Objectives: The increase in carbapenemase-producing organism (CPO) transmission among hospitalized patients is a growing concern. Studies investigating the transmission of CPO to epidemiologically linked contacts are scarce. We conducted an interim subgroup analysis of the ongoing multicenter household transmission of CPO in Singapore (CaPES-C) study to identify the acquisition rate of CPO among epidemiologically linked contacts of hospitalized CPO patients. Methods: This multicenter prospective cohort study was conducted between January and December 2021. We recruited CPO-positive patients and their epidemiologically linked contacts. Stool samples were collected from the patients at baseline, day 3, day 7, and at weeks 2, 3, 4, 5, 6, 12, 24, 36, and 48. Additionally, a sample was collected at the time of discharge from the hospital. Xpert Carba-R test was used to detect CPO genotypes in the stool samples. In this interim analysis, we calculated the acquisition rate of CPO among the epidemiologically linked hospital contacts of CPO positive patients using Stata version 15 software. Results: We recruited 22 (56.4%) CPO-positive index patients [blaNDM, n = 7 (31.8%); blaIMP, n = 3 (13.6%); blaOXA-48, n = 10 (45.5%), others, n = 2 (9.1%)] and 14 (35.9%) epidemiologically linked hospital contacts. The median age of CPO-positive patients was 72.5 years (IQR, 62–82) and 15 (68.2%) were female. The median age for the epidemiologically linked contacts was 82.5 years (IQR, 70–85) and 4 (28.6%) were female. After 1,082 patient days, 2 (14.3%) epidemiologically linked contacts tested positive for CPO giving an acquisition rate of 1.85 per 1,000 patient days (95% CI, 0.46 – 7.39). One of these participants acquired a concordant genotype (blaOXA-48) at day 7 and the other acquired a discordant genotype (CPO positive index, blaIMP; epidemiologically linked contact, blaNDM) at week 12 of follow-up. Conclusions: This small interim analysis revealed a high conversion rate among epidemiologically linked hospital contacts. A larger study is needed to understand the influence of genotypes, hospital environment, and human behavior on the transmission of CPO in hospitals.

Objectives: The dissemination of Escherichia coli–producing extended-spectrum β-lactamase (ESBL-Ec) is evident in the community. In this population-based spatial analysis, we sought to describe the distribution of ESBL-Ec and to identify predictors of incidence in the community. Methods: The study population was defined as individuals with the ESBL-Ec isolate in Queensland, Australia, from 2010 to 2019. Annual choropleth maps and a global Moran index were constructed to describe ESBL-Ec distribution. Getis-Ord Gi* was performed to identify “hot spots” of statistical significance. Using demographic factors and incidence per postal area from 2016, multivariable analyses with or without spatially structured random effects were performed. Results: In total, 12,786 individuals with ESBL-Ec isolate were identified. The incidence rate increased annually from 9.1 per 100,000 residents in 2010 to 49.8 per 100,000 residents in 2019. The geographical distribution changed from random to clustered in 2014. Hot spots were more frequently identified in the Outback and Far North Queensland, where remote communities and hotter weather are prevalent. Multivariable spatial analysis suggests that communities with higher socioeconomic status (RR, 0.66; 95% CI, 0.55–0.79 per 100 units) and employment in the agricultural industry (RR, 0.79; 95% CI, 0.67–0.95 per 10%) were protective of lower ESBL-Ec incidence. After accounting for multiple demographic factors, the residual, structured, random-effects model indicated that hot spots were still detected in more remote communities but also in several city regions. Conclusions: The change in distribution of ESBL-Ec across Queensland suggests the presence of area-level specific risk factors that enhance spread in the community. Risk factors for spread appear different between remote and city settings, and future research should be tailored to understand the respective area-level risk factors. Factors such as local temperature, antibiotic consumption, and access to services should be validated. Future public health measures to reduce transmission should be focused on the identified hot spots.

Background: COVID-19 is associated with symptoms, clinical findings, and laboratory abnormalities that raise concern for secondary infections. Excess antimicrobial use despite low rates of secondary infections has been reported and presents a continuing challenge for antimicrobial stewardship programs (ASPs), particularly during COVID-19 surges. The objective of this study was to analyze the appropriateness of antimicrobial use in patients with extended antimicrobial therapy during 2 distinct COVID-19 hospital surges. Methods: We conducted an observational, retrospective, cohort study of COVID-19 patients admitted to our 548-bed community teaching hospital between November and December 2021 (ie, the SARS-CoV-2 delta-variant predominant phase) and January–February 2022 (ie, the SARS-CoV-2 omicron-variant predominant phase) and who received antibiotics for >4 days without positive cultures. Demographic and clinical data were obtained from the institutional data warehouse. Infectious diseases–trained researchers evaluated the appropriateness of antimicrobials based on diagnostic and clinical reporting and institutional antimicrobial stewardship guidelines. Patients were considered to have probable secondary bacterial infection if they had 2 of the following symptoms: fever, unexplained leukocytosis, worsening secretions, or hypoxia and/or imaging. The outcomes of interest included confirmed infections and excess antimicrobial days. Categorical and continuous variables were analyzed using χ2 tests, Fisher exact tests, and Mann-Whitney U tests, respectively. Statistical significance was defined as P ≤ .05. Results: In total, 87 patients were included in the study. Moreover, 56 patients were identified in the SARS-CoV-2 delta-variant predominant phase and 31 patients were identified in the SARS-CoV-2 omicron-variant predominant phase. The groups were similar, with higher vaccination rates in the SARS-CoV-2 omicron-variant predominant group (37.5% vs 64.5%; P = .016). Patients in the SARS-CoV-2 omicron-variant predominant group required less mechanical ventilation (39.3% vs 16.1%; P = .025). There were no significant differences in infectious diseases consultation, immunomodulator or remdesivir use, antimicrobials classes prescribed, or antimicrobial days of therapy or duration between cohorts. There were no significant differences in length of stay, 30-day mortality, or 30-day readmissions. Infections were confirmed in 78.6% in the delta-variant group versus 83.9% in the omicron-variant group (P = .55). Pneumonias accounted for 60.7% in the delta-variant group and 40.9%, in the omicron-variant group. Excess antibiotic use occurred in 14.3% of patients in the delta-variant group and in 3.1% of patients in the omicron-variant group (P = .149). There was no significant difference in the duration of inappropriate antimicrobial use between groups in patients without infections: 5 days in the delta-variant group versus 5 days in the omicron-variant group (P = .24). Conclusions: Results demonstrated that most antimicrobial use was appropriate in a challenging patient population lacking positive cultures to guide therapy. Inappropriate antimicrobial utilization occurred demonstrating continued opportunities for our institutional ASP.

Disclosures: None

We analyzed efficacy of a centralized surveillance infection prevention (CSIP) program in a healthcare system on healthcare-associated infection (HAI) rates amid the coronavirus disease 2019 (COVID-19) pandemic. HAI rates were variable in CSIP and non-CSIP facilities. Central-line–associated bloodstream infection (CLABSI), C. difficile infection (CSI), and surgical-site infection (SSI) rates were negatively correlated with COVID-19 intensity in CSIP facilities.

Background: Many clinical guidelines recommend that clinicians should use antibiograms to decide on empiric antimicrobial therapy. However, antibiograms aggregate epidemiologic data without consideration for any other factors that may affect the risk of antimicrobial resistance (AMR), and little is known about an antibiogram’s reliability in predicting antimicrobial susceptibility. We assessed the diagnostic accuracy of antibiograms as a prediction tool for E. coli clinical isolates in predicting the risk of AMR for individual patients. Methods: We extracted microbiologic and patient-level data from the nationwide clinical data warehouse of the Veterans Health Administration (VHA). We assessed the diagnostic accuracy of the antibiogram for 3 commonly used antimicrobial classes for E. coli: ceftriaxone, fluoroquinolones, and trimethoprim-sulfamethoxazole. First, we retrospectively generated facility-level antibiograms for all VHA facilities from 2000 to 2019 using all clinical culture specimens positive for E. coli, according to the latest Clinical & Laboratory Standards Institute guideline. Second, we created a patient-level data set by including only patients who did not have a positive culture for E. coli in the preceding 12 months. Then we assessed the diagnostic accuracy of an antibiogram for E. coli to predict resistance for the isolates in the following calendar year, using logistic regression models with percentages in the antibiogram as dependent variables. We also set 5 stepwise thresholds at 80%, 85%, 90%, 95%, and 98%, and we calculated sensitivity, specificity, and accuracy for each antimicrobial. Results: Among 127 VHA hospitals, 1,484,038 isolates from 704,779 patients were available for analysis. The area under the ROC curve (AU-ROC) was 0.686 for ceftriaxone, 0.637 for fluoroquinolones, and 0.578 for trimethoprim-sulfamethoxazole, suggesting their relatively poor prediction performances (Fig. 1). The sensitivity and specificity of the antibiogram widely varied by antimicrobial groups and thresholds, with substantial trade-offs. Along with AU-ROC, these metrics suggest poor prediction performances when antibiograms are used as the sole prediction tool (Fig. 2). Conclusions: Antibiograms for E. coli have poor performances in predicting the risk of AMR for individual patients when they are used as a sole tool, and their contribution to the clinical decision making may be limited. Clinicians should also consider other clinical and epidemiologic data when interpreting antibiograms, and guideline statements that suggest antibiogram as a valuable tool for decision making in empiric therapy may need to be reconsidered. Further studies are needed to evaluate the contribution of antibiograms when combined with other patient-level factors.

Disclosures: None

Objectives: Over the past 2 years, many infection prevention and control (IPC) resources have been diverted to manage the COVID-19 pandemic. Its impact on the incidence of antimicrobial-resistant organisms has not been adequately studied. We investigated the impact of the pandemic on the incidence of carbapenem-resistant Enterobacterales (CRE) in Singapore. Methods: We extracted data on unique CRE isolates (clinical and/or surveillance cultures) and patient days for 6 public hospitals in Singapore from the carbapenemase-producing Enterobacteriaceae (CaPES) study group database, and we calculated the monthly incidence of CRE (per 10,000 patient days). Interrupted time-series (ITS) analysis was conducted with the pre–COVID-19 period defined as before February 2020, and the COVID-19 period defined as after February 2020. Statistical analyses were performed using Stata version 15 software. Results: From January 2017 to March 2021, 6,770 CRE isolates and 9,126,704 patient days were documented. The trend in CRE monthly incidence increased significantly during the pre–COVID-19 period (0.060; 95% CI, 0.033–0.094; P < .001) but decreased during the COVID-19 period (−0.183; 95% CI, −0.390 to 0.023; P = .080) without stepwise change in the incidence (−1.496; 95% CI, −3.477 to 0.485; P = .135). The trend in monthly incidence rate of CRE clinical cultures increased significantly during the pre–COVID-19 period (0.046; 95% CI, 0.028–0.064; P < .001) and decreased significantly during COVID-19 period (−0.148; 95% CI, −0.249 to −0.048; P = .048) with no stepwise change in the incidence (−0.063; 95% CI, −0.803 to 0.677; P = .864). The trend in monthly incidence rate of CRE surveillance cultures decreased during the pre–COVID-19 period (−0.020; 95% CI, −0.062 to 0.022; P = .341) and the COVID-19 period (−0.067; 95% CI, −0.291to 0.158; P = .552) without stepwise change in the incidence (−1.327; 95% CI, −3.535 to 0.881; P = .233). Conclusions: The rate of CRE in clinical cultures decreased during COVID-19 but not the rate in surveillance cultures. Further studies are warranted to study the impact of COVID-19 on CREs.

(DUPLICATE DELETED)

De-implementation is defined as the process of discontinuing, removing, reducing, or replacing a harmful, ineffective, or low-value clinical practice or intervention. The goal of de-implementation strategies is to minimize patient harm, maximize use of resources, and reduce healthcare costs and inequities. Both antibiotic and diagnostic stewardship programs focus on reducing low-value interventions (tests or antimicrobials). Stewardship interventions commonly involve de-implementation and deprescribing strategies. This commentary explores unique aspects of deimplementing low-value testing and unnecessary antimicrobial use, similarities between de-implementation and stewardship approaches, multilevel factors that impact de-implementation, and opportunities for future research.

Background: Nationally, >30% of all outpatient antibiotics are unnecessary or inappropriate, and only 52% of outpatients with sinusitis, otitis media, or pharyngitis receive recommended first-line antibiotics. The Minnesota All Payer Claims Database (MN APCD) collects medical claims, pharmacy claims, and eligibility files from private and public healthcare payers. We analyzed claims to describe overall and firstline antibiotic prescribing for acute bronchitis, adult acute sinusitis, and pediatric patients. Results: We analyzed 3,502,013 respiratory events from 1,612,501 members. Acute bronchitis accounted for 179,723 events (5.1%), acute sinusitis accounted for 236,901 adult events (10%), and otitis media accounted for 232,226 pediatric events (19%). Also, 73,385 bronchitis diagnoses (~40%) had no associated antibiotic. Antibiotics were associated with 199,445 adult sinusitis events (84.2%), of which 89,386 (44.8%) were firstline antibiotics, and 190,962 pediatric otitis media events (82.2%), of which 126,859 (66.4%) were firstline antibiotics. Common antibiotic classes used when a firstline drug was not selected were macrolides (28.9%) and tetracyclines (26.8%) for adult acute sinusitis and cephalosporins (61.4%) and macrolides (30.6%) for pediatric otitis media. Compared to the least vulnerable quartile, the most vulnerable social vulnerability index (SVI) quartile had lower odds of receiving firstline antibiotics for adult acute sinusitis if antibiotics were prescribed (OR, 0.90; 95% CI, 0.87–0.94) and higher odds of receiving firstline antibiotics for pediatric otitis media if antibiotics were prescribed (OR, 1.16; 95% CI, 1.12–1.21). Conclusions: Improvement is needed in avoiding antibiotics for acute bronchitis and selecting firstline drugs for sinusitis and otitis media. Additional analyses adjusting for demographic, geographic, and prescriber factors are planned to better understand differences in prescribing appropriateness among Minnesotans.

Disclosures: None

Objectives: Although antimicrobial resistance (AMR) disproportionately affects low- and middle-income countries (LMICs), primary clinical data on AMR burden from LMICs are lacking, particularly from the Pacific Islands. We adapted recent World Health Organization methodology to measure the impact of third-generation cephalosporin (3GC) resistance on mortality and excess length of hospital stay among inpatients with Enterobacterales bloodstream infection (BSI) in Fiji. Methods: We conducted a prospective cohort study of inpatients with Enterobacterales BSIs at Colonial War Memorial Hospital in Suva. We collected demographic, clinical, and microbiological data, and we stored bacterial isolates for confirmatory testing and molecular genomics in Melbourne, Australia. We employed cause-specific Cox proportional hazards models to estimate the effect of 3GC-resistance on hazard of in-hospital mortality and discharge alive (competing outcomes), and we used multistate modelling to estimate the excess length of hospital stay associated with 3GCR. Results: From July 2020 to February 2021, we identified 162 consecutive Enterobacterales BSIs, and 66 (40.7%) were 3GC resistant. The crude mortality rates for patients with 3GC-susecptible and 3GC-resistant BSIs were 16.7% (16 of 96) and 30.3% (20 of 66), respectively. Also, 3GC resistance was not associated with either in-hospital mortality (aHR, 1.67; 95% CI, 0.80–3.49) or discharge alive (aHR, 0.75; 95% CI, 0.50–1.12). However, patient comorbidities and acuity of illness were associated with in-hospital mortality. Furthermore, 3GC-resistance was associated with an increased length of stay of 2.6 days (95% CI, 2.5–2.8). Overall, 3GC-resistance was more common among patients with hospital-associated than community-acquired infection, but genomics did not identify clonal transmission. Conclusions: Among patients with Enterobacterales BSIs, mortality was relatively high, and 3GC resistance was common. Also, 3GC resistance was associated with increased hospital length of stay but not with in-hospital mortality after adjusting for potential confounders. Accurate estimates of the burden of AMR are important, especially from LMICs. Such knowledge can inform policy decisions, guide allocation of limited resources, and assist the evaluation of future interventions to address AMR.

Objectives: In Vietnam, the burden of healthcare-associated infections (HAIs), especially by multidrug-resistant organisms (MDROs), can be greater at national hospitals where a high number of severe patients from lower-tier hospitals are received for treatment. To reduce MDRO and HAI incidences at the tropical diseases intensive care unit (ICU) of Cho Ray Hospital, the final line of treatment for 20 southern provinces of Vietnam, a comprehensive infection prevention bundle was developed and implemented in 2019. In this retrospective study, we evaluated the effectiveness of this intervention in preventing MDRO infections and HAIs among patients at risk. Methods: The infection prevention bundle included elements to improve administration controls, environmental controls, and personal protective equipment usage. The bundle was implemented via training and active monitoring. Medical data, such as microbiology results, length of hospital stay, treatment cost of all patients admitted to the targeted ICU, and data on adherence to the bundle elements, were collected via routine monitoring from July to December 2019. These data were reviewed and analyzed. An independent 2-sample t test was used to calculate the significance of the differences in MDRO and HAI rates before and after the intervention. Results: The mean number of MDRO infections decreased significantly after implementation of the infection prevention bundle (7.0 vs 3.3; P = .011). HAI and ventilator-associated pneumonia (VAP) rates also decreased significantly (5.9 vs 3.7; P= .013 and 20.4 vs 13.7; P=0.047, respectively). The mean total treatment cost per patient was reduced by 1.8 million VND (US $76.76). Bundle-element adherence was high throughout the intervention period, ranging from 72.7% (putting MDRO sign on beds) to 100% (hand hygiene, cohort patients, environment cleaning). Conclusions: Implementation of an appropriate infection prevention bundle with a high adherence rate by healthcare workers helped to effectively reduce MDRO infection and HAI rates in the tropical diseases ICU.

Background: Patients receiving outpatient parenteral antimicrobial therapy (OPAT) are often medically complex and require carefully tailored treatments to address severe and often concomitant infections. Our objective was to illustrate the heterogeneity in antimicrobials used for patients in OPAT, within and across infection diagnosis groups. Methods: We abstracted electronic health record data regarding day-level treatment into a registry of 2,358 OPAT courses (n = 2,072 unique patients) treated in the University of North Carolina Medical Center OPAT program during 2015–2022 (total, 11,861 person weeks; average, 7 OPAT weeks per patient). We classified infection diagnoses into 10 hierarchical or mutually exclusive categories (eg, bacteremia only, diabetic foot infection (DFI) only, osteomyelitis only) (Fig., vertical axes). Accounting for 64 antimicrobial medications and 520 cocktails administered for at least 1 patient day in our OPAT registry, we also defined 18 hierarchical or mutually exclusive classifications of treatment (eg, “daptomycin alone” or “daptomycin and any other antibiotic(s)” (Fig. key). We conducted 2 stratified analyses to describe the heterogeneity across infection diagnoses with respect (1) to medications used at OPAT initiation (patient as unit of analysis) and (2) to medications used throughout OPAT (person time as unit of analysis, allowing for differential OPAT course to other treatment classifications during follow-up). We present stacked bar charts to visualize the intersection between infection diagnosis and treatment group. Results: Among patients in this OPAT registry, 34.6% had osteomyelitis and/or DFI, 4.8% had bacteremia, and 44.6% had multiple infections (Fig. 1). The most common medications in initial OPAT regimens were vancomycin (30.8% of OPAT patients), ceftriaxone (15.0%), and daptomycin (10.9%). We observed overall similarity between the distribution of treatment groups at initiation compared to cumulative person-time during the OPAT course (Figs. 1 and 2). However, we observed heterogeneity in medications by infection diagnosis (Figs. 1 and 2); for example, vancomycin was used in 39% of osteomyelitis cases but only 14% for endocarditis (Fig. 2). For several infection groups (eg, osteomyelitis, DFI, multiple infections, “other” single infections), no treatment classification exceeded 20% use (Figs. 1 and 2). Conclusions: Day-level data on medication use in this monitored registry of patients provided evidence of heterogeneity in the types of medications used throughout treatment in OPAT, which varies within and across infection diagnoses. These data highlight the need for multilayered ascertainment of medication exposure in this medically complex patient population to inform surveillance for adverse effects and guide comparative effectiveness research for postdischarge antibiotic treatment.

Disclosures: None

Objectives: Carbapenem-resistant Enterobacteriaceae (CRE) is a multidrug-resistant gram-negative bacteria (MDR-GNB) that is rapidly emerging as a life-threatening nosocomial disease in many countries. We sought to identify the risk factors associated with mortality for carriage of CRE in patients at a tertiary-care teaching hospital. Methods: A retrospective observational study was conducted between January 2020 to December 2021 in a tertiary-care teaching healthcare facility, University Malaya Medical Centre in Malaysia. The study included all inpatients aged ≥18 years who had a CRE infection or were colonized during the study period. The genotype was identified by polymerase chain reaction (PCR). Statistical analysis of data including a multivariate logistic regression analysis was conducted using SPSS version 23.0 software. Results: In total, 176 cases of CRE (130 infection and 46 colonized) were identified, and the mortality rate was 31.8%. The main sources of CRE were rectal swab (61.9%), blood (11.9%), and respiratory sources (11.9%). Klebsiella pneumoniae (55.7%) was the predominant species, followed by Escherichia coli (21.6%). Among the isolates, 17.7% were non–CPE-CRE and 82.3% were CPE-CRE: NDM (69.3%) and OXA (10.8%). In multivariate analysis, the factors associated with mortality were older age (OR, 1.040; 95% CI, 1.012–1.069), longer length of stay (OR, 0.974; 95% CI, 0.955–0.994), use of central venous catheter (OR, 0.287; 95% CI, 0.094–0.878), and arterial lines (OR, 0.292; 95% CI, 0.095–0.891). Conclusion: Patients with CRE had a high mortality rate. Older age, longer duration of stay, indwelling CVC and arterial line were independent risk factors for death. Infection prevention and control measures to reduce CRE, such as active surveillance, contact precautions, compliance to intravenous catheter care bundles, healthcare worker education, and hand hygiene adherence, should be implemented.

Background: Clostridioides difficile infection (CDI) poses a health burden to patients and a financial burden to hospital systems. Timely identification of CDI patients can reduce the impacts by allowing for prompt treatment and ensuring that proper isolation precautions are in place to prevent spread. It also ensures correct CDI event categorization according to the NHSN. Community-onset (CO) CDI cases are tested on or prior to hospital day 3, and hospital-onset (HO) CDI are tested on or after hospital day 4. The objective of this study was to determine the effectiveness of utilizing an electronic health record (EHR) report to reduce CDI standard infection ratios (SIRs) by identifying potential CDI cases prior to hospital day 4. Methods: From August of 2021 to September 2022, an EHR report was implemented in a 5-hospital healthcare system in the Midwest to identify patients with 3 or more type 6 or 7 stools in a 24-hour period based on Bristol stool chart classification. All inpatