Background: Carbapenem resistance in gram-negative organisms is an important public health problem. The CDC conducted Sentinel surveillance in 2018–2019 to characterize these organisms from 9 facilities in 9 different states. Methods: Carbapenem-resistant Enterobacterales (CRE), Pseudomonas aeruginosa (CRPA), and Acinetobacter spp (CRA) obtained from clinical samples of patients in acute-care or long-term care facilities were submitted to the CDC. Identification was confirmed using matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF), and antimicrobial susceptibility testing (AST) was performed via broth microdilution for 27 antibiotics. All confirmed CRE and CRPA were tested for carbapenemase production (CP) using the modified carbapenem inactivation method (mCIM). The isolates that were mCIM-positive were assessed by real-time PCR for presence of blaKPC, blaNDM, blaVIM, and blaIMP. CP-CRE were also assessed for blaOXA-48-like. All confirmed CRA were tested for the same genes as CRPA and blaOXA-23–like, blaOXA-24/40-like, blaOXA-58–like, and blaOXA-235–like genes. Difficult-to-treat resistance (DTR) was defined as resistance to all β-lactams (excluding newer β-lactam combination agents) and quinolones tested. Results: The CDC confirmed 208 CRE, 161 CRPA, and 94 CRA. Table 1 summarizes AST results for a selection of drugs. We identified 112 (53.8%) mCIM-positive CRE and 6 (3.7%) mCIM-positive CRPA. The PCR results are summarized in Table 2. One mCIM-positive and PCR-negative isolate was positive in a metallo-β-lactamase screen. Conclusions: Resistance among CRE and CRPA to newer β-lactam combination agents was detected. Options for treating CRA are limited. Of 112 CP-CRE, 85.7% harbored blaKPC; CP-CRPA were rare (3.7%); and most CRA harbored blaOXA-23-like (55.3%) or blaOXA-24/40-like (30.9%). Whole-genome sequencing is planned to better understand gene variants, sequence types, and additional resistance markers present among the isolates.

Funding: None

Disclosures: None

Hospital epidemiologists, infection preventionists, and antimicrobial stewards are integral to the pandemic workforce. However, regardless of pandemic surge or postsurge conditions, their workload remains high due to constant vigilance for new variants, emerging data, and evolving public health guidance. We describe the factors that have led to burnout and suggest strategies to enhance resilience.

Background: The IDSA recommends that when determining empiric treatment for a given patient, clinicians should consider previous organisms and associated antibiotic susceptibility data in the prior 6 months when multidrug-resistant/extended-spectrum β-lactamase gram-negative rods (MDR /ESBL GNR) is suspected. The presence of MDR GN in cultures on previous hospitalization or in surveillance cultures leads to use of broader-spectrum empiric antibiotic in subsequent admissions. We sought to determine whether the presence of an MDRO in 1 hospital admission dictates the need for broader-spectrum antibiotic use in subsequent hospitalizations when a different site is considered the source of infection. Methods: The study was conducted at a 700-bed tertiary-care level I trauma center. Infection control records were reviewed for patients aged 21–99 years admitted between July 1, 2013, through January 31, 2014, who had any ESBL or MDR GNR at any site including surveillance culture. The charts were surveyed until January 31, 2016, for ESBL or MDR gram-negative organisms at any site during subsequent hospitalizations. The CDC definition of MDR/ESBL was applied. Results: Of the 50 people followed, 34 patients (68%) had ESBL/MDR recovered at a single sentinel site, 6 patients (12%) had ESBL/MDR in multiple sites, and in 10 patients (20%) had ESBL/MDR only in surveillance cultures during their primary hospitalization. In 34 patients with a single sentinel site MDRO/ESBL in their primary hospitalization, 16 (47%) were identified in urine, 13 (38%) were identified by bronchoalveolar lavage (lung), 4 (12%) were identified on skin or MS, and 1 (3%) was identified at another site. When lung and urine were the sole sites of recovery of MDRO/ESBL in primary hospitalization, respectively, 3 patients (23%) and 2 patients (12.5%) subsequently developed MDROs in a secondary site on subsequent admissions. Overall, 10 bacteremia episodes occurred in 8 patients among 189 total admissions. MDRO/ESBL Klebsiella spp were identified in the cultures of 5 patients (50%), MDRO Acinetobacter spp were identified in the cultures of 3 patients (30%), and ESBL E. coli was identified in the cultures of 2 patients (20%). The organism causing bacteremia was present at a different sites during the same admission in only 3 (30%) of these cases: 2 were cultured from urine and 1 was cultured from a pulmonary source. Conclusions: The presence of MDRO at 1 site in a previous hospitalization may not be sufficient evidence to justify very broad antibiotics for patients during subsequent admissions, especially when another site is considered the source of infection. More studies are needed on natural history of MDRO in different patient populations with different risk factors.

Funding: None

Disclosures: None

Background: Persistent colonization with resistant gram-negative bacteria (R-GNB) increases risk of clinical infection and intra-facility transmission among nursing home (NH) patients. Limited data exist on the roles of age and function on duration of R-GNB colonization. Methods: Secondary data analysis was performed from a cohort study of patients admitted to 6 Michigan NHs between November 2013 and May 2018. Swabs obtained upon enrollment, day 14, day 30, then monthly until NH discharge from 6 anatomical sites were cultured for GNB. R-GNB were defined as resistant to ciprofloxacin, ceftazidime, or imipenem. Positive R-GNB culture from a single visit followed by negative cultures for the same organism from ≥2 subsequent visits were defined as transient R-GNB colonization. All other patients with positive R-GNB cultures from multiple visits were considered persistently colonized. Demographic data, antibiotic use, device use, and physical self-maintenance scales (PSMSs) were obtained upon enrollment. Characteristics were compared between patients with transient versus persistent R-GNB and uncolonized patients using multinomial logistic regression. Results: We recruited 896 patients (median age, 75 years; 41% male; 46% nonwhite) and followed them for 2,437 total visits. Of 896 patients, 407 (45.4%) were colonized with ≥1 R-GNB during their stay. Of 171 patients with ≥ 2 follow-up visits after R-GNB detection, 94 (55%) remained persistently colonized with the same R-GNB (Table 1). Escherichia coli (30%) and Proteus mirabilis (22%) were the most frequently identified R-GNB. The most common anatomical colonization sites were perirectal (368 [24.3%] of 1,147) groin (340 [14.3%] of 2,046), and hands (115 [4.8%] of 2283). Compared to uncolonized patients, patients with persistent (1.09; 95% CI, 1.00–1.19, P = .048) and transient R-GNB colonization (1.13; 95% CI, 1.04–1.23; P = .003) had lower PSMS (Tables 2 and 3). Compared to uncolonized and transiently colonized patients, patients with persistent R-GNB colonization had prolonged lengths of NH stay (1.01; 95% CI, 1.00–1.01; P = .003). Conclusions: R-GNB colonization in vulnerable NH patients is common (407 [45.5%] of 896 and often persistent (94 [55%] of 171 patients with sufficient follow-up to assess persistence). Patients with persistent R-GNB had lower functional status, longer LOS, and higher readmission rates than those without. R-GNB decolonization should be investigated as a strategy to potentially improve outcomes among NH patients.

Funding: None

Disclosures: None

Background: Carbapenem-resistant Enterobacterales (CRE) are an increasing threat to patient safety but only a small percentage of CRE identified are NDMs. Since 2018, clinical CRE isolates have been submitted to the Ohio Department of Health for sequencing and NDM cases have notably increased since that time. Candida auris is an emerging pathogen with similar risk factors for colonization as CRE. Methods: A point-prevalence study was initiated after an index patient was identified with NDM CRE infection or colonization during their inpatient admission. Two patient populations were included in the study: current patients on the same unit as the index patient and currently hospitalized patients who overlapped on any unit with the index patient for at least 72 hours. Patients had perirectal screening for CRE (via PCR) and axilla or groin screening for C. auris (via Xpert Carba-R Assay). Patients were excluded if they had been discharged, expired, or refused testing. Results: We completed 5 point-prevalence studies from March 21, 2021, to October 15, 2021. The index patients were admitted at different times and across 2 campuses including medical, cardiac, and surgical ICUs as well as medical-surgical and inpatient rehabilitation units. Moreover, 3 species of NDM were identified from urine and 2 species were identified from bronchoalveolar lavage: Enterobacter hormaechei, Citrobacter freundii, and Enterobacter cloacae complex. C. freundii and E. cloacae complex both had dual mechanisms of NDM and KPC. Although some of the index patients overlapped temporally within the health system, none overlapped in the same unit or building. None of the patients had recently received health care outside the United States, although 1 patient had emigrated from Togo >5 years prior and 4 had had prior local healthcare exposure within 12 months of admission. Also, 147 patients were identified for screening; 105 consented, 32 declined, and 10 were excluded due to being discharged, deceased, or unable to consent. Inpatient point-prevalence screening tests for all patients tested (n = 105) were negative for NDM CRE and C. auris. Conclusions: Despite an increase of inpatients with NDM CRE, evidence of patient-to-patient transmission was not identified, likely resulting from adherence to standard precautions. The diversity of species and lack of international travel suggests that these patients likely acquired NDM CRE from a local reservoir in the community or healthcare settings. Given the continued increase in NDM CRE without traditional risk factors, it is critical for hospitals and public health agencies to collaborate to identify these organisms and that they develop surveillance programs to clarify risk factors for colonization.

Funding: None

Disclosures: None

Background: Prosthetic joint infections (PJIs) are costly and cause increased morbidity and mortality for patients. Staphylococcus epidermidis is a common cause of both early postoperative and late-presenting PJIs. Although S. epidermidis is a normal part of the human skin microflora, its ability to form biofilm on implanted medical devices make it an important causative pathogen of PJIs. We investigated genetic, epidemiologic, and environmental factors contributing to S. epidermidis PJIs by performing whole-genome sequencing and clinical epidemiologic investigation of isolates collected from infected patients between 2017 and 2020. Methods: Patients with S. epidermidis isolated from a prosthetic joint that was placed at our orthopedic specialty hospital were identified using the microbiology laboratory records and electronic medical records. Whole-genome sequencing and single-nucleotide polymorphism (SNP)–based clonality analyses were performed using the epiXact service at Day Zero Diagnostics. These analyses included species identification, in silico MLST typing, phylogenomic analysis, as well as genotypic assessment of the prevalence of specific antibiotic resistance genes, virulence genes, and other relevant genes. For clonal isolates, additional reviews of surgical history and clinical data were performed. Results: In total, 62 S. epidermidis joint isolates were identified from 46 patients. Among these isolates, 52 were of sufficient purity to be used for genomic analysis (Fig. 1). A number of genes appeared in every isolate including sepA, smr, cap, sesB, sesG, and embp. Also, 6 S. epidermidis samples had a discrepancy between phenotypic resistance to oxacillin and the presence of the mecA resistance gene. We also identified 6 distinct clusters of isolates, all of which had SNP distances <10 base pairs (Fig. 2). Each cluster consisted of 2–4 patients. Cluster isolates accounted for 29.8% of all S. epidermidis prosthetic joint isolates. Most clonal isolates occurred in patients who were heavily exposed to different healthcare settings. Further epidemiologic investigation showed that some of these clonal isolates had ties to aspirations or procedures, whereas no clear connection could be determined for others. Conclusions:S. epidermidis isolated from clinical prosthetic joint samples contains a high degree of genetic resistance, including a mismatch between presence of mecA and phenotypic oxacillin resistance and genetic propensity for chlorhexidine resistance. Mupirocin resistance was not observed. Of all isolates, 29.8% belonged to multiple clusters, confirming hospital spread of this commensal organism in some cases.

Funding: None

Disclosures: None

Antibiotic stewardship initiatives usually occur in the inpatient setting and should be optimized during transitions of care. In this study, we assessed the appropriateness of oral antibiotic treatment duration at the time of discharge from our institution based on national guidelines and clinical parameters for common infections.

Background:Staphylococcus aureus is a common cause of healthcare associated infections and is associated with high mortality. S. aureus colonization of skin and mucosa contributes to its pathogenesis. Universal S. aureus decolonization reduces methicillin-resistant S. aureus (MRSA) and other bloodstream infections among ICU patients. However, universal decolonization in acute-care settings has not shown a similar benefit. We describe a targeted decolonization protocol implemented at a large academic hospital across acute-care and intensive care settings. We assessed the impact of decolonization on S. aureus–related infections. Methods: Adults admitted in 2018–2019 to the medicine, oncology, transplant, and ICU services were screened for S. aureus colonization using nasal swabs for MRSA/MSSA by culture. Those with S. aureus detected underwent decolonization with 5 days of chlorhexidine 2% baths and mupirocin intranasal ointment. Decolonization was considered complete if given for 5 days. The primary outcome was S. aureus invasive infection from hospital day 3 until discharge, defined by positive clinical cultures from sterile sites. Secondary outcomes included 30-day readmission and 30-day mortality. The control population was patients with negative MRSA/MSSA nasal screening in the same hospital units. Results: In total, 4,465 (23%) of 19,065 screening tests were positive for MSSA (75%) or MRSA (25%). The median age was 69 years (IQR, 56–80), and the median length of stay (LOS) was 6 days (IQR, 4–10). Among patients with LOS ≥3 days, 541 (16%) completed decolonization and 2,161 (64%) received no decolonization. The rate of complete decolonization increased to 35% among those with LOS ≥ 7 days. In total, 802 screened patients developed invasive S. aureus infections. Of 4,437 colonized patients, 536 (12%) had invasive infections, compared with 265 (2.1%) invasive infections in 12,917 noncolonized patients. Among patients with S. aureus colonization, 24% of decolonized patients developed invasive infection and 13% of patients who were not decolonized developed invasive infection. Rates of 30-day readmission and mortality were 28% and 10%, respectively, among fully decolonized patients, versus 20% and 6.6% among those receiving no decolonization. Conclusions: These data provide an assessment of the efficacy of a targeted screening and decolonization program. Although decolonization did not reduce rates of invasive infection or secondary outcomes, further analysis is needed. Patients with longer lengths of stay are more likely to receive full decolonization but are also at higher risk of invasive infection, which may contribute to our unexpected results.

Funding: None

Disclosures: None

Objective: We evaluated the impressions and perceived workflow consequences following installation of a touch-free aseptic stethoscope barrier dispenser in the clinical environment. Methods: Beginning in 2020, we conducted a volunteer survey of aseptic stethoscope diaphragm barrier (AseptiScope, San Diego, CA) users in multiple departments at 7 US healthcare facilities. A 10-question survey was presented on an iPad near the aseptic barrier dispenser, which was usually located in the patient exam room, to be available immediately after the practitioner completed their examination, which included the use of the stethoscope barrier. This evaluation was considered a quality improvement project and was exempt from institutional review board approval. For this analysis, only 1 survey per practitioner was included. Results: Overall, 147 surveys were obtained from 7 institutions geographically distributed across the United States, immediately after placement of the DiskCover system in the patient care environment. Responses were generally positive and included ease of use (95.2% rated easy or very easy), comparison to a disposable stethoscope (97.9% as similar to, improved over, or significant improvement), workflow changes (53.7% improvement, 97.3% no impact, or improved), and perceived effect on patient safety (90.3% felt that patient safety was improved or significantly improved). Conclusions: The use of a touch-free aseptic stethoscope barrier system was reported to be easy to use, superior to a disposable stethoscope, and an improvement to practitioner workflow and perceived patient safety.

Funding: AseptiScope, Inc.

Disclosures: None

Background: Urine testing is one of the more frequently ordered diagnostic tests among hospitalized patients. Many hospitals have implemented urinalysis with reflex culture (UARC) as a method of diagnostic testing stewardship to guide appropriate use of urine testing. Isolated urine culture, or urine culture without preceding urinalysis, is the most appropriate diagnostic test for patients who are neutropenic, pregnant, or those about to undergo an invasive urologic procedures. This testing is often used beyond these indications in hospitals though, potentially leading to overdiagnosis of UTI and overtreatment of asymptomatic bacteriuria. Methods: We compared outcomes in the preimplementation period (December 2018–November 2019) to those in the postintervention period (December 2019–October 2020) at an academic medical center. The intervention was the addition of an indication selection (ie pregnancy, neutropenia, etc) to the isolated urine-culture order in the electronic medical record (EMR). The primary outcomes were isolated urine culture rate per 1,000 patient days and urine-culture positivity. Our exploratory analysis included a review of selected indications after the intervention was implemented and a chart review of a subset of these tests for appropriateness. The primary analysis was performed using interrupted time-series negative binomial regression. Results: There was no significant change in isolated urine-culture rates after the intervention (11.18 cultures per 1,000 patient days before the intervention versus to 7.75 cultures per 1,000 patient days after the intervention; P > .90), and there were as no significant pre- or postintervention trends. We detected no significant change in isolated urine-culture positivity: 26.9% before the intervention versus 26.7% after the intervention (P > .90). These results are shown graphically in Fig. 1. In the exploratory analysis, of 661 isolated urine-culture tests ordered in the postintervention period, the indication for testing was left blank in 71.9% of tests. The other most common reasons for testing included other (16%), pregnancy (5.7%), and neutropenia (4.4%). In the 100 tests reviewed for appropriateness, only 8% had a documented diagnosis corresponding with the selected indication for testing. Discussion: The addition of an indication selection for isolated urine culture testing did not change the rates of culture ordering or the culture’s subsequent likelihood of positivity. In the exploratory analysis, most providers were incorrectly selecting this testing rather than UARC as prompted. Next steps could potentially be removing the “other” category and requiring a selected answer or requiring approval from stewardship team prior to ordering. Continued education of providers is paramount to the appropriate use of diagnostic testing.

Funding: None

Disclosures: None

We estimated the predictive value of methicillin-resistant Staphylococcus aureus (MRSA) nasal polymerase chain reaction (PCR) for blood, bone, and soft-tissue cultures. The specificities were 88.8%, 88.5%, and 92.7% for all cultures, blood cultures, and bone and soft-tissue cultures respectively, and the negative predictive values were 99.3%, 99.8%, and 92.7% respectively.

Background:Candida bloodstream infections (candidemia) have significant mortality and morbidity rates, as well as healthcare cost implications. Emerging multidrug-resistant Candida spp such as Candida auris, as well as increasing resistance among non–albicans species, which are becoming more prevalent, also raise concern. Understanding the epidemiology of this infection could enhance prevention and management efforts. We studied risk factors for candidemia. Methods: This matched case–control study was conducted at a university hospital from December 2019 through May 2021. Cases of candidemia were identified using positive blood-culture results. Controls were matched 5:1 to cases by age, sex, and month and year of admission. Risk factors of interest included total parenteral nutrition (TPN), central venous access (CVA), neutropenia, Clostridium difficile, pancreatic disease, Candida in urine culture, cancer, invasive procedures, H2 blockers, chemotherapy, antibiotic use, immunosuppression, and antifungal use. Bivariate conditional logistic regression models were used to study the association of individual factors with candidemia. Multivariable conditional logistic regression models were performed using factors with a P Results: Overall, 101 patients with candidemia and 505 matched controls were included. In the bivariate analysis, associations were detected between candidemia and TPN, CVA, pancreatic disease, invasive procedures, H2 blocker use, antibiotic use, and antifungal use (all Ps Conclusions: Associations of candidemia with recent antifungal use and pancreatic disease were relatively novel findings. Neutropenia was not an independent risk factor for candidemia in this study. Future directions include further evaluations of previous antifungal use in patients with candidemia to identify opportunities for possible intervention and antifungal stewardship.

Funding: None

Disclosures: None

Antibiotic overuse is common in ambulatory care settings, underscoring the importance of outpatient antibiotic stewardship to ensure safe and effective antibiotic prescription. In response to this need, the Agency for Healthcare Research and Quality (AHRQ) developed the AHRQ Safety Program for Improving Antibiotic Use in Ambulatory Care. The Safety Program successfully assisted 389 outpatient practices across the United States to establish ambulatory antibiotic stewardship. Herein, we have used lessons learned from the AHRQ Safety Program to describe a step-by-step framework to assist practices with establishing antibiotic stewardship in the outpatient setting. Steps include obtaining support from practice leadership; establishing an antibiotic stewardship team; garnering support from practice members; determining how to access antibiotic prescribing data; building communication skills around antibiotic use in the practice; implementing educational content around an infectious syndrome; monitoring antibiotic prescription data; and implementing a sustainability plan.

Background: In March–April 2021, 23 patients at a 906-bed hospital in Delaware had surgical implantation of a bone graft product contaminated with Mycobacterium tuberculosis; 17 patients were rehospitalized for surgical site infections and 6 developed pulmonary tuberculosis. In May 2021, we investigated this tuberculosis outbreak and conducted a large, multidisciplinary, contact investigation among healthcare personnel (HCP) and patients potentially exposed over an extended period in multiple departments. Methods: Exposed HCP were those identified by their managers as present, without the use of airborne precautions, in operating rooms (ORs) during index spine surgeries or subsequent procedures, the postanesthesia care unit (PACU) when patients had draining wounds, inpatient rooms when wound care was performed, and the sterile processing department (SPD) on the days repeated surgeries were performed. We created and assigned an online education module and symptom screening questionnaire to exposed HCP. Employee health services (EHS) instituted a dedicated phlebotomy station to provide interferon-γ release assay (IGRA) testing for HCP at ≥8 weeks after last known exposure. EHS managed all exposed HCP, including nonemployees (eg, private surgeons) via automated e-mail reminders, which were escalated through supervisory chains as needed until follow-up completion. The infection prevention team notified exposed patients, defined as those who shared semiprivate rooms with case patients with transmissible tuberculosis. The Delaware Division of Public Health performed IGRA testing. Results: There were 506 exposed HCP in ORs (n = 100), the PACU (n = 87), inpatient units (n = 140), the SPD (n = 54), and other locations (n = 122); 83% were employed by the health system. Surgical masks and eye protection were routinely used during patient care. All exposed HCP completed screening by December 17, 2021. Furthermore, 2 HCP had positive IGRAs without symptoms or chest radiograph abnormalities, indicating latent tuberculosis infection, but after further review of records and interviews, we discovered that they had previously tested positive and had been treated for latent tuberculosis infection. In addition, 5 exposed patients tested negative and 2 remain pending. Conclusions: This large investigation demonstrated the need for a systematic process that encompassed all exposed HCP including nonemployees and incorporated administrative controls to ensure complete follow-up. We did not identify any conversions related to this outbreak despite high burden of disease in case patients and multiple exposures to contaminated bone-graft material and infectious bodily fluids without respirator use. Transmission risk was likely reduced by baseline surgical mask use and rapid institution of airborne precautions after outbreak recognition.

Funding: None

Disclosures: None