1, Nirma Khatri Vadlamudi1, David M Patrick1,2,3, Andrew M Morris4, Michael Silverman5, Manish Sadarangani6,7, Fawziah Marra1,3
1Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada; 2British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada; 3School of Population and Public Health, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 4Sinai Health System, University Health Network, and University of Toronto, Toronto, Ontario, Canada; 5Faculty of Medicine, University of Western Ontario, London, Ontario, Canada; 6Vaccine Evaluation Centre, British Columbia, Canada Children’s Hospital Research Institute, Vancouver, British Columbia, Canada; 7Department of Pediatrics, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada
Objective: Acute otitis media (AOM) is one of the most prevalent pediatric infections and is often associated with inappropriate antibiotic prescribing, despite being a self-limiting condition. This study sought to investigate the effects of a universal pneumococcal immunization program on disease burden and antibiotic prescribing rates in British Columbia (BC). With the introduction of the 7-valent pneumococcal conjugate vaccine (PCV7) in 2003 and the switch to the 13-valent (PCV13) in 2010, we examined rates of AOM as well as associated antibiotic use across pre- and post-vaccine eras.
Methods: In a retrospective cohort study from 2000 to 2015, we examined administrative health care databases in the outpatient setting. Change in annual incidence for both AOM and antibiotic prescribing was evaluated using the Cochran–Armitage test. Poisson regression was used to calculate incidence rate ratios (IRRs) to compare changes across each period of interest (pre-vaccine, PCV7, PCV13).
Results: From 2000 to 2015, this study identified just over 2.4 million cases of AOM, with approximately 1.4 million prescriptions of an antibiotic in the outpatient setting. In comparison to the pre-vaccine period, the introduction of PCV7 was associated with a significant decline in AOM incidence (IRR 0.83 [95% CI 0.83 to 0.84]) as well as antibiotic use (IRR 0.81 [95% CI 0.80 to 0.81]). Subsequently, the switch to PCV13 saw further reductions in AOM (IRR 0.69 [95% CI 0.69 to 0.70]) as well as associated prescribing (IRR 0.65 [95% CI 0.65 to 0.66]). By the end of the study period, rates of AOM had reduced from 42 to 29 cases per 1,000 population, with a similar decrease in antibiotic prescribing (pre-vaccine, 25 prescriptions per 1,000; post-PCV13, 16 prescriptions per 1,000).
Conclusion: The introduction of PCV into the BC infant immunization program has been associated with significant reductions in both the incidence of AOM and associated antibiotic prescribing.
Thomas Fear1,2, Alyssa R Golden1, Heather J Adam1,2, Melanie R Baxter1, Irene Martin3, Walter Demczuk3, Michael Mulvey3, James A Karlowsky1,2, George G Zhanel1
1University of Manitoba, Winnipeg, Manitoba, Canada; 2Shared Health Manitoba, Winnipeg, Manitoba, Canada; 3National Microbiology Laboratory, Winnipeg, Manitoba, Canada
Objective: The new 15-valent pneumococcal conjugate vaccine (PCV) will cover serotypes 22F and 33F. We characterized serotypes 22F and 33F causing invasive pneumococcal disease (IPD) in Canada from 2011 to 2018.
Methods: The SAVE study is a partnership between the Canadian Antimicrobial Resistance Alliance (CARA) and the National Microbiology Laboratory. From 2011 to 2018, eight provincial public health laboratories contributed S. pneumoniae isolated from sterile sites, along with patient demographic information. Serotyping and antimicrobial susceptibility testing (Clinical and Laboratory Standards Institute [CLSI] microdilution) were performed for each isolate. Whole genome sequencing was performed on select isolates. The Cochran–Armitage test was used to identify significant trends over time.
Results: MIC testing was performed on11,044 SPN isolates. The most common serotype identified was 22F with 9.3% (1,024/11,044) of isolates causing 11.0% and 10.0% of IPD in patients <5 years and >65 years, respectively. Susceptibilities for 22F were as follows: ceftriaxone (CTX) 99.8% (meningitis breakpoints [BP]), meropenem (MER) 99.7%, penicillin (PEN) 99.3% (oral BP), levofloxacin (LEV) 99.3%, SXT 98.9%, doxycycline (DOX) 98.8%, and clindamycin (CD) 97.8%. Overall, susceptibility to clarithromycin (CLR) in 22F was 68.6%, declining from 2011 (80.3%) to 2018 (52.9%, p < 0.001). Multidrug resistance (MDR) in 22F was 1.2% (12/1,024). Serotype 22F was highly clonal, where 96% of sequenced strains were related to ST433. Serotype 33F accounted for 3.8% (416/11,044) of isolates with susceptibilities to CTX 100%, LEV 100%, PEN 99.8%, MER 99.8%, DOX 89.2%, CD 87.1%, SXT 24.9%, and CLR 22.2%. MDR was seen in 10.8% (45/416) of 33F isolates. Commonly CLR– and SXT–non-susceptible 33F isolates related to ST100 increased in prevalence from 2011 (50.0%) to 2018 (84.8%, p < 0.006).
Conclusions: Serotype 22F was commonly isolated, highly clonal and generally demonstrated little antimicrobial resistance with the exception of CLR. Serotype 33F demonstrated 10.8% MDR, and there was an increasing prevalence of antimicrobial-non-susceptible clone ST100.
Sean H Ling1, Ekua Amponsah Agyemang1, Peter Anto Johnson1,2, Mikayla Lindsell1,2, William Banh2, Melody Cordoviz2, Chad Herbers2, Alexandra C McFarlane1,2, Aruna U Chandran1,2
1University of Alberta, Edmonton, Alberta, Canada; 2Infection Prevention & Control, Royal Alexandra Hospital, Edmonton, Alberta, Canada
Objective: Direct, overt observations are the gold standard for evaluating hand hygiene (HH) rates; however, the Hawthorne effect may affect the accuracy of these observations. This study compared HH rates from covert and overt direct observations by indication (HH moment), health care worker (HCW) group and clinical area.
Methods: HH reviewers were trained with a combination of online modules and in-person competency checks. Reviewers observed HCWs covertly on seven patient care units (including adult and neonatal critical care units) and the emergency department, providing no “in the moment” feedback. Overt observations occurred simultaneously in the same areas by pre-established procedures. Compliance was defined as appropriate hand washing with alcohol-based hand rub or soap and water during one of the four HH moments per the Alberta Health Services provincial guidelines. HH rates were compared using the two-tailed Fisher’s exact test; p < 0.05 was considered to be statistically significant.
Results: Observations were made during 8 weeks from July to September 2019. The covert HH rate was 31.0% (415/1,338); the overt rate was 80.4% (1,173/1,459; p < 0.05). There was also a significant difference between covert and overt HH rates by HH moment, HCW group (physicians and nurses), and clinical area. Similar trends were observed in a related study in 2018; however, the 2019 covert HH rates were also significantly lower than the 2018 covert HH rates (overall and by moment, HCW group and clinical area).
Conclusions: In 2019, there was a 49.4% overestimation of overall HH rates due to the Hawthorne effect, compared with 28.5% in 2018. This must continue to be considered not only when interpreting overtly collected HH data in these hospital areas, but also when developing new HH initiatives and when reassessing and evaluating existing initiatives attempting to address moment-, HCW-, and area-specific HH rates.
Samuel D Chorlton1, Gordon Ritchie1,2, Marc G Romney1,2, Christopher F Lowe1,2
1Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 2Division of Medical Microbiology and Virology, Providence Health Care, Vancouver, British Columbia, Canada
Objectives: Infections caused by C. diphtheriae are increasing, with more cases reported globally in 2018 than any year since 1996. Genomic investigations of diphtheria outbreaks have relied on core genome multilocus sequence typing (cgMLST) to identify potential epidemiological links between cases; however, these schemes have not been published for public use. Previously, we found that the choice of cgMLST scheme can lead to different allele distances between C. diphtheriae isolates. We investigated a novel scheme for uniform typing and outbreak investigation.
Methods: We downloaded all publicly available, complete and chromosome-level C. diphtheriae assemblies from Refseq (n = 24). An additional assembly, GCA_900312965.1, representing the type strain of the recently proposed C. diphtheriae subsp lausannense subsp nov, was also included. Plasmid sequences were removed. A cgMLST scheme was generated and alleles called with chewBBACA version 2.1.0 (2019). Loci present in 100% of isolates were retained and paralogous loci removed.
Results: We identified a core set of 1,474 loci present in all complete C. diphtheriae genomes. The conservation of core loci was evaluated in all contig- and scaffold-level C. diphtheriae assemblies from Refseq (n = 198). Ninety-five percent or more of core loci were found in 99.0% (n = 196) of all assemblies. The two assemblies missing more than 5% core loci were in the bottom 5% of all assemblies when ranked by completeness. We applied our novel scheme to 2 pairs (n = 4) of C. diphtheriae cases reported in the literature with a probable direct transmission between patients. There was a 1 and 4 allele difference between the isolates with direct transmission compared with 1,338 to 1,342 allele difference between isolates across the two studies.
Conclusion: We developed and evaluated a novel cgMLST scheme for C. diphtheriae molecular epidemiological investigations. Further analyses of epidemiologically linked cases are needed to validate the scheme.
Nirma Khatri Vadlamudi1, David M Patrick1,2, Linda Hoang1,2, Fawziah Marra1
1University of British Columbia, Vancouver, British Columbia, Canada; 2British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada
Objective: Acute otitis media (AOM) commonly afflicts young children. AOM can have bacterial or viral etiology. Prior to the implementation of pneumococcal conjugate vaccines (PCV), the bacterial pathogen Streptococcus pneumoniae was isolated in 30%–60% of culture-tested AOM cases. In 2003, British Columbia (BC), Canada implemented the 7-valent pneumococcal conjugate (PCV7) vaccine in their provincial infant immunization program for prevention of pneumococcal infections, which was replaced by the 13-valent pneumococcal conjugate (PCV13) vaccine in 2010. Several studies reported that the introduction of the PCV7 vaccine in their infant immunization programs led to a decline in AOM in children under 2 years of age. However, there is a paucity of data following the switch to the PCV13 vaccine. We studied the change in all-cause AOM rates in children aged 2 years and younger following the implementation of the PCV13 vaccine in the infant immunization program in BC.
Methods: In this retrospective cohort study, we report AOM rates using the health care administrative databases for the years 2000–2015. We assessed the change in annual incidence of AOM following the implementation of PCVs using the Cochran–Armitage test. Finally, incidence rate ratios (IRRs) comparing the PCV7 (2004–2010) and PCV13 (2011–2015) eras to the pre-vaccine era (2000–2003) were evaluated using Poisson regression.
Results: In the 16 years, 334,356 cases occurred in children aged 2 years and younger. AOM rates have been declining substantially from 227 per 1,000 in 2000 to 107 per 1,000 in 2015 (p < 0.0001). A significant decline in AOM rates is noted in the PCV7 era compared with the pre-vaccine era (IRR 0.77 [95% CI 0.77 to 0.78]). A further decline is observed in the PCV13 era compared with the pre-vaccine era (IRR 0.58 [95% CI 0.58 to 0.59]).
Conclusion: A significant reduction in all-cause AOM rates was observed in the children under 2 years of age following the implementation of PCV.
Brittany Buffone1, Jonathan Schwarz1, Jennifer Grant2, Salomeh Sharjari2, Yu-Chen Lin1,3
1Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada; 2Vancouver Coastal Health, Vancouver, British Columbia, Canada; 3Lower Mainland Pharmacy Services, Fraser Health, North Vancouver, British Columbia, Canada
Objectives: Historically, prescribers would avoid β-lactam antibiotics for patients reporting an allergy to penicillins due to concerns for cross-reactivity. Recent studies suggest that type-1 hypersensitivity cross-reactivity between β-lactam antibiotics is due to side chain similarity and not the common β-lactam ring. As a result, the prescriber alerting rules of an electronic medical record (EMR) system were adjusted to only flag prescribers when prescribing penicillins or β-lactams with similar side chains (such as cephalexin, cefadroxil, and cefoxitin) to patients with a documented allergy to penicillins. This study was conducted to assess and confirm the safety of the adjusted alerting rules, and the primary outcome is the prevalence of anaphylaxis upon β-lactam re-exposure.
Methods: Retrospective chart review was conducted on patients who, under the reformed alerting rules, received a β-lactam antibiotic post-documentation of an allergy to penicillins in their EMR from April 2018 to July 2019 at a Canadian hospital. Given the volume of eligible patients, a 25% sample was randomly selected for review from initiation of the β-lactam antibiotic up to 30 days post-exposure to determine the prevalence of anaphylaxis.
Results: Of the 325 charts reviewed, 92% (300/325) received a β-lactam antibiotic with a different side chain than penicillins (not alerted upon prescribing). Chart review of these 300 patients confirmed no reports of anaphylaxis secondary to β-lactam exposure (0%), while 2 patients developed non-anaphylactic delayed reactions (rash).
Conclusion: There were no reports of immediate life-threatening anaphylaxis under the reformed alerting rules despite 8% of patients receiving an alerted drug such as piperacillin-tazobactam. The reformed alerting rules better reflect current literature and reduce the risk of prescriber alerting fatigue without compromising patient safety. The occurrence of delayed reactions reinforced the need to monitor for these reactions upon β-lactam antibiotic prescribing.
Jonathan Schwarz1, Brittany Buffone1, Jennifer Grant2, Salomeh Shajari2, Yu-Chen Lin1,3
1Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada; 2Vancouver Coastal Health, Vancouver, British Columbia, Canada; 3Lower Mainland Pharmacy Services, Fraser Health, North Vancouver, British Columbia, Canada
Objectives: Approximately 10% of inpatients report a penicillin allergy; however, when tested only 5% have a positive penicillin skin test (PST). The PST can effectively rule out type 1 hypersensitivity reactions. Literature shows that antidepressants, antipsychotics, and benzodiazepines have varying histamine (H1) blockade and may interfere with the histamine positive control and render the test indeterminate. However, with limited data, the screening and management of these drugs remains unstandardized. This study aims to examine histamine blockades in patients’ skin-tested on psychotropics and improve screening processes and resource utilization.
Methods: We conducted a retrospective chart review of all patients receiving a PST at a Canadian hospital between April 2018 and July 2019. Antidepressant, antipsychotic, or benzodiazepine usage (drug, dose, route, and frequency) was recorded. Chart notes were used to review patient cases and assess the primary outcome of histamine positive control results.
Results: Twenty-two patients were eligible and included in our study. Among patients taking a psychotropic medication and those not taking any potentially interfering medication, 3 of 10, and 4 of 11 patients, respectively, had a negative skin prick histamine control result. However, only 1 patient (on amitriptyline 10 mg daily) proceeded to also have a negative intradermal histamine control result, which indicated medication interference. The patient was re-tested after holding amitriptyline for 2 weeks and had negative skin prick histamine control, positive intradermal histamine control, and unimpeded overall test result.
Conclusion: Our results support holding tricyclic antidepressants for 2 weeks is sufficient to avoid interference with histamine control. Continuing other antipsychotics and antidepressants does not appear to jeopardize PST results. Proceeding to the intradermal step despite negative skin prick histamine control appeared to be practical and safe. Effective medication screening can optimize the use of limited PST appointment resources.
Patrick J Kim1, David C Bulir2,1, Kathy E Luinstra1, Julia Maciejewski1, Melissa Richard-Greenblatt1, Marek Smieja1
1Research Institute, St. Joseph’s Healthcare-Hamilton, Hamilton, Ontario, Canada; 2McMaster University, Hamilton, Ontario, Canada
Objective: Sapovirus is responsible for sporadic cases and outbreaks of acute gastroenteritis around the world. Being less prevalent compared with other gastroenteritis-causing pathogens such as norovirus and rotavirus, few investigations of sapovirus have been conducted in Canadian communities. This study’s aim was to describe the epidemiological and clinical characteristics of sapovirus cases in Hamilton, Ontario, Canada.
Methods: Three-hundred-seventy-five stool samples from patients of all ages were collected from four Hamilton hospitals from April 2017 to March 2018. A novel RT-PCR probe assay was developed to detect sapovirus. Samples positive for sapovirus were sequenced for genotyping. Sapovirus cases were compared to other cases of acute gastroenteritis based on clinical charts.
Results: Sapovirus was detected in 1.6% (6/375) of samples. All sapovirus-positive samples were found in fecal samples of children aged ≤5 years and were community-acquired infections. Among patients aged ≤5 years, the prevalence was 5.6% (6/108). Sapovirus was found only in the spring and summer months. When compared with norovirus, sapovirus was significantly less likely to be hospital-acquired (OR 0.097; p < 0.01). Genotypes identified through sequencing were I.1, II.1, II.3, and II.4.
Conclusion: Our results show that sapovirus was present in young children in the Hamilton community from 2017 to 2018. Sapovirus will be considered for future implementation in routine hospital testing once successfully multiplexed with existing norovirus testing protocols.
Nicky Rai1, Jennifer Grant2, Suzanne C Malfair1,3, Salomeh Shajari2, Yu-Chen Lin1,3
1Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada; 2Vancouver Coastal Health, Vancouver, British Columbia, Canada; 3Lower Mainland Pharmacy Services, Fraser Health, North Vancouver, British Columbia, Canada
Objective: Changes in electronic medical record systems can impact the duration of antimicrobial orders. A new electronic medical record system was implemented in a community hospital in April 2018 with Soft Stop and reminders. Soft Stop orders would remain active beyond the Soft Stop date. In October 2019, Soft Stop was replaced with Hard Stop. To assess the impact of this change, data from the electronic system was analyzed and compared.
Methods: An antimicrobial point prevalence survey was conducted in a community hospital from September to November 2019. The survey included all inpatients admitted at 8:00 am on the survey date receiving an anti-infective within the 24 hours prior to the survey date. Duration of antimicrobial orders were recorded to include: duration to date of the active order, prescribed duration of the active order, duration of therapy for the same indication, and duration to IV to PO step down. The survey was repeated on selected wards after implementation of Hard Stop.
Results: From 218 patients surveyed, 68 patients (31.2%) met the inclusion criteria with 88 antimicrobial prescriptions. Comparing 2018 to 2019, the survey identified a decrease in average duration of the antimicrobial order on the survey date from 4.9 days to 4.4 days, and 2.8 days with Hard Stop. Duration of treatment for the same indication was 6 days with Soft Stop and decreased to 3.7 days with Hard Stop. Duration to IV to PO stepdown was 3.1 days and 3.2 days with Soft Stop and Hard Stop respectively.
Conclusions: These results suggest that the implementation of Hard Stop for antimicrobials led to a 1.6-day decrease in antimicrobial prescription duration compared with Soft Stop. Opportunities remain for further streamlining of the electronic system to improve antimicrobial stewardship initiatives. Further developments may be assessed with additional point prevalence surveys.
Barbara LS Gomes, Fernanda S Fendler, Marcelo AL Cotta, Maria Eduarda PM Pinto, Nicolle M Almeida, Gabrielle C Nonato, Bráulio RGM Couto, Carlos Starling
Centro Universitário De Belo Horizonte, Belo Horizonte, Brazil
Background: Although low, the risks of contamination of the surgical site during or after performing clean surgeries do exist and require attention, given the vulnerability to which patients may be submitted and the burden that these consequences of infection generate to the hospital system.
Objective: The objective was to analyze the incidence and risk factors associated with the development of surgical site infection (SSI) in patients undergoing clean surgeries.
Materials and Methods: A retrospective, multicentre study was based on revised data from medical records of clean surgery patients between July 2016 and June 2018. Patients were from 7 hospitals in Belo Horizonte, a city of 3,000,000 people in Brazil. Data were gathered by standardized methods defined by the National Healthcare Safety Network (NHSN)/CDC procedure-associated protocols for routine SSI surveillance. Twenty-six preoperative and operative categorical and continuous variables were evaluated by univariate and multivariate analysis (logistic regression). The outcome variables were SSI, hospital death, and hospital length of stay. Variables were analyzed using Epi Info and applying statistical two-tailed test hypothesis with a significance level of 5%.
Results: A sample of 45,990 patients was submitted and analyzed (SSI risk = 1.4%). The mean hospital length of stay for noninfected patients was 4 (SD 9) days. The mean hospital length of stay for infected patients was 15 (SD 22) days (p < 0.001). The mortality rate in patients without infection was 1.2%, while hospital death of infected patients was 9.4% (p < 0.001). Main risk factors for SSI were general anesthesia (SSI = 1.9%, relative risk [RR] 1.4, p < 0.001); preoperative hospital length of stay more than 4 days (SSI = 3.8%, RR 3.2, p < 0.001); duration of procedure higher than 2 hours (SSI = 2.0%, RR 1.6, p < 0.001); ASA score >2 (SSI = 3.1%, RR 2.8, p < 0.001); emergency surgery (SSI = 2.8%, RR 2.0, p = 0.001); prosthesis (SSI = 2.1%, RR 1.8, p < 001); age >70 years (SSI = 2.1%, RR 1.7, p < 0.001). Video laparoscopy is a protective factor (SSI = 0.9%, RR 0.6, p = 0.008).
Conclusions: We identified patients at high risk of SSI after clean surgery.
Yahya Shabi1, Audra Russell-Tattrie2, Amrita Bharat3, David Haldane2, Glenn Patriquin2
1Dalhousie University, Halifax, Nova Scotia, Canada; 2Dalhousie University, Halifax, Nova Scotia, Canada; 3National Microbiology Laboratory, Winnipeg, Manitoba, Canada
Objectives: Clinical isolates of Candida species typically undergo susceptibility testing (to azoles and echinocandins). Multiple echinocandins are commercially available for in vitro susceptibility testing using an agar gradient diffusion method including micafungin, caspofungin, and anidulafungin. After institutional formulary (and therefore susceptibility testing) was changed from micafungin to caspofungin, we experienced an abrupt increase in echinocandin resistance among clinical isolates of C. glabrata. In this study, we determined and quantified discrepancies in echinocandin testing among three agents and found that caspofungin testing results had overestimated echinocandin resistance.
Methods: We queried the laboratory information system for caspofungin-resistant C. glabrata isolates and retrospectively tested these isolates for sensitivity to micafungin and anidulafungin using an agar diffusion method. We compared these results with those generated prior to the institutional change to caspofungin, and to those generated after adopting European Committee on Antimicrobial Susceptibility Testing (EUCAST) testing guidelines (inferring caspofungin sensitivity from a combination of results from micafungin and anidulafungin). The study period was a total of 44 weeks. Breakpoints were determined by Clinical and Laboratory Standards Institute (CLSI) guidelines. Descriptive statistics were used.
Results: A total of 44 C. glabrata isolates from various body sites were tested (mostly from sterile sites). The baseline echinocandin non-susceptibility rate prior to the change to caspofungin testing was 5.3% (n = 19), which increased to 91.7% (n = 12) coinciding with the change to caspofungin testing (p < 0.0001). Institution of EUCAST guidelines resulted in a return to baseline echinocandin non-susceptibility rates of 7.7% (n = 13). Retrospective testing of micafungin/anidulafungin-sensitive isolates with caspofungin confirmed erroneous detection of echinocandin resistance based on caspofungin MICs.
Conclusion: This study quantifies and emphasizes the unreliability of caspofungin testing by agar diffusion for determining echinocandin resistance in C. glabrata, which may affect patient management and antifungal choice.
Evan Wheeler1, Gerald McDonald2, Peter Daley1
1Memorial University, St. John’s, Newfoundland, Canada; 2Eastern Health, St. John’s, Newfoundland, Canada
Objectives: Community hospitals may not have internal antimicrobial stewardship programs, yet they may benefit from remote audit and feedback. Appropriateness of antibiotic prescription in community hospitals is unknown and may be different from urban hospitals.
Methods: The National Antimicrobial Prescribing Survey tool was used to retrospectively review all systemic antimicrobials prescribed on April 24, 2019 in three community hospitals for indication, documentation of infection, review date, and appropriateness. Appropriateness was assessed by an infectious disease physician.
Results: We received 101 prescriptions and 21 were excluded (topical antibiotics, antivirals). Antibiotic treatment prevalence was 58/120 (48.3%) of occupied beds. Overall appropriateness was 37/80 (46.3%). The most prescribed antibiotics were ceftriaxone (17/80, 21.3%, 47.1% appropriate), piperacillin/tazobactam (10/80, 12.5%, 10.0% appropriate), and moxifloxacin (9/80, 11.3%, 0% appropriate). We also found that 36/80 (45.0%) of prescriptions were for respiratory tract infections, 50/80 (62.5%) prescriptions were intravenous, and 50/80 (62.5%) had an indication documented, and 71/80 (88.8%) had a stop/review date. As well, 20/80 (25.0%) prescriptions were considered optimal, 21/80 (26.3%) were suboptimal, 17/80 (21.3%) were adequate, and 22/80 (27.5%) were inadequate.
Conclusions: We observed a high treatment prevalence and low appropriateness among common drugs. Appropriateness was lower compared with an audit performed in urban hospitals. The stewardship priorities in community hospitals are piperacillin/tazobactam and moxifloxacin.
Kieran Shah1,2, Maggie OY Wong1, Kevin Afra1, Kennard Tan1
1Fraser Health, Surrey, British Columbia, Canada; 2Lower Mainland Pharmacy Services, Vancouver, British Columbia, Canada
Objective: Infections from carbapenemase-producing organisms (CPO) have been steadily increasing in British Columbia. The antimicrobial susceptibility profile from clinical isolates (i.e., antibiogram) has the potential to reduce mortality by reducing time to empirically active antimicrobials. This study describes CPO from clinical isolates and assesses their susceptibilities through a multi-year cumulative antibiogram.
Methods: We analyzed antimicrobial susceptibility data of CPO isolated from clinical specimens (i.e., sputum, urine, blood and bodily fluids) from January 2013 to December 2018. Only CPO confirmed to contain NDM, KPC and/or OXA48 genes were included. An enhanced antibiogram based on modified Clinical & Laboratory Standards Institute (CLSI) M34-4 criteria was generated. Data over multiple years were combined, and only the first isolate of a given species per patient was included. The antibiogram included additional parameters (i.e., CPO gene, year, specimen type, facility) to permit supplemental analysis.
Results: We included 144 CPO isolates in the antibiogram. The most common species were Klebsiella pneumoniae (45%), Escherichia coli (25%), and Enterobacter cloacae (13%). We found that 62% of isolates were from urine, and 76% of isolates had the NDM gene. The antibiotic profile of all CPO isolates found that the most active agents were colistin (94%), fosfomycin (71%), tigecycline (49%), and amikacin (41%). Compared with K. pneumoniae, E. coli was more susceptible to fosfomycin (96% versus 48%) and tigecycline (92% versus 30%). Isolates with KPC gene were more susceptible to amikacin (94%) than colistin (86%).
Conclusion: Within our population, the majority of CPO carry the NDM gene and are susceptible to colistin. This antibiotic should be considered as part of the empiric treatment for patients with severe infections with CPO. However, if the pathogen carries the KPC gene, amikacin should be considered instead of colistin, considering the higher toxicity profile of colistin.
Nirma Khatri Vadlamudi1, David M Patrick1,2, Linda Hoang1,2, Fawziah Marra1
1University of British Columbia, Vancouver, British Columbia, Canada; 2British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada
Objectives: Acute sinusitis is a common upper respiratory tract infection with viral or bacterial etiology. Streptococcus pneumoniae, a bacterial pathogen, is isolated in 30% of the cases. While few studies reported impact of the 7-valent pneumococcal conjugate (PCV7) vaccine on acute sinusitis rates, the impact of 13-valent pneumococcal conjugate (PCV13) vaccine is unknown. In this retrospective cohort study, we assess the change in acute sinusitis rates in the population following the implementation of the PCV13 vaccine in the infant immunization program in British Columbia, Canada.
Methods: Using the health care administrative databases for the years 2000–2015, acute sinusitis cases in all ages were identified with ICD9/10 codes: 461/J01 and the change over time in the annual incidence of acute sinusitis was evaluated using the Cochran–Armitage test. Incidence rate ratios (IRRs) were compiled comparing the PCV7 (2004–2010) and PCV13 (2011–2015) periods to the baseline (2000–2003).
Results: A total of 1,945,672 cases occurred in the study duration of 16 years. A significant reduction in incidence of acute sinusitis was seen between 2000 (27 per 1,000) and 2015 (25 per 1,000) (p < 0.0001). A significant decrease was noted in all children in the PCV7 era compared with the baseline, especially those aged ≤2 years (IRR 0.89 [95% CI 0.86 to 0.91]). In contrast, a slight increase was seen among adults, leading to an increase in overall rates in the PCV7 era compared with the pre-vaccine era (IRR 1.07 [95% CI 1.06 to 1.07]). In the PCV13 era, a small decline in the overall acute sinusitis rates was observed compared with the pre-vaccine era (IRR 0.98 [95% CI 0.97 to 0.98]); this related to further decreases in rates in children but a 3% reduction was seen in adults aged 18–49 years.
Conclusions: A small reduction is noted in acute sinusitis rates in all ages following the implementation of PCVs in the infant immunization program.
Nirma Khatri Vadlamudi1, David M Patrick1,2, Linda Hoang1,2, Fawziah Marra1
1University of British Columbia, Vancouver, British Columbia, Canada; 2British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada
Objective: Acute laryngitis and pharyngitis are common upper respiratory tract infections with mainly viral etiology, but in some instances can be caused by bacteria, such as Streptococcus pneumoniae. Pneumococcal conjugate vaccines (PCV) target S. pneumoniae but theirimpact on acute laryngitis and pharyngitis rates is unknown.
Methods: We report acute laryngitis and pharyngitis rates using the physician billing and hospital discharge databases for the years 2000–2015. We evaluated the changes in annual incidence using the Cochran–Armitage test. Then, incidence rate ratios (IRRs) were evaluated comparing the PCV7 (2004–2010) and PCV13 (2011–2015) eras to the pre-vaccine era (2000–2003), using Poisson regression.
Results: During the 16 years, 370,293 cases of acute laryngitis and 2,937,948 cases of acute pharyngitis were reported. Compared with the pre-vaccine era, acute laryngitis rates have significantly declined in the PCV7 era (IRR 0.84 [95% CI 0.84 to 0.85]) with further declines in the PCV13 era (IRR 0.76 [95% CI 0.75 to 0.76]). Acute pharyngitis rates have also significantly declined in the PCV7 era (IRR 0.89 [95% CI 0.89 to 0.89]) with additional decline in the PCV13 era (IRR 0.70 [95% CI 0.69 to 0.70]).
Conclusion: A significant reduction in acute laryngitis and pharyngitis have occurred in both the PCV7 and PCV13 eras at the population level.
Lisa Li1,2, Diana Whellams2,3, Marthe K Charles1,2, Jennifer Grant1,2
1Division of Medical Microbiology and Infection Prevention and Control, Vancouver General Hospital, Vancouver, British Columbia, Canada; 2Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 3LifeLabs, Surrey, British Columbia, Canada
Objectives: Discrepancies in antibiogram data between a community and a hospital laboratory in Lower Mainland British Columbia prompted a collaboration to examine susceptibility testing (AST) results obtained by different testing methods for isolates of Streptococcus pneumoniae for tetracycline and β-lactam antibiotics.
Methods: We tested 89 clinical isolates of S. pneumoniae collected at both sites by ETEST (bioMérieux, St. Laurent, PQ, Canada), disk diffusion (KB), and broth microdilution (BMD) using the Sensititre™ Streptococcus STP6F panel (penicillin, ceftriaxone and tetracycline) and the Sensititre™ Myco RAPMYCO panel (doxycycline) (ThermoFisher, Burnaby, BC, Canada). Susceptibility rates (overall, by anatomic site, and by location) and categorical agreement (CA) were calculated using Clinical and Laboratory Standards Institute (CLSI) breakpoints.
Results: By BMD, susceptibility rates were as follows: tetracycline 77%, doxycycline 75%, penicillin IV non-meningitis 92%, meningitis 78%, and ceftriaxone 95% (both meningitis and non-meningitis breakpoints). For all antibiotics, isolates from sterile sites were more susceptible than those from non-sterile sites. Testing by KB for doxycycline showed a decrease of >9% in susceptibility compared with testing by other methods or by tetracycline KB. For penicillin and ceftriaxone, CA between ETEST and BMD ranged from 87% to 99%, but with high very major error rates in some cases; however, all sterile site isolates tested susceptible to ceftriaxone by both ETEST and BMD.
Conclusions: Sterile site isolates of S. pneumoniae were more susceptible than those from non-sterile sites, which could be due to a fitness cost associated with resistance genes. Reassuringly, invasive isolates remained susceptible to ceftriaxone, the empiric treatment in suspected meningitis. Percent susceptibility varied by testing method, with decreased susceptibility for doxycycline KB compared with other methods, and increased susceptibility for penicillin and ceftriaxone ETESTs compared with BMD, which could have an impact on antibiotic stewardship.
Nicole L Veltri1, Nicole Le Saux1, Prameet M Sheth1,2
1Queens University, Kingston, Ontario, Canada; 2Kingston Health Sciences Centre, Kingston, Ontario, Canada
Objectives: We present the fourth reported case of Mucor circinelloides fungemia, and the first reported case to our knowledge of transient Mucor circinelloides fungemia in a patient with active intravenous drug use (IVDU).
Case Report: A 48-year-old male with a history of IVDU was admitted to the intensive care unit for acute agitation and confusion requiring intubation and vasopressor support. He recently completed prednisone and antibiotics for a chronic obstructive pulmonary disease (COPD) exacerbation and also suffered a recent nasal septal fracture. He was febrile, and initial bloodwork showed leukocytosis. Computed tomography (CT) of the head was suggestive of sinusitis. He was treated with broad-spectrum antibiotics but recovered rapidly with extubation to room air within 24 hours of presentation. Diagnosis was thought to be sympathomimetic toxidrome and opioid withdrawal.
Two of 2 aerobic peripheral blood cultures drawn on admission flagged positive 45 hours after collection and direct Gram stain showed septate fungal elements. M. circinelloides was identified by the Public Health Ontario Laboratory Mycology section. The patient was treated with liposomal amphotericin B for 4 days. One subsequent set of blood cultures grew M. circinelloides in 1 of 2 aerobic bottles. Work-up including echocardiogram, chest radiograph, ophthalmalogic examination, and anterior rhinoscopy revealed no foci of infection. 0/16 further blood cultures grew M. circinelloides. The patient remained well and was discharged home with no complications. The portal of entry was thought to be IVDU versus nasal septal fracture in a colonized patient.
Conclusions: Despite having positive blood cultures, this patient did not have invasive mucormycosis, but transient fungemia in the setting of mild immune compromise due to prednisone. Infection with this organism and growth in blood cultures is rare but typically seen in those with profound immune compromise. Clinicians should treat aggressively until invasive infection is ruled out, even in patients without a typical risk factor profile.
Daniel Doyle1, Gerald McDonald2, Peter Daley1
1Memorial University, St. John’s, Newfoundland, Canada; 2Eastern Health, St. John’s, Newfoundland, Canada
Objectives: The Spectrum™ app provides antibiotic decision support based on local resistance epidemiology and treatment recommendations. We determined the impact of regional implementation on inpatient antimicrobial appropriateness.
Methods: We performed two 1-day hospital-wide point prevalence surveys using the National Antimicrobial Prescribing Survey tool, using two and four appropriateness categories, in two urban teaching hospitals, before (June 25, 2018) and 6 months after (June 25, 2019) app dissemination to prescribers. All inpatient prescriptions for systemic antimicrobials were included. Appropriateness assessments were performed by an internal medicine resident and reviewed by an infectious disease physician. Two-sided Pearson chi-square was used to compare pre- and post-survey categories.
Results: The app was accessed 20,016 times between February 1 and June 24, 2019. The pre-survey included 184 (8/184 [4.3%] were not assessable for appropriateness), and the post-survey included 197 prescriptions (5/197 [2.54%] were not assessable for appropriateness). Antibiotic treatment prevalence was 131/420 (31.2%) occupied beds pre-, and 139/429 (32.4%) occupied beds post- (+1.2%, p = 0.70). Overall appropriateness was 97/176 (55.1%) in the pre-survey and 126/192 (65.6%) in the post-survey (+10.5%, p = 0.051). Each appropriateness category (optimal, adequate, suboptimal, inadequate) improved but none met statistical significance. The most- prescribed antibiotic was piperacillin-tazobactam (27/184 [14.6%], 48.1% appropriate pre-; 37/197 [18.8%], 41.9% appropriate post-) (−6.2%, p = 0.83).
Conclusion: We observed a trend toward improvement in overall appropriateness following Spectrum™ implementation; however, piperacillin/tazobactam appropriateness worsened.
Claire Pratt1, Zahra Rehan1, Lydia Xing1, Brenda Fillier2, Laura Gilbert2, Peter Daley1
1Memorial University, St. John’s, Newfoundland, Canada; 2Eastern Health, St. John’s, Newfoundland, Canada
Objectives: Inappropriate antibiotic treatment for CA-ASB is common among inpatients. We conducted a randomized, unblinded superiority trial of modified reporting of positive urine cultures with outcomes of treatment, adverse events, and mortality.
Methods: Consecutive positive urine cultures collected from catheterized inpatients from two teaching hospitals were randomized in the laboratory between standard (identification and susceptibility) and modified (call for full report) culture reports between November 2018 and June 2019. Exclusion criteria were pregnancy, current antibiotic treatment, ICU or urology admission, or neutropenia. Patients were followed for 7 days after the report. Catheter-associated urinary tract infection (CA-UTI) and CA-ASB were diagnosed using IDSA criteria by prospective chart review.
Results: We considered 543 positive urine cultures; 443 were excluded, 100 were randomized and included in intention-to-treat (ITT) analysis, and 90 were included in per-protocol (PP) analysis. We diagnosed 75/100 (75%) patients with CA-ASB and 25/100 (25%) with CA-UTI. We treated 45/75 (60%) patients with CA-ASB and 25/25 (100%) patients with CA-UTI. The lab was called to report the identification and susceptibility in 20/54 (37%) of modified reports. Appropriate treatment (untreated CA-ASB + treated CA-UTI) was higher in the modified report arm than in the standard report arm, at 31/54 (57.4%) versus 23/46 (50.0%)(+7.4%, p = 0.45, RR 1.15). Untreated CA-ASB was higher in the modified report arm, at 19/42 (45%) versus 11/33 (33%)(+12%, p = 0.30, RR 1.36). PP analysis had similar results. There were 4/54 (7.4%) deaths and 21/54 (38.9%) adverse events in the modified arm, and 3/46 (6.5%) deaths and 24/46 (52.2%) adverse events in the standard arm.
Conclusions: Collection of urine culture in patients not meeting criteria for CA-UTI is common. Treatment of CA-ASB is common. Modified reporting reduced the treatment of CA-ASB, but not significantly. Modified reporting was safe.
Zahra Rehan1, Claire Pratt1, Kim Babb2, Brenda Fillier2, Laura Gilbert2, Peter Daley1
Objectives: We conducted a prospective, randomized, unblinded superiority trial of the safety and efficacy of modified reporting of positive urine cultures in long-term care facilities (LTCFs).
Methods: Consecutive positive urine cultures collected from non-catheterized patients admitted to eight LTCFs were randomized equally in the microbiology laboratory between standard (identification and susceptibility) or modified (without identification and susceptibility) report, between November 2018 and June 2019. Exclusion criteria included current antibiotic treatment, neutropenia, or transfer to acute care. Patients were followed for 30 days. The diagnosis of urinary tract infection (UTI) or asymptomatic bacteriuria (ASB) were made using published criteria.
Results: We considered 169 positive urine cultures; 69 were excluded, 100 were randomized and included in intention-to-treat analysis, and 96 were included in per-protocol analysis. We diagnosed 62/100 (62%) patients with ASB (41/62 [66%] treated) and 38/100 (38%) with UTI (35/38 [92%] treated). The lab was called to report the identification and susceptibility in 30/51 (59%) modified reports. The rate of appropriate treatment (untreated ASB + treated UTI) was higher in the modified report arm, at 31/51 (61%) versus 25/49 (51%)(+10%, p = 0.33, RR 1.2). Untreated ASB was higher in the modified report arm, at 13/21 (62%) versus 8/41 (20%)(+42%, p = 0.25, RR 1.5). There were 2/51 (3.9%) deaths and 15/51 (29%) adverse events in the modified arm, and no deaths (p = 0.16) and 11/49 (22%) adverse events (p = 0.43) in the standard arm. Three patients with untreated UTI survived.
Conclusions: Treatment of ASB is common. Modified reporting reduced treatment of ASB, but not significantly. A high percentage of doctors called to receive standard reports, causing crossover to the standard report arm. One death in the modified arm was treated ASB, and 1 death in the modified arm was untreated ASB. Modified reporting may not be suitable for LTCF implementation.
Matthew Magyar1,2,3, Allyson Shephard1,4, Pat Bedard1,4, Ken Tang5, Gyaandeo Maharajh1,3,6, Nisha Thampi1,3,4,7
1Children’s Hospital of Eastern Ontario, Ottawa, Ontario, Canada; 2Department of Pediatrics, Ottawa, Ontario, Canada; 3University of Ottawa, Ottawa, Ontario, Canada; 4Infection Prevention and Control Program, Ottawa, Ontario, Canada; 5Clinical Research Unit, Children’s Hospital of Eastern Ontario Research Institute, Ottawa, Ontario, Canada; 6Department of Surgery, Division of Cardiovascular Surgery, Ottawa, Ontario, Canada; 7Department of Pediatrics, Division of Infectious Diseases, Ottawa, Ontario, Canada
Objectives: Among pediatric patients, surgical site infections (SSIs) following open heart surgery with cardiopulmonary bypass (CPB) are associated with significant morbidity and mortality. At our pediatric hospital, an epidemiologic review was conducted to describe an increase in SSI incidence and identify risk factors and areas of practice variation to inform improvement initiatives.
Methods: We conducted a retrospective cohort study of all children undergoing CPB from January 2016 to December 2019. Cases were identified through routine surveillance using National Healthcare Safety Network definitions. Risk factors were identified through a common cause analysis with stakeholders across the preoperative, intraoperative, and postoperative care continuum. The rate of SSIs and 95% confidence interval were estimated, and univariate logistic regressions were fitted to estimate unadjusted odds ratios (OR) for the association between the anticipated risk factors and developing an SSI.
Results: A total of 195 patients underwent CPB surgery. Eighteen patients (9.2%) developed a cardiac SSI, with the highest rates occurring in 2018 and 2019 (7/53 [13.2%] and 6/45 [13.3%]). Twelve cases were organ-space SSIs (66%). There were no predominant pathogens; 5/18 cases were associated with methicillin-susceptible Staphylococcus aureus. Risk factors—including age at surgery (OR 0.843 [0.698, 1.017]), sex (OR 1.039 [0.391, 2.757]), preoperative length of stay (OR 1.001 [0.975, 1.027]), risk adjustment for congenital heart surgery (RACHS) score (OR 1.353 [0.698, 2.623]), and total CPB duration (OR 1.39 [0.963, 2.005])—did not demonstrate increased risk of SSI. Each additional day of peritoneal dialysis during the first 3 days post-operatively was associated with a 45% increased risk of SSI (OR 1.45 [1.037, 2.027]). There was variability in postoperative wound documentation, occurring in only 66% of all patients.
Conclusions: No unique organism or process explained the increased SSI incidence at our institution. Determining areas of practice variability informed the development of a sternal wound infection prevention bundle.
Gerald McDonald1, Jennifer Phillips1, Tammy Benteau2, Peter Daley3
1Eastern Health, St. John’s, Newfoundland, Canada; 2Choosing Wisely Newfoundland and Labrador, St. John’s, Newfoundland, Canada; 3Memorial University, St. John’s, Newfoundland, Canada
Objectives: The Spectrum™ app provides antibiotic decision support based on local resistance epidemiology and treatment recommendations. We determined the impact of regional implementation on antimicrobial stewardship indicators.
Methods: The app was disseminated on February 1, 2019. Inpatient antimicrobial use (AMU) data at two urban hospitals was collected monthly using PYXIS™ drug dispensing system and reported in defined daily doses (DDD). Clostridiodes difficile infection (CDI) rate, including hospital and community-acquired cases, was obtained from monthly population-based health region surveillance and reported in cases/100,000 inhabitants. CDI testing was performed using multiplex polymerase chain reaction (PCR). Trends were analyzed using linear regression.
Results: The app was accessed 20,016 times between February 1 and June 24, 2019, with a mean of 598 unique monthly active users (range 214–826), including 25% physicians, 16% residents, 14% pharmacists, and 8% nurse practitioners. AMU declined from 582.6 DDD/1,000 bed days in January 2019 (baseline) to a mean of 530.9 DDD/1,000 bed days in June–December 2019 (follow-up period) (−51.7 DDD, −9.7%) (p = 0.019) (β during follow-up = −0.50). The most commonly used antibiotic was piperacillin/tazobactam (56.2 DDD/1,000 bed days in January to mean 54.0 DDD/1,000 bed days during follow-up (−2.2 DDD, −4.1%) (p = 0.84). CDI incidence data using a common lab test was available starting January 2019, and declined from 6.30/100,000 inhabitants in January 2019 to 3.15/100,000 inhabitants in July 2019 (−3.15/100,000, −50%; reduction of 10 CDI cases during the trial period, β [January–July 2019] = −0.41, β [January–November 2019] = −0.29).
Conclusions: Spectrum use was correlated with improvements in antimicrobial stewardship indicators in two hospitals; however, unmeasured confounders may explain these trends.
Peter Daley1, Robert Wilson2, Brendan Barrett1, Patrick Parfrey1, Asghar Mohammadi2
1Memorial University, St. John’s, Newfoundland, Canada; 2Choosing Wisely Newfoundland and Labrador, St. John’s, Newfoundland, Canada
Objectives: Newfoundland and Labrador (NL) has the highest outpatient antimicrobial use (AMU) rate in Canada (970 prescriptions/1,000 inhabitants in 2017) and AMU rate is increasing, based on Canadian Antimicrobial Resistance Surveillance System 2018 estimates using selected pharmacies. We measured outpatient AMU in NL using provincial pharmacy network data.
Methods: Data on prescriptions for oral antimicrobials given to outpatients were provided by the NL Centre for Health Information. J01 class antimicrobials were included, with the addition of oral metronidazole, oral vancomycin, and oral nystatin. Indications were not available.
Results: The pharmacy network represents 100% of pharmacies in NL. Between July 1, 2017 and June 30, 2019, 912,435 prescriptions were written, representing 2,841 unique prescribers (range 1–4,652 prescriptions/prescriber/year). We excluded 70,026 non-oral prescriptions (7.7%). The rate of prescriptions/1,000 inhabitants decreased from 823 in 2017–2018 to 786 in 2018–2019 (−37/1,000, p = 0.037). The annual rate among those ≥65 years was 897/1,000 inhabitants, and among those aged 0–9 years it was 1,000/1,000. Family practitioners gave 686,820 prescriptions (82%), and 556,585 prescriptions (61%) were written for females. The most commonly prescribed antimicrobials were amoxicillin (242,986 prescriptions, 29%), azithromycin (106,266, 13%), and ciprofloxacin (79,537 prescriptions, 9%).
Conclusions: Our survey representing all provincial pharmacies demonstrates that NL has a lower AMU rate than reported, and AMU rate is significantly decreasing. AMU rate among seniors is below the national rate, but prescription rate among children is almost double the national rate. Ciprofloxacin use rate is above the national average.
Rasha Sarhan1, Evan Wilson2
1Brockville General Hospital, Brockville, Ontario, Canada; 2Division of Infectious Diseases, Queen’s University, Kingston, Ontario, Canada
Objectives: Small community hospitals often have limited staffing and infrastructure resources to dedicate toward antimicrobial stewardship (AS). We describe the implementation and results of a pharmacist-led (0.3 FTE) AS program over a 3-year period in a 150-bed community hospital using multiple AS interventions.
Methods: Initial assessment of antibiotic use was performed in summer 2016. The overuse of piperacillin-tazobactam (TZP) and fluoroquinolones (FQ) was identified as a priority area for intervention. The AS program utilizes multiple strategies: local guidelines for empiric antibiotic prescribing; specific clinical recommendations for fluoroquinolones, vancomycin, and piperacillin-tazobactam, including alternate agents; publishing of an annual local antibiogram; and IV to PO therapeutic interchange. Prospective audit and feedback are also performed on a limited basis. Infectious diseases consultation is available remotely.
Results: In the 3 years since starting the AS program, FQ use has decreased 68% (25 versus 74 DDD/1,000 patient days) and TZP use has decreased 46% (17.8 versus 33 DDD/1,000 patient days). When FQs were prescribed, a 12% decrease in parenteral administration was observed. There has been a demonstrated trend toward narrow-spectrum antibiotic use with first- and second-generation cephalosporins representing 47% of all antimicrobials prescribed compared with 25% at baseline.
Conclusions: We demonstrate that an antimicrobial stewardship program can be successfully implemented in a community hospital with a significant impact in antimicrobial prescribing. Wider adoption of ASPs in community hospitals is essential to minimizing antimicrobial resistance and improving patient safety.
Camille Pelletier Vernooy1,2, Philippe Morency-Potvin1,2,3, Anita Ang1,2,3, Annie Routhier1,2, Sandra Chapados1,2, Pierre-Louis Desaulniers1,2
1Université de Montréal, Montréal, Québec, Canada; 2Centre hospitalier de l’Université de Montréal, Montréal, Québec, Canada; 3Centre de recherche du CHUM, Montréal, Québec, Canada
Objectives: Antimicrobial resistance (AR) is one of the most critical threats to global health. One of its root causes, misuse of antibiotics, can stem from prescribers’ preconceived ideas, differing attitudes, and lack of knowledge. Canadian data on this subject is scarce. This study aimed to understand the culture of antimicrobial prescribing in order to optimize strategies targeting prescribers in the local antimicrobial stewardship program (ASP).
Methods: An anonymous online survey was distributed to antimicrobial prescribers at a 772-bed acute care teaching hospital to conduct a prospective cross-sectional study.
Results: The survey was completed by 240 respondents (95% CI MOE: 5.47%). Participation among attending physicians, residents, and specialized nurse practitioners was 16%, 37%, and 29%, respectively. All agreed that AR is a significant challenge in Canada. However, only 46% of respondents believed that antibiotics are misused locally. Most (93%) agreed that ASPs can decrease AR. Several knowledge gaps were identified through clinical questions; for example, respondents failed to identify treatment indications for asymptomatic bacteriuria 25% of the time, 61% chose an unnecessarily broad antibiotic when presented a susceptibility report from a common clinical situation, and only 28% identified an appropriate length of therapy. Prescribers’ confidence did not correlate with knowledge. The number of correct knowledge questions appears to decrease with years of practice (p < 0.01). Overall, respondents felt more confident about when to start antimicrobial therapy (88%) than when to stop it (62%), yet 29% did not know where to find hospital resources for optimal antimicrobial prescribing.
Conclusions: Respondents recognized AR as a critical issue, but awareness and knowledge on antibiotic misuse were lacking. Our results are consistent with prior similar surveys published globally. Barriers to optimal antimicrobial prescribing were identified and strategies for improving the effectiveness of the ASP will be developed accordingly.
Nadine Khalil1, Ana Cabrera2, Michael Silverman1, Michael Payne2, Megan Devlin1, Johan Delport2, Danielle Ouellette3, Lise Bondy1
1Division of Infectious Diseases, Western University, London, Ontario, Canada; 2Department of Microbiology, London Health Sciences Centre, London, Ontario, Canada; 3Department of Medicine, Western University, London, Ontario, Canada
Objectives: Q fever, caused by Coxiella burnetii, is transmitted by inhalation of particles that contain Coxiella organisms. We present an interesting case of possible transmission of Q fever from kissing and multiple subcutaneous abscesses secondary to Coxiella burnetii.
Method: A42-year-old previously well male presented with multiple bilateral strokes. Lumbar puncture demonstrated varicella zoster virus DNA by polymerase chain reaction (PCR), and treatment with IV acyclovir led to no improvement. He had bronchoscopies negative for infectious causes and normal transesophageal echocardiograms. Multiple chest wall and extremity cold abscesses developed, which were drained. Bacterial, fungal, and mycobacterial cultures were negative.
Results: 16S rRNA amplification and Sanger sequencing from the abscesses yielded C. burnettii. The original cerebrospinal fluid (CSF) sample and blood were also PCR positive for C. burnetii. He had positive Q fever titres (phase 1 IgG 1:256, IgM 1:64, and phase 2 IgG 1:4,096, IgM <1:16). C. burnetii culture from the abscesses was confirmed with immunofluorescence from France and was susceptible to doxycycline. Despite treatment with doxycycline and hydroxychloroquine, and then additional rifampin and ciprofloxacin, the patient deteriorated and was transferred to hospice care before his death. No zoonotic risk was identified. Contact and droplet precautions were maintained throughout his hospitalization but discontinued in hospice. His partner gave her final goodbyes with kissing but no sexual intercourse or oral sex. She developed fevers and respiratory symptoms 5 days later, persisting for 4 weeks. Her Q fever serology showed phase 1 IgG 1:512, P1 IgM 1:32, phase 2 IgG 1:4,096, P2 IgM 1:16.
Conclusions: This case suggests possible Q fever transmission through kissing in the presence of chronic pulmonary infiltrates, which has not previously been reported, although person-to-person transmission via intercourse has rarely been reported. Q fever in the absence of valvular lesions with cold chest wall abscesses and profound lymphopenia has not been previously described in adults.
Megan McCreary1, Alena Tse-Chang2,3, Karen L Forbes2,4, Jessica L Foulds2,4
1Department of Medicine, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, Alberta, Canada; 2Department of Pediatrics, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, Alberta, Canada; 3Division of Pediatric Infectious Disease, Department of Pediatrics, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, Alberta, Canada; 4Division of Pediatric Hospital Medicine, Department of Pediatrics, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, Alberta, Canada
Objective: An antimicrobial stewardship (AS) intervention was implemented for pediatric medicine units using a twice-weekly in-person rounds-based approach to provide stewardship recommendations and education from an AS physician and AS pharmacist. We assessed the effects of this intervention by qualitatively exploring the experiences of pediatricians and pediatric residents 1 year after implementation.
Methods: This was an exploratory qualitative study. Purposeful sampling was used to recruit participants for individual interviews. Pediatricians and residents who attended ≥1 stewardship rounds (SR) were included. A semi-structured interview guide was created that focused on perceptions of AS, personal experiences at SR, and perceived impacts on patient care. Using a constant comparative analysis approach, codes were developed and collapsed into subthemes then themes.
Results: Eight pediatricians and 10 residents completed interviews. Qualitative analysis yielded three themes: “Insights into Clinical Reasoning,” “Opportunity for Growth and Learning,” and “Establishing and Exploring Professional Relationships.” SR encouraged participants to critically evaluate antimicrobial choices and review evidence to make informed decisions. The in-person approach allowed participants to explain their rationale and engage in discussion with the AS team. Furthermore, participants felt a sense of validation at SR and gained confidence for prescribing antimicrobials in the future. The educational aspect was an important benefit, particularly for senior residents. SR provided face-to-face interaction with infectious disease (ID) physicians and some participants felt more comfortable consulting ID service because of this. In contrast, others worried physicians may avoid ID consultation and wait for input at SR.
Conclusions: Participating clinicians found SR to be an effective strategy for education and development of clinical reasoning skills for optimal antimicrobial prescribing. The effects of our intervention on timing and frequency of ID consultation is an interesting finding shared by participants. Further research into patient important outcomes and consultation practices are needed to explore this.
Georgia L Carr, Wendy Gouliquer, Aaron Skillen, David Welbourne, Gregory Gamble
Thunder Bay Regional Health Sciences Centre, Thunder Bay, Ontario, Canada
Objectives: Timely detection or exclusion of respiratory pathogens is paramount to targeted patient management. While diagnostic solutions draw more funds from limited global budgets, they can contribute to overall corporate efficiencies by, among other things, decreasing length of stay (LOS). Although respiratory virus testing is funded free of charge by Public Health Ontario Laboratories (PHOL), our remote geographical location inherently causes delayed turn-around times (TAT). As a solution, respiratory multiplex PCR testing was adopted in-house to improve patient flow. LOS and costs were compared.
Methods: A retrospective chart review was performed comparing outcomes 1 year pre- (2017-04-03 to 2018-03-31) with 1 year post-implementation (to 2019-03-31) of bioMérieux’s FILMARRAY Respiratory Panel (bioMérieux, St. Laurent, PQ, Canada). This test identifies the same pathogens as the PHOL test, 17 viruses and 3 bacteria that cause upper respiratory tract infections, with an overall sensitivity and specificity of 95% and 99%, respectively. No physician-order restrictions were put into place. We evaluated TAT, clinical, and economic impact; that is, emergency department (ED) LOS and hospitalization LOS in patients with a physician diagnosis of suspected respiratory tract infection.
Results: We included 771 specimens and 1,376 specimens in the pre- versus post-period. The average pre-TAT was10.32 days compared with 0.14 days (3.36 h) in the post-period, accounting for an absolute shortened TAT of 10.18 days (98.6% decrease). Specimens had a positivity rate of 30.0% pre- versus 47.8% in the post-period. LOS decreased by 1.36 days for inpatients and 2 hours for ED patients. Corporate cost savingsare an estimated $778,235.84, based on patient flow efficiencies and funded bed allotments.
Conclusions: Investment in rapid molecular diagnostics for upper respiratory infections creates opportunities to improve patient care, bed flow, and efficiencies for the corporation, especially in remote settings such as Thunder Bay. Pathogen-specific diagnosis can support the antimicrobial stewardship teams by allowing informed therapy adjustment (antimicrobial or antiviral) in a timely manner.
Ji-in Hum1, Benjamin Hon1, Min-Kuang Lee1, Navdeep Chahil1, Muhammad G Morshed1,2
1BC Centre for Disease Control—Public Health Laboratory, Vancouver, British Columbia, Canada; 2University of British Columbia, Vancouver, British Columbia, Canada
Objective: To enhance the current surveillance of tick-borne disease in British Columbia, Canada, we aim to expand our existing Borrelia 23S Taqman polymerase chain reaction by integrating two additional tick-borne pathogens, Ehrlichia and Anaplasma, into the assay. This triplex real-time PCR test was designed to cover most of the species under these three genera, which serves as a pan-detection of tick-borne pathogen screening panel.
Methods: The primer and probe were designed to target 23S rRNA of Borrelia spp and 16S rRNA of Ehrlichia spp, and Anaplasma spp DNA was extracted using QIAGEN Blood; tissue extraction and real-time PCR was performed on the ABI Taqman 7500 using ABI TaqMan 2X Fast Advanced PCR Master Mix. Synthesized DNA gBlocks were used for analytical validation, and 50 specimens for clinical validation (including 20 specimens from Mayo Clinic in the United States for blind testing).
Results: The reportable range of the assay is 1E2 to 1E6 and the R-squared values for all three targets are greater than 0.99, indicating a great linearity throughout the reportable range. The detection limit for all tests is 1E2 copies per reaction. Precision was calculated from triplicates over four runs. All targets have a coefficient of variation (% CV) ≤5.12% (within the acceptable 15% range). The clinical sensitivity is 100% for Ehrlichia and 87.5% for Anaplasma. The clinical specificity for both targets is 100%.
Conclusion: The real-time PCR assay described in this report is found to be a reliable method for screening of Borrelia, Anaplasma, and Ehrlichia species. Using retrospective samples, we were able to detect the newly identified Candidatus Ehrlichia khabarensis, which was identified using next-generation sequencing in the past. Implementation of this triplex assay will enhance vector surveillance program in British Columbia.
Carson Ka-Lok Lo1, Prameet M Sheth2,3
1Department of Medicine, Queen’s University, Kingston, Ontario, Canada; 2Department of Pathology and Molecular Medicine, Queen’s University, Kingston, Ontario, Canada; 3Division of Microbiology, Kingston General Hospital, Kingston, Ontario, Canada
Objectives: Carnobacterium species are non–spore-forming, lactic acid–producing gram-positive rods regarded as non-pathogenic to humans. They are frequently isolated from the environment and currently under investigation as a bio-preservative in the food/fish industry. The use of live bacteria in food, and bio-preservation, have recently led to several studies highlighting cases of bacteremia/sepsis associated with lactic acid bacteria used in probiotics (e.g., Lactobacillus spp). We report a case of pneumonia in an immunosuppressed cancer patient with Carnobacterium inhibens isolated in blood culture of uncertain clinical significance. We reviewed case reports reported in the literature of human infections with Carnobacterium spp.
Methods: A comprehensive search was conducted on all English-written articles using Embase, Ovid MEDLINE, PubMed, and Google Scholar.
Results: To date, only 4 cases of Carnobacterium spp isolated from humans have been reported. Two cases were identified with other mixed bacterial isolates from traumatic wounds. Two cases were isolated from blood cultures: 1) an immunocompetent man with 1 positive blood culture set presenting with febrile illness and an extensive history of handling and consuming fish; and 2) an immunosuppressed woman on enteral nutrition post–cardiac arrest with multiple positive blood cultures for C. divergens. Our case described an 81-year-old retired clergyman with metastatic castrate-resistant prostate cancer on enzalutamide and chronic steroids who was admitted for pneumonia empirically treated with ceftriaxone and vancomycin. Extensive computed tomography imaging confirmed multifocal pneumonia without other foci of infection. He was discharged with amoxicillin-clavulanate to complete a 7-day course and achieved clinical recovery. One of 2 blood culture sets from admission resulted positive for C. inhibens.
Conclusion: The pathogenicity and disease spectrum of Carnobacterium spp in humans remain unknown. The use of gram-positive bacteria in the food industry for their bio-preservative or fermentative capacity presents a potential source of unique organisms leading to disease, especially in immunocompromised patients.
Matthew Clifford-Rashotte1, Quynh-Dao Dinh1,2, Aleksandra Stefanovic2, Victor Leung2
1University of British Columbia Division of Infectious Diseases, Vancouver, British Columbia, Canada; 2Providence Health Care, Vancouver, British Columbia, Canada
Objective: Staphylococcus aureus bacteremia (SAB) is a frequently encountered problem associated with a high rate of complications and mortality. A recent publication employed a RAND-modified Delphi procedure to identify 25 quality indicators for the management of SAB, including timing of repeat blood cultures, investigation for endocarditis, appropriate antimicrobial therapy, and infectious diseases consultation. We sought to evaluate adherence to these quality indicators in our institution, an urban tertiary academic referral centre.
Methods: We performed retrospective chart review of all cases of SAB identified from May to October 2019 in our hospital.
Results: One-hundred-fifteen SAB cases were identified, of which 106 met the criteria for analysis (5 patients were receiving palliative care only and 4 were transferred to another facility). Of those, 98 (92.5%) had community onset bacteremia. There was a high rate of repeat blood cultures (96.2%), infectious diseases consultation (95.3%), and use of cloxacillin/cefazolin as definitive therapy for MSSA (91.7%). Transthoracic echocardiogram (TTE) was completed in 87 patients (82.1%), a median of 6 days (IQR 3–10) after blood culture positivity. 19 patients (17.9%) did not undergo TTE, and among this group 73.7% were discharged against medical advice. Transesophageal echocardiogram (TEE) was completed in 12 patients (11.3%) after a median of 13 days (IQR 8.75–18.25).
Conclusions: Our institution has a high rate of adherence to many quality indicators for the management of SAB. The rate of echocardiography was lower than expected, and this could be partly attributed to patients who were discharged against medical advice. Additionally, there was a low rate of utilization of TEE. Wait times for both TTE and TEE were outside of the recommended range of 3–5 days. Further evaluation should attempt to elucidate patient and system factors associated with echocardiogram wait times and non-completion, as well as to better define the role of TEE.
Robinson Truong1,2, Vincent Tang1, Troy Grennan3, Darrell HS Tan1,2,4,5
1Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada; 2Centre for Urban Health Solutions, St. Michael’s Hospital, Toronto, Ontario, Canada; 3Division of Infectious Diseases and Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; 4Division of Infectious Diseases, St. Michael’s Hospital, Toronto, Ontario, Canada; 5Department of Medicine, St. Michael’s Hospital, Toronto, Ontario, Canada
Objectives: There is interest in daily doxycycline as pre-exposure prophylaxis (PrEP) against sexually transmitted infections (STIs) but concern about antimicrobial resistance (AMR). We conducted a systematic review (CRD42017077320) of the impact of oral tetracycline class antibiotics on the development of AMR in normal flora.
Methods: We searched MEDLINE, Embase, the Cochrane Library (1940–2019) and conference proceedings (2014–2019) for randomized controlled trials in adults comparing daily oral tetracycline class antibiotics to non-tetracycline controls. The primary outcome was AMR to tetracyclines, measured using the burden of resistant isolates, minimum inhibitory concentrations and/or the emergence of resistance genes. Secondary outcomes included cross-resistance to non-tetracyclines. We extracted data onto standardized forms and assessed articles for risk of bias. Data were not appropriate for meta-analysis, so we analyzed findings descriptively.
Results: Our search yielded 5,602 abstracts, of which 7 articles fulfilled inclusion criteria. Most were at moderate/high risk of bias, generally due to inadequate methodologic reporting. Studies used doxycycline, tetracycline, oxytetracycline, or minocycline for 2–18 weeks. Most observed an increased burden of tetracycline resistance, including in subgingival (n = 3 studies), gastrointestinal (n = 2), and upper respiratory (n = 1) flora; a single study of skin flora found no change in mean growth score for the prevalence of tetracycline-resistant Propionibacteria after 18 weeks of oxytetracycline/minocycline. Four studies reassessed for AMR at 2–50 weeks post-intervention and reported varying degrees of AMR persistence. Three articles reported on cross-resistance to non-tetracycline antibiotics, of which one found a transient increase among Escherichia coli in the gastrointestinal tract while the other two showed no difference from control.
Conclusion: Though effects may be transient, limited data suggest that oral tetracyclines for 2–18 weeks produce an increased burden of resistance in subgingival, gastrointestinal and upper respiratory flora. It is critical that STI PrEP trials study the impact of long-term doxycycline on AMR in commensal bacteria.
Ana Clara Arantes Vieira, Carolyne Sthéfany Gonçalves Silva, Ester Olivia De Oliveira Menino, Gabriela Cristina Libanio, Luiza Buchemi Cardoso, Thiago Breder Leandro Pereira, Bráulio RGM Couto
Centro Universitário de Belo Horizonte, Belo Horizonte City, Minas Gerais, Brazil
Background: The objective of our study is to answer two questions: a) What is the surgical site infection (SSI) risk after abdominal aortic aneurysm repair (AAA)? b) What are the risk and protective factors for hospital death, SSI, and extra length of stay in patients undergoing AAA repair?
Methods: A retrospective cohort study, based on revised data from medical records between July 2016 and June 2018 of 600 patients from one general hospital in Belo Horizonte, a 3,000,000 inhabitants city from Brazil. Data were gathered by standardized methods defined by the National Healthcare Safety Network (NHSN)/Centers for Disease Control (CDC) procedure-associated protocols for routine SSI surveillance. 20 preoperative and operative variables were evaluated by univariate and multivariate analysis (logistic regression).
Results: Nineteen SSI were diagnosed (risk = 3.2% [95% CI 2.0% to 4.9%]). Among the variables related to SSI, the factors that had the greatest impact were more than one surgical procedure, besides AAA (relative risk [RR] = 5.5; p < 0.001), and preoperative hospital length of stay greater than 24 hours (RR = 4.5; p = 0.027).
Conclusion: We identified one intrinsic risk factor for SSI after AAA repair (multiple surgical procedures), and one extrinsic risk factor, the preoperative length of hospital stay. The incidence of infection can be reduced significantly when patients are submitted to surgery at the first day of hospital admission.
Joan L Robinson, Dolores E Freire Jijon, Liza Bialy
University of Alberta, Edmonton, Alberta, Canada
Objective: To compare interventions in cerebrospinal shunt infections.
Methods: The search included Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE, Ovid Embase, Wiley’s Cochrane Library, CINAHL Plus with Full Text via EBSCOhost, Scopus advanced search, and Web of Science Core, for 1990 through May 2019. Studies were included if they reported the results of two different management strategies. Studies that compared complete versus incomplete shunt removal were excluded as complete shunt removal is clearly optimal.
Results: The search identified 2,208 records, of which 8 met the inclusion criteria. All were moderate quality cohort studies. Seven included only children. One analyzed delay in cerebrospinal fluid (CSF) sterilization while the others analyzed relapse or reinfection rates; none distinguished between relapse and reinfection. Four compared duration of antibiotics; none showed that a longer course prevented relapse/reinfection. Two analyzed addition of rifampin, one showing a decrease in relapse/reinfection and the other with a small sample size. No studies analyzed addition of intraventricular antibiotics, but one showed equally good results with twice-daily versus once-daily administration. One study reported no difference in relapse/reinfection with placement of an antibiotic impregnated catheter following a shunt infection. There was no advantage to shunt replacement 7 days or more versus less than 7 days after CSF became sterile, or to shunt replacement with sterile CSF cultures 48 hours versus 24 hours after antibiotics were stopped. A new shunt entry site did not decrease the risk of relapse/reinfection.
Conclusion: There is a paucity of high-quality studies, and all have small sample sizes. Addition of rifampin for staphylococcal infections may decrease relapse but needs to be further studied. There is a need for studies that compare the durations of antibiotics and the timing of shunt replacement. Studies should distinguish between relapse and reinfection.
Xena X Li1, Donald B Duncan2, Nishma Singhal1
1Division of Infectious Diseases, Department of Medicine, McMaster University, Hamilton, Ontario, Canada; 2Division of Microbiology, Department of Pathology and Molecular Medicine, McMaster University, Hamilton, Ontario, Canada
Objectives: Salmonella species are known to cause mycotic aneurysms, and typically involve the aorta. Isolated carotid artery aneurysm is rare. We aim to review the literature and describe our own case highlighting the need to consider Salmonella species as an etiology of carotid aneurysms, and the importance of tissue for diagnosis.
Methods: A 72-year-old immune-competent female with peripheral vascular disease presented with 3 weeks of right neck pain and swelling leading to odynophagia and airway compromise. She had no systemic or localized infectious symptoms and no travel history. Imaging showed a right carotid pseudoaneurysm and she underwent emergent repair with a Dacron graft. One week post-operatively, she developed acute pain and swelling requiring a re-do graft repair. Cultures were sent despite no obvious signs of infection.
Results: Pathology from her initial surgery showed acute neutrophilic inflammation with focal fibrinoid vessel wall necrosis with an occlusive thrombosis and focal atherosclerotic plaque. Tissue cultures from her graft grew Salmonella enterica serovar typhimurium. Upon retrospective questioning, she had experienced an acute diarrheal illness 6 weeks prior, following a meal at a fast food restaurant. She was empirically started on ciprofloxacin and ceftriaxone and ultimately completed a 6-week course of ceftriaxone without complications. Subsequent CT-A did not show arteritis elsewhere, and her echocardiogram was normal. Blood cultures, VDRL, and Lyme serologies were negative.
Conclusions: Salmonella mycotic aneurysms are an uncommon complication of systemic Salmonella infection, and isolated carotiditis is very rare. There are approximately 20 cases per decade of extracranial infected carotid artery aneurysms in the literature, of which Salmonella is the causative agent in approximately 10%. Culture and pathology are key to suspecting an infection when the initial diarrheal illness has already resolved and may have been forgotten.
Dolores E Freire Jijon1,2, Joan Durand3,4, Shirley Perry3,4, Alena Tse-Chang1,3, Crystal Lefebvre3,4, Marta Rojas-Vasquez3,4, Bonita E Lee1,3
1University of Alberta, Edmonton, Alberta, Canada; 2Universidad de Guayaquil, Guayaquil, GY, Ecuador; 3Stollery Children’s Hospital, Edmonton, Alberta, Canada; 4Alberta Health Services, Edmonton, Alberta, Canada
Objective: Our objective was to compare the epidemiology of community-acquired (CA) and hospital-acquired (HA) bloodstream infections (BSI) in pediatric patients with malignancies and central lines admitted to a tertiary pediatric centre.
Methods: Retrospective chart review was performed for patients with a diagnosis of malignancy and a central line and positive blood cultures admitted to the oncology unit from January 2017 to June 2019. Oncological diagnoses, information regarding central line placement, and microbiological data were recorded. Cases were reviewed and classified using the NHSN definition for BSI into central line-associated BSI (CLABSI), mucosal barrier injury (MBI), and secondary BSI. RStudio (RStudio, PBC, Boston, MA) was used for statistical analysis.
Results: A total of 49 BSI events were reviewed: 7 were excluded according to NHSN criteria for commensal organisms; 27 were HA with 11 (41%) as CLABSI, 15 (56%) MBI and 1 (3.7%) secondary; and 15 were CA with 11 (73%) CLABSI, 2 (13%) MBI, and 2 (13%) secondary (p < 0.05). Four BSI had more than one organism isolated: 3 (11%) were HA and all were MBI; 1 (7%) was a CA CLABSI. The proportion of gram-positive and gram-negative was similar between HA and CA BSI, with Staphylococcus aureus and Escherichia coli as the top two organisms in both groups. The ratio of hematological malignancy versus solid tumour was 22:3 for HA and 2:3 for CA BSI. The median time of BSI since central line insertion for HA and CA BSI was 29 days (interquartile range [IQR] 11–80) and 138 days (IQR 79–238) (p < 0.001), and the time to clearance for HA and CA BSI was 1 day (IQR 1–2) and 3 days (IQR 2–4) (p < 0.05), respectively.
Conclusion: There are different clinical characteristics in HA versus CA BSI in patients under oncology care with central lines. It is important to include both groups under surveillance for program evaluation and quality improvement.
Torrance Oravec1, Samna Khan1, Jennifer Leigh2, Liam Matthews3, Raheel Inayatullah3, Cheryl Main1, Dominik Mertz1, Peter Daley3, Bahareh Ghadaki4, Anjali Shroff1
1McMaster University, Hamilton, Ontario, Canada; 2University of Ottawa, Ottawa, Ontario, Canada; 3Memorial University, St. John’s, Newfoundland, Canada; 4St. Catharines General Hospital, St. Catharines, Ontario, Canada
Objectives: Infective endocarditis (IE) caused by Streptococcus agalactiae (GBS) is increasingly reported and associated with an aggressive course and high mortality rate. Existing literature on GBS IE is limited to case series; we compared the characteristics of patients with GBS IE to patients with GBS bacteremia without IE to identify risk factors for development of endocarditis.
Methods: Retrospective case control study across seven centres in three Canadian cities over 18 years. A chart review was conducted on patients with GBS bacteremia to identify those with probable or definite IE (per Modified Duke Criteria) and three controls per case, matched by time and location. Information was collected regarding patient demographics, treatment, and outcomes. Uni- and multivariate analyses were completed using a generalized estimation equation (GEE) to adjust for the case control design.
Results: Of 520 patients with GBS bacteremia, 28 cases of IE were identified. Among cases, 68% (19/28) met criteria for definite IE, surgical management was pursued in 29% (8/28), and the overall in-hospital mortality rate was 29% (8/28). Multivariate analysis demonstrated that patients with IE were more likely to have native valve disease (35% versus 8%, p = 0.001), a history of alcoholism (14% versus 1%, p = 0.047), lack of identifiable source (50% versus 23%, p = 0.02), and were younger (p = 0.05). Univariate analysis also identified prosthetic valves and injection drug use as risk factors, but these variables were excluded from the multivariate analysis due to collinearity.
Conclusions: Among patients with GBS bacteremia, those with native valve disease, no identifiable source, younger age, and those with addictions are at higher risk of IE. These results emphasize the importance of recognizing GBS as a causative organism in IE, and recommendations for future studies include characterization of virulence factors for strains associated with endocarditis.
Ryan Marinovich1, Santiago Perez-Patrigeon2, Evan Wilson2
1Department of Medicine, Queen’s University, Kingston, Ontario, Canada; 2Division of Infectious Disease, Queen’s University, Kingston, Ontario, Canada
Objectives: Mycobacterium avium complex (MAC) infections are a rare but serious complication of solid organ transplantation. Due to a variety of drug–drug interactions between antimicrobials and immunosuppressants used in transplantation, treatment is often challenging. We describe a case of disseminated MAC in a renal transplant recipient with gastrointestinal involvement and multiple drug intolerances to first line agents. We also present a literature review of MAC infections in renal transplant patients.
Methods: A comprehensive search was conducted on English language literature published from 1990 to December 2019 on MAC infections in adult kidney transplant recipients, using the PubMed database.
Results: Our case follows a 55-year-old female with a remote history of renal transplant for polycystic kidney disease who presented with vague abdominal discomfort, weight loss, anemia, and decline in graft function. Upper endoscopy revealed diffuse duodenitis and MAC was confirmed in duodenal, bone marrow, and graft biopsies. Treatment was complicated by azithromycin intolerance, hives secondary to clarithromycin, and rifabutin-related hepatitis. Fluoroquinolone and linezolid resistance were later confirmed. Our literature review identified 13 cases of MAC infections in adult kidney transplant recipients: 6 cases (46%) were localized infections, while 7 cases (54%) were disseminated. Intestinal involvement was described in 3 cases. The median time from transplant to diagnosis was 4 years (range 1–20 years). All cases received antimicrobial treatment, and 7 cases (54%) required decreasing immunosuppression. MAC-related death was reported in 4 cases (30%), and 5 cases (38%) experienced graft rejection. Remission of MAC infection with maintained graft function was reported in only 5 cases (38%).
Conclusions: Disseminated MAC infection, particularly gastrointestinal (GI) involvement, is rare in renal transplant recipients. The presentation may be insidious and diagnosis requires thorough investigation, often with invasive biopsy. Treatment is further challenged in settings of drug intolerances and interactions.
Daniel B Gregson1,2, Ian Lewis3, Ashlee Earl4, Rauf Salamzade4, Thomas Rydzak1, Jenna Poelzer1, Diana Doyle2, Sydney Walker2, Tara Winstone2, Deirdre Church1,2
1University of Calgary, Calgary, Alberta, Canada; 2Alberta Precision Laboratories, Calgary, Alberta, Canada; 3University of Calgary, Calgary, Alberta, Canada; 4Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA
Objectives: Penicillin is considered the therapy of choice for penicillin-susceptible MSSA. There is controversy regarding the best phenotypic test method to exclude the presence of blaZ in MSSA. Our objective was to determine the best phenotypic method to detect blaZ-positivity for use in a laboratory without access to blaZ-PCR.
Methods: Isolates from 154 patients with MSSA bacteremia previously analyzed by whole-genome sequencing (WGS) were included in this study. Isolates were divided into blaZ-positive and negative strains based on this sequencing. VT2 results from the patient’s initial blood culture were included as the MIC and interpretation. Technologists performed DST and cliff edge testing according to EUCAST (P1) and CLSI (P10) methods and nitrocephin testing at the DST zone edge, blinded to prior results.
Results: The blaZ gene was detected in 111 (72%) of these isolates. All isolates resistant on VT2 (99) were blaZ-positive. However, 12/55 (22%) sensitive on Vitek II testing were blaZ-positive. In 7/12 of these cases the MIC was 0.12 µg/mL. P1 and P10 disk testing detected 110 (99%) and 106 (95.5%) of blaZ-positive isolates respectively (p = 0.10). Similarly, all nitrocephin-positive isolates were blaZ-positive. Nitrocephin testing was negative in 13 and 8 of blaZ-positive isolates on the P1 and P10 tests, respectively. Nitrocephin testing, and/or the presence of a cliff edge, did not add to the disk zone size interpretation. P1 and P2 testing of isolates sensitive on VT2, reduced the very major error rate from 22% to 0.02% and 0.09%, respectively (p = 0.09)
Conclusions: MSSA bacteraemias in our region are largely resistant to penicillin as determined by the presence of blaZ on WGS. Very major errors are common with VT2 and penicillin susceptibility needs to be confirmed with DST. There was no statistical difference between the EUCAST and CLSI DST breakpoints for the detection of blaZ.
Vicki Doyle1, Michelle Whittle1, Claire Pratt2, Zahra Rehan2, Carla Penney2, Ashley Waghmare3, Peter Daley2
1Eastern Health, St. John’s, Newfoundland, Canada; 2Memorial University, St. John’s, Newfoundland, Canada; 3Lawton’s Drugs, St. John’s, Newfoundland, Canada
Objectives: We provided education to nurses in long-term care facilities (LTCFs) on urine culture collection and treatment of asymptomatic bacteriuria, and measured impact on total antimicrobial use (AMU) rate and appropriateness of treatment for urinary tract infection.
Methods: All antimicrobial prescriptions given in all LTCF in one city in 2015 (pre-intervention period) and 2018 (post-intervention period) were provided by the local pharmacy. Prescriptions given for urinary tract infection were reviewed retrospectively using published appropriateness criteria.
Results: Education was provided to 352 nursing staff in 15 LTCFs between January and July 2017. In 2018, 2,159 prescriptions were given to 868 patients in eight LTCFs, representing 20.05 defined daily doses (DDD)/1,000 patient days, or 2.50 prescriptions/patient, with median duration 9.5 days. This represents an absolute reduction compared with 2015 of −1.11 DDD (p = 0.53). All 79 prescriptions given for urinary tract infection in one LTCF were analyzed; 40/79 (50.6%) were appropriate using Loeb 2001 criteria and 31/79 (39.2%) were appropriate using Happe 2017 criteria. This represents an absolute improvement compared with 2015, of +26.9% (p = 0.001) using the Loeb 2001 criteria.
Conclusions: AMU remains high in LTCFs. Education was associated with a non-significant reduction in total AMU and a significant increase in appropriateness in treatment of UTI.
Ana C Blanchard, Sandra Isabel, Kayur Mehta, Laurie Streitenberger, Renee Freeman, Yvonne Yau, Aaron Campigotto, Michelle Science
The Hospital for Sick Children, Toronto, Ontario, Canada
Background: Infections caused by carbapenemase-producing Enterobacteriaceae (CPE) are associated with increased morbidity and mortality. Despite global emergence in adults, the epidemiology of CPE in children has not been well studied.
Objective: The aim of this study was to describe CPE colonization in a tertiary care pediatric hospital in Canada.
Methods: Two hospital-wide inpatient point prevalence screens (PPS) were conducted at our institution in April 2017 and 2019. Stools were collected, preferably in a sterile container or using a rectal Eswab™ (Copan Diagnostics, Inc., Murrieta, CA, USA). Specimens were tested using CHROMID-Carba Smart (bioMérieux, St. Laurent, PQ, Canada)). Any bacteria growing was identified by MALDI-TOF, and susceptibility testing was performed by broth microdilution (Sensititre), meropenem, and ertapenem disks. Carbapenem-resistant isolates were sent for molecular confirmation at the provincial reference laboratory.
Results: In the 2017 PPS, specimens were obtained from 242 (89.6%; 83 stools and 159 rectal swabs) of 270 admitted patients. CPE was isolated in one patient (0.4%) from the oncology unit; the organism was an Escherichia coli with an OXA-48-like gene. This patient had risk factors for CPE, including prolonged hospitalizations in the previous year in a country where significant OXA-48 outbreaks have been reported. In the follow-up 2019 PPS, specimens were obtained from 278 (93.0%; 164 stools and 114 rectal swabs) of 299 admitted patients. CPE were isolated in two (0.7%) patients; both were E. coli with OXA-48-like genes. While Canadian-born and without a history of hospitalization outside of Canada, both patients had a history of long-term travel (>1 month) to India.
Conclusions: In our centre, only OXA-48-like β-lactamases were identified in E. coli. The majority of cases were identified on high acuity units, which supports the need for a screening program with careful consideration for screening criteria as not all patients had a history of hospitalization outside of Canada.
Ifueko J Adeghe1, Maureen Buchanan-Chell2, Geoffrey D Taylor3, Stephanie W Smith3
1Department of Medical Microbiology, University of Alberta, Edmonton, Alberta, Canada; 2Department of Infection Control and Prevention, Alberta Health Services, Edmonton, Alberta, Canada; 3Department of Infectious Diseases, University of Alberta, Edmonton, Alberta, Canada
Background: Although Enterococcus is a normal flora of the gastrointestinal tract and female genital tract, it has the potential to cause nosocomial infections, particularly in the immunocompromised. Over time, we have seen increasing resistance to vancomycin, mostly in hospitalized patients. Currently, there is significant variation in vancomycin-resistant Enterococcus (VRE) screening in Canada. Some hospitals continued to screen while others have discontinued.
Objectives: The University of Alberta conducts prospective surveillance for all VRE infections. In 2015, the hospital modified its screening policy whereby only transplant and hematology/oncology patients are screened due to the high morbidity associated with VRE infection in these groups. We assessed the utility of our current screening policy by evaluating the incidence of VRE infection from 2013 to 2019.
Methods: This is a single-centre retrospective cohort study using data prospectively collected by the Infection Prevention and Control Department at the University of Alberta Hospital and Mazankowski Heart Institute, a tertiary care 700-bed hospital. The hospital provides a wide variety of surgical and medical services, including a large transplant program and management of hematologic malignancies. All patients with positive cultures for VRE were reviewed to determine presence of infection. VRE infection was determined using standard National Healthcare Safety Network (NHSN) definitions. Attributable mortality was determined retrospectively by chart review. We compared rates of hospital-acquired VRE infection in the screened and non-screened populations and attributable mortality rates between 2013 and 2019.
Results: Between 2013 and 2019, 161 patients had hospital-acquired VRE infection, with 115 non-bloodstream infections (NBSI) and 46 bloodstream infections (BSI). During this time frame, the rates of VRE BSI and NBSI in the screened (transplant/hematological oncology) and the unscreened (general) population increased significantly. Prior to 2015, infection rate in the hematology/oncology/transplant population was 2.4/10,000 patient days, but it increased to 6.8/10,000 patient days between 2015 and 2019. The infection rate in the unscreened population was 0.46/10,000 patient days and increased to 1.14/10,000 patient days post change. The attributable mortality was 15.6% and 10.9% in the screened and non-screened groups, respectively (OR 1.52, 95% CI 0.58 to 3.98).
Conclusion: Despite screening and isolation, VRE infection rates continued to be higher in hematological/oncology and transplant patients. VRE infection was associated with higher attributable 30-day attributable mortality in this group, although not statistically significant. However, rates of increase appeared to be lower in the screened group than in the non-screened group. Our current screening policy captures this vulnerable population; however, further work needs to be done to identify optimal prevention.
Robert Taylor1,2, Philippe RS Lagacé-Wiens2,3, Christine Turenne2,3, Joelle Carlson2, Markus Stein2,3, Heather J Adam2,3, Andrew Walkty2,3, James A Karlowsky2,3
1Max Rady College of Medicine, Rady Faculty of Health Sciences, Winnipeg, Manitoba, Canada; 2Shared Health, Winnipeg, Manitoba, Canada; 3University of Manitoba, Winnipeg, Manitoba, Canada
Objectives: The current study was undertaken to verify the Vacutainer Plus C&S preservative tube (Becton–Dickinson) for the recovery of urinary pathogens from transported urine samples before its introduction into our clinical laboratory system. The performance of our current urine transport container, UriSwab™ (UriSponge™, Copan Diagnostics, Inc., Murrieta, CA, USA), was also evaluated and performance of the two containers compared.
Methods: The Vacutainer and the UriSwab™ were compared using criteria defined in section 9.2 of the Clinical and Laboratory Standards Institute (CLSI) M40-A2 Quality Control of Microbiology Transport Systems, Approved Standard. Nine organisms associated with urinary tract infections were tested using 10-fold dilutions of 105–108 CFU/L. Transport containers were stored at 4oC and room temperature (21oC) over 5 days. Aliquots were plated to selective agars in triplicate, incubated for 20 hours at 35 ± 2oC, and colony counts performed.
Results: The two transport containers behaved almost identically at 4oC, having minimal growth variation even after 5 days, with the single exception of Streptococcus agalactiae for the Vacutainer, which had a 21-fold decrease in colony counts at 106 CFU/L in 24 hours. At room temperature (21oC), the Vacutainer maintained the integrity of the organisms tested for 5 days, while the UriSwab™ did not. The UriSwab™ caused marked overgrowth at 21oC for certain microorganisms: with a starting dilution of 105 CFU/L, Proteus mirabilis had a 2-fold increase in colony counts in 24 hours, Staphylococcus saprophyticus had a 11-fold increase in 48 hours and Enterococcus faecalis had an 8-fold increase in 24 hours, compared with less than 1-fold colony count increases seen for the Vacutainer.
Conclusions: This work demonstrated that samples delivered to the laboratory in the Vacutainer Plus C&S preservative tube, including samples with lower pathogen counts transported at room temperature for extended periods of time, are more likely to result in representative culture results than similar samples received using UriSwab™.
Céline M Allen1, Jennie Johnstone1,2,3, Frances B Jamieson3, Kevin May3, Hafid Soualhine4, Catherine Yoshida4, Daniel Kein4, Christopher E Kandel1,5, Eric Coomes1, Luke Devine1,2, Jeremy Kobulnik1,2, Susan M Poutanen1,2,5
1University of Toronto, Toronto, Ontario, Canada; 2Sinai Health System, Toronto, Ontario, Canada; 3Public Health Ontario, Toronto, Ontario, Canada; 4National Microbiology Laboratory, Winnipeg, Manitoba, Canada; 5University Health Network, Toronto, Ontario, Canada
Objectives: A global clonal outbreak of Mycobacterium chimaera infection has been linked to contaminated Stöckert 3T heater–cooler devices (HCD) used for cardiopulmonary bypass during surgery manufactured in Germany by LivaNovo between 2008 and September 2014. Health Canada released a safety alert in October 2016 with recommendations for risk mitigation measures. Three cases of M. chimaera infection (2 adults, 1 pediatric) linked to the use of contaminated HCD in Canada occurring prior to the 2016 safety alert have been reported. We describe an M. chimaera infection in Canada closely related to the HCD M. chimaera outbreak strain post-recommendations for risk mitigation measures.
Methods: The patient’s medical record was reviewed to abstract clinical information. Whole genomic sequencing (WGS) of the M. chimaera isolate was completed and the isolate was compared with the outbreak reference strain (M. chimaera ZUERICH-1) as well as other strains linked to the global outbreak using single-nucleotide polymorphism analysis.
Results: An 80-year-old man presented with weight loss, chronic cough, and generalized weakness in July 2019. He had undergone a bioprosthetic aortic valve replacement in 2006 and coronary-artery bypass surgery in 2017. Investigations revealed thrombocytopenia, lymphopenia, elevated liver enzymes, and non-necrotizing granulomas on bone marrow biopsy. Transthoracic and transesophageal echocardiograms were negative for valvular abnormalities. The patient returned 3 months later with ongoing malaise. Bone marrow aspirate and biopsy cultures grew M. chimaera. Treatment was initiated with rifampin, ethambutol, azithromycin, and moxifloxacin. WGS revealed the isolate was closely related to M. chimaera ZUERICH-1 and other strains associated with the HCD outbreak.
Conclusions: This case highlights the importance of recognizing the risk of M. chimaera infection persists in patients with a history of surgery with cardiopulmonary bypass despite the recommendation for mitigation measure implementation. Further study on the effectiveness of mitigation measures is needed.
Gautham J Nair1, Bryn Hazlett2, Susy Hota1,3, Julianne Kus4, Ian Brasg1, Susan M Poutanen1,2
1University of Toronto, Toronto, Ontario, Canada; 2University Health Network/Sinai Health Department of Microbiology, Toronto, Ontario, Canada; 3University Health Network, Toronto, Ontario, Canada; 4Public Health Ontario, Toronto, Ontario, Canada
Objective: Candida auris is a newly emerged multi-drug-resistant organism associated with nosocomial outbreaks. Older identification systems risk misidentifying C. auris isolates. A case of C. auris candidemia with concomitant C. lusitaniae candidemia based on identifications using VITEK MS (IVD knowledge base [KB] 3.0) (bioMérieux, St. Laurent, PQ, Canada) was identified in our laboratory. Re-identification of the C. lusitaniae isolate using VITEK MS (IVD KB 3.2) confirmed that the C. lusitaniae isolate was a misidentified C. auris. C. lusitaniae has not previously been documented as a possibly misidentified species using the VITEK MS. To confirm that other non-auris Candida isolates were not misidentified C. auris, a retrospective review was completed.
Methods: We reviewed all archived Candida isolates between 1999 and 2019 identified in our laboratory information system and the VITEK MS database for species identified by Public Health Ontario (PHO) and the U.S. Centers for Disease Control as being possibly misidentified C. auris, regardless of methodology. Isolates were subcultured twice onto sheep blood agar from −80°C and then re-identified using the VITEK MS (IVD KB 3.2).
Results: Of the 249 isolates tested, 3 isolates were identified as C. auris using VITEK MS (IVD KB 3.2). All three were from the original index patient. Two had no identification using VITEK MS (IVD KB 3.0) and were identified in real-time at PHO, and the third had an initial low identification as C. haemulonii but was identified as C. lusitaniae with 99.9% confidenceupon refiring using IVD KB 3.0. Confirmation of its identification as C. auris was completed at PHO.
Conclusions: VITEK MS users should be aware that the IVD KB 3.0 has the potential to have misidentified C. auris as C. lusitaniae. It is reassuring that our retrospective review did not detect other cases in which this misidentification occurred. VITEK MS users should consider completing retrospective reviews of their own archived isolates.
Bryn Hazlett1, Tony Mazzulli1,2, Susan M Poutanen1,2
1University Health Network/Sinai Health System Department of Microbiology, Toronto, Ontario, Canada; 2University of Toronto Departments of Laboratory Medicine and Pathobiology and Medicine, Toronto, Ontario, Canada
Objectives: Rapid detection of carbapenemase-producing organisms (CPO), extended-spectrum β-lactamases (ESBL), and vancomycin-resistant Enterococcus (VRE) is important. Data suggest nucleic acid amplification tests (NAAT) are more sensitive than culture. We evaluated the Allplex Entero-DR assay (Seegene Inc., Seoul, South Korea). CPO-testing was done previously on 200 well-characterized isolates, 53 rectal Eswabs (Copan Diagnostics, Inc., Murrieta, CA, USA), and 11 rectal swab broth cultures (AMMI CACMID 2019 Abstract L01). The purpose of this study was to verify the assay’s ability to detect CTX-M ESBLs and VRE.
Methods: ESBL and VRE testing was completed using: (a) 64 well-characterized isolates including 33 vanA, 6 vanB, and 22 CTX-M; (b) 37 rectal Eswabs, including 14 CTX-M, 10 vanA, and 14 non-ESBL/VRE; and (c) 20 rectal swab on-target positive broth cultures (10 CTX-M and 10 vanA). (d) LOD for VRE was calculated using 3 QC strains following manufacturer direct-from-specimen-protocols using 10E7–10E9 CFU/L concentrations with Xpert vanA/vanB (Cepheid, Sunnyvale, CA, USA) (our laboratory’s current method) as reference.
Results: For CPO-testing, all targets were detected and all non-targeted-CPO/non-CPO were negative for each type as previously presented. For ESBL and VRE testing: (a) of the 64 isolates, all VRE and CTX-M were detected. VRE-detection sensitivities/specificities (Sn/Sp) were 100%; respective Sn 95% CI were vanA (87.61 to 100) and vanB (55.72 to 100), and CTX-M ESBL-detection Sn/Sp were 100% (Sn 95% CI 82.45 to 100); (b) of the 37 rectal Eswabs, all target vanA and CTX-M were detected all non-target were negative. vanA-detection Sn/Sp were 100% (Sn 95% CI 67.91 to 100) and CTX-M detection Sn/Sp were 100% (Sn 95% CI 74.85 to 100); (c) of the 20 rectal swab on-target positive broth cultures, all vanA and CTX-M were detected. vanA-detection Sn/Sp were 100% (Sn 95% CI 67.91 to 100) and CTX-M detection Sn/Sp were 100% (Sn 95% CI 67.91 to 100); and (d) LOD (cfu/L) are shown (Table A061).
Conclusion: The Seegene Allplex Entero-DR Assay was highly accurate for CPO, VRE, and CTX-M detection.
Kathleen Simms1,2, Bryn Hazlett1, Jordan Fruitman1,2, Melissa Kissoon2, Satyender Singh3, Susy Hota3,4, Susan M Poutanen1,2,4
1Department of Microbiology, University Health Network/Sinai Health, Toronto, Ontario, Canada; 2Department of Laboratory Medicine and Pathobiology, University, Toronto, Ontario, Canada; 3Infection Prevention and Control Department, Toronto, Ontario, Canada; 4Department of Medicine, Toronto, Ontario, Canada
Objectives: The Microbiota Therapeutics Outcomes Program (MTOP) is a fecal microbiota transplantation (FMT) program that accepts healthy stool donors with no personal or family history of chronic illness. Negative microbiology screens including stool antimicrobial-resistant organism (ARO) cultures and enteric pathogens (EP) are required, with repeat screening prior to donation release. During routine screening, one donor was unexpectedly positive for an extended-spectrum β-lactamase (ESBL)-producing Klebsiella pneumoniae, and another was positive for a carbapenem-resistant Enterobacteriaceae. The purpose of this study was to determine the prevalence of intermittent ARO and EP carriage among healthy stool donors as well as the optimal frequency of donor rescreening required to detect this carriage.
Methods: All active donors between March 2017 and June 2019 were included. Stool aliquots stored at −80°C from each donation were thawed and tested for ARO [methicillin-resistant Staphylococcus aureus, vancomycin-resistant enterococci, ESBL, and carbapenemase-producing organisms (CPO)] using standard culture methods and 22 EP using a multiplex polymerase chain reaction platform. The prevalence of ARO and EP and optimal donor rescreening frequency were determined.
Results: Seven donors actively donated during the study period, with participation ranging from 2 to 16 months. Of 146 donations, 12 (8.2%; 95% CI 4.6% to 13.9%) were positive for ARO or EP. Rescreening donors every 2 months would detect only 25% of ARO and EP within this donor population, while donor rescreening every 2 weeks would increase the detection yield to 83.3%. However, 100% detection of all ARO and EP would require screening every donation.
Conclusions: There is frequent transient carriage of AROs and EP in healthy pre-screened stool donors. Clinicians involved with FMT need to recognize that FMT administration involves a risk of ARO and EP transmission and that donor rescreening at frequencies less than every donation will not detect all ARO and EP carriage.
Bryn Hazlett1, Pranav Tandon1, Kayla Miike1, Kathleen Simms1,2, Susan M Poutanen1,2
1University Health Network/Sinai Health Department of Microbiology, Toronto, Ontario, Canada; 2University of Toronto Departments of Laboratory Medicine and Pathobiology and Medicine, Toronto, Ontario, Canada
Objectives: Routine susceptibility testing methods do not include more recently approved agents. Fosfomycin IV has recently become available for use in North America. However, many labs do not have a verified way to test it. We verified various methods of fosfomycin susceptibility testing against gold-standard agar dilution.
Methods: An array of multi-drug resistant bacteria (44 Enterobacteraceae [Ent], 20 Pseudomonas aeruginosa [Pae], 19 Acinetobacter [Acb], 7 Aeromonas [Aer], 10 non-fermenters [NF], 6 Burkholderia cepacia, and 9 Stenotrophomonas maltophilia) were tested to determine very major errors (VME), major errors (ME), minor errors (MinE), and categorical/essential agreements (CA/EA) using ThermoFisher’s custom frozen and dry Sensititre microdilution panel (ThermoFisher, Burnaby, BC, Canada), Liofilchem’s commercial agar dilution and gradient strip (Liofilchem S.r.I., Roseto degli Abruzzi, Teramo, Italy), bioMérieux’s ETEST strip (bioMérieux, St. Laurent, PQ, Canada), and ThermoFisher’s antimicrobial susceptibility disk (KB) compared with in-house agar dilution following Clinical and Laboratory Standards Institute (CLSI) standards. Acceptability was determined using ³90%EA/³90%CA/ < 3%VME/< 3%ME/< 10%MinE thresholds and calculating 95% CI using CLSI fosfomycin Ent breakpoints.
Results: Both the frozen and dry panels had acceptable EA/CA/VME/ME/MinE for all bacteria except for Pae, Acb, and S. maltophilia. Liofilchem’s agar dilutions had acceptable EA/CA/VME/ME/MinE for all bacteria groups tested—although some Ent and Acb had errors, the number of isolates were small. Both the Liofilchem and ETEST strips had acceptable EA/CA/VME/ME/MinE with Ent, Aer, NF, and B. cepacia. KB had unacceptable EA/CA/VME/ME/MinE for all bacteria groups except Ent, Aer, and B. cepacia. A summary of results for Ent/Pae/Acb/Aer/NF within range (green), within range with CI (orange) and out of range (red) is shown in Figure A063.
Conclusion: Liofilchem’s agar dilution is the most accurate method for testing fosfomycin. All other methods have limitations and should only be used for select bacteria groups. Confirmatory testing using more organisms would be helpful.
Bryn Hazlett1,2, Pranav Tandon1,2, Kayla Miike1,2, Susan M Poutanen1,2,3
1University Health Network, Toronto, Ontario, Canada; 2Sinai Health, Toronto, Ontario, Canada; 3University of Toronto, Toronto, Ontario, Canada
Objectives: Routine susceptibility testing methods do not include newer agents for resistant gram-negative organisms (MDRGN). We verified bioMérieux’s ETEST strips (bioMérieux, St. Laurent, PQ, Canada) and Liofilchem’s gradient strips (Liofilchem S.r.I., Roseto degli Abruzzi, Teramo, Italy) against gold-standard custom frozen broth microdilution (BMD).
Methods: MDRGN (44 Enterobacteraceae[Ent], 20 Pseudomonas aeruginosa [Pae], 19 Acinetobacter [Acb], 7 Aeromonas, 10 non-fermenters [NF], 6 Burkholderia cepacia, and 9 Stenotrophomonas maltophilia) were tested to determine very major errors (VME), major errors (ME), minor errors (MinE), and categorical/essential agreements (CA/EA) using bioMérieux’s ETEST and Liofilchem’s gradient strips for ceftazidime/avibactam [C/A], ceftolozane/tazobactam [C/T], and tigecycline [T], as well as Liofilchem’s meropenem/vaborbactam [M/V] strip and bioMérieux’s piperacillin/tazobactam [P/T] and trimethoprim/sulfamethoxazole[TMP/SXT] ETEST against BMD. Acceptability was determined using ³90%EA/³90%CA/< 3%VME/< 3%ME/< 10%MinE thresholds and calculating 95% CI using Clinical and Laboratory Standards Institute (CLSI) preferentially (or EUCAST) breakpoints.
Results: Liofilchem’s and bioMérieux’s ETEST T strips did not meet threshold criteria for any MDRGN group. All strips other than T had acceptable EA/CA/VME/ME/MinE with Ent, and acceptable EA/CA/ME/MinE with Aeromonas (except Aeromonas VME due to lack of resistant isolates). Only M/V and P/T had acceptable EA/CA/VME/ME/MinE for NF. Most strips were unreliable with Pae and Acb (Figure A064). M/V had acceptable EA/CA/VME/ME/MinE for both B. cepacia and S. maltophilia and C/T strips and TMP/SXT ETEST had acceptable EA/CA/VME/ME/MinE with B. cepacia. A summary of EA/CA/VME/ME/MinE results for Ent/Pae/Acb within range (green), within range taking CI into account (orange) and out of range (red) is shown in Figure A064.
Conclusion: Laboratories should be aware of the limitations of gradient strip supplemental testing of resistant Gram-negative organisms and should ideally perform verification testing prior to implementation.
Bryn Hazlett1,2, Pranav Tandon1,2, Kayla Miike1,2, Barbara Willey1,2, Susan M Poutanen1,2,3
1University Health Network, Toronto, Ontario, Canada; 2Sinai Health, Toronto, Ontario, Canada; 3University of Toronto, Toronto, Ontario, Canada
Objectives: We created a custom BMD R-GN panel with ThermoFisher (ThermoFisher, Burnaby, BC, Canada) and verified a lyophilized version of the panel against its frozen (gold-standard) counterpart.
Methods: The R-GN panel permit Clinical and Laboratory Standards Institute (CLSI) and EUCAST breakpoints for Enterobacteraceae [Ent], Pseudomonas aeruginosa [Pae], Acinetobacter [Acb], Aeromonas [Aer], non-fermenters [NF], Burkholderia cepacia [BC], and Stenotrophomonas maltophilia [SM]. Ceftazidime/avibactam (C/A), ceftolozane/tazobactam (C/T), colistin (C), imipenem/relebactam (I/R), meropenem/vaborbactam (M/V), plazomicin (P), and tigecycline (T) were included. R-GN (44 Ent, 20 Pae, 19 Acb, 7 Aer, 10 NF, 6 BC, and 9 SM) were tested on the lyophilized panel to determine very major errors (VME), major errors (ME), minor errors (MinE), and categorical/essential agreements (CA/EA) compared with frozen BMD. Acceptability was determined using ³90%EA/³90%CA/< 3%VME/< 3%ME/ < 10%MinE thresholds and calculating 95% CI using CLSI (preferentially)/EUCAST breakpoints.
Results: The lyophilized panel performed well with some exceptions. A summary of EA/CA/VME/ME/MinE results for newer agents within range (green), within range taking 95% CI into account (orange) and out of range (red) is shown (Figure A065). For Ent/Pae/Acb/Aer/NF: ceftazidime/avibactam, imipenem/relebactam, and plazomycin all had acceptable EA/CA/VME/ME/MinE for all species; cefolozane/tazobactam has acceptable EA/CA/VME/ME/MinE except for Pae/Acb; colistin has acceptable EA/CA/VME/ME/MinE except Pae; meropenem/vaborbactam had acceptable EA/CA/VME/ME/MinE except for NF; tigeycycline had acceptable EA/CA/VME/ME/MinE except for Ent/Acb/Aer. For BC/SM, imipenem/relebactam, meropenem/vaborbactam, and tigecycline had acceptable EA/CA/VME/ME/MinE but ceftazidime avibactam, ceftolozane/tazobactam, colistin, and plazomicin had at least one parameter outside of its threshold.
Conclusions: Although the lyophilized panel generally performed well, it does not replace the gold-standard frozen BMD for all drug–bug combinations.
Susy Hota1,2, Satyender Singh1, Tiffany Liang1, Bryn Hazlett1,3, Susan M Poutanen1,2,3
1University Health Network, Toronto, Ontario, Canada; 2University of Toronto, Toronto, Ontario, Canada; 3Sinai Health, Toronto, Ontario, Canada
Objectives: Our randomized clinical trial (RCT) previously found limited effectiveness of a single fecal microbiota transplantation (FMT) by enema following treatment for acute recurrent CDI. Other studies have shown improved outcomes primarily with multiple FMT given at a time other than acute recurrence. Through MTOP, we offer multiple FMT to patients with rCDI at a time distant from acute recurrence. MTOP records of patients with rCDI treated with FMT were reviewed to document clinical outcomes.
Methods: All patients between 2017 and 2019 enrolled in MTOP for rCDI were included. Records were reviewed to capture clinical outcomes, gender, rCDI history/risk factors, and comorbidities. Clinical outcomes were determined based on the most recent follow-up visit, with failure defined as symptomatic, laboratory-confirmed CDI. Differences in results and FMT administration were compared between MTOP and our RCT.
Results: Between 2017 and 2019, 32 patients with rCDI have received FMT through MTOP. Of the 32, 26 [81.3% (95% CI 64.3 to 91.5%)] had successful outcomes as of their last follow-up visit (range 1–24 months, median 6 months) compared with 7 out of 16 [43.8% (95% CI 23.1% to 66.9%)] at 3 months in our RCT. Data for MTOP/RCT were: average age of 62.1 y/75.7y; % female, 75.0%/68.8%; number of previous recurrences, 3.0/4.4; previous vancomycin taper 100%/81.3%; Charlson comorbidity index, 2.7/5.3; PPI, 3.1%/43.8%; immunosuppression, 12.5%/0%. FMT administration details are summarized in Table A066.
|FMT||50 g in 300 mL||50 g in 500 mL|
|No. of FMT||Three spaced over a week||One|
|Bowel preparation||Pico-Salax before first FMT||None|
|Timing of FMT relative to last recurrence||Distant||Immediately post|
|Vancomycin pre-treatment||Maintenance dosing||Treatment dosing|
FMT = fecal microbiota transplantation; MTOP = Microbiota Therapeutics Outcomes Program; RCT = randomized clinical trial
Conclusions: FMT treatment of rCDI through MTOP is associated with 81.3% success, an increase compared with 43.8% in our RCT. Which of the differences in FMT administration between MTOP and our RCT is most responsible for this success requires further study.
Eric Bhatti1, Eric Eckbo2, Paula Mahon3, S Rod Rassekh3, Lisa Evans3, Lisa Jacques3, Kiera Bailie3, Susanna Piasecki2, Peter Tilley2, David M Goldfarb2
1University of Ottawa, Ottawa, Ontario, Canada; 2Department of Pathology and Laboratory Medicine, BC Children’s Hospital/University of British Columbia, Vancouver, British Columbia, Canada; 3Division of Hematology/Oncology/BMT, BC Children’s Hospital, Vancouver, British Columbia, Canada
Objectives: Accurate detection of blood stream infections (BSI) is critical for effective treatment and management of a variety of childhood infections. We sought to establish the impact of implementing a weight-based pediatric blood culture collection protocol at a tertiary care pediatric hospital on the mean volume of blood collected and the overall BSI pathogen yield.
Methods: Education sessions were held with nurses and phlebotomy staff campus-wide prior to implementation of the new protocol. Laminated lanyard cards were distributed that summarized the new collection guidelines. Mean blood volume collected was defined as total blood volume collected for blood culture in a 24-hour period from a single patient. The volume was determined by weighing blood culture bottles before and immediately after collection with a calibrated scale. BSI pathogen yield (number of collections positive for ≥1 pathogen/total number of collections) was compared for the year pre- versus post-protocol implementation.
Results: For the mean blood volume evaluation, there were 139 collections evaluated in the pre-period versus 110 collections in the post-period. Mean blood volume collected per kilogram patient weight increased by 165% (0.26 mL/kg versus 0.69 mL/kg, p <0.00001). For the pathogen yield evaluation, a total of 6,605 blood culture collections from 2,986 patients were evaluated from the pre-implementation phase (July 1, 2017, to June 20, 2019), and 6,254 blood culture collections from 3,012 patients were evaluated in the post-implementation phase (July 8, 2018, to July 7, 2019). Total pathogen yield per collection increased 22% for all collections (231/6,605 versus 269/6,254, p = 0.02) (see Figure A067). The largest increases in pathogen yield were seen in the Oncology/Hematology wards (44% increase), followed by the Emergency Department (38% increase).
Conclusions: Implementation of a weight-based institutional blood culture collection protocol was associated with increased blood volume collection and improved BSI pathogen yield.
Antoine Corbeil1,2, Gautham J Nair1, Zeyad Elias1, Bryn Hazlett1, Aimee Paterson1, Angel Li1, Allison McGeer1,2, Tony Mazzulli1,2, Susan M Poutanen1,2
1University Health Network/Sinai Health, Department of Microbiology, Toronto, Ontario, Canada; 2University of Toronto, Department of Laboratory Medicine and Pathobiology, Toronto, Ontario, Canada
Objectives: Candida auris is a global threat. Screeningis being implemented in many hospitals, yet laboratory procedures have not been standardized. Our objective was to determine the optimal C. auris screening culture procedure.
Methods: Eleven C. auris and 23 non-C. auris (C. albicans, C. lusitaniae, C. krusei, C. glabrata, C. tropicalis, C. duobushaemulonii, C. haemulonii, C. guillermondii, Kodamea ohmeri) obtained from clinical isolates and the US AR Isolate Bank were inoculated into residual nasal-axillary-groin-perineum swabs at a final concentration of 104 CFU/mL. Serial experiments compared six key parameters: inoculum volume (30/150/300 µL), agars (BBL CHROMagar Candida™ [BD] versus Brilliance Candida™ [ThermoFisher] versus a custom dulcitol-based selective agar [ThermoFisher]), incubation temperature (37/40/42˚C), incubation time (1–14 days), direct agar inoculation versus broth enrichment with Auris Enrichment Broth (ThermoFisher), and broth daily inversion versus continuous shaking.
Results: Of all directly inoculated agars using all inoculum volumes and incubation times, 11/11 (100%; 95% CI 70%–100%) C. auris isolation was only achieved using the custom agar with minimal to absent growth of non-C. auris; all were isolated with 30 µL inoculum after 3/4/6 days of incubation at 37/40/42˚C, respectively. Auris enrichment broth using continuous shaking (but not daily inversion) for 2 days at 37˚C or 3 days at 42˚C also achieved 100% yield using either CHROMagar (fastest growth at 24h post-subculture) or the custom agar but not Brilliance; breakthrough with non-auris Candida species was greater than with custom agar direct inoculation.
Conclusion: Direct inoculation using 30 µL onto the custom dulcitol-based selective agar at 37˚C provided 100% (95% CI 70–100%) C. auris isolation within a 3-day turnaround time and with minimal to absent growth of non-C. auris. Auris enrichment broth onto CHROMagar at 37˚C provided similar results and turnaround-time, but it required more manipulations, necessitated the use of a shaker, and had greater breakthrough.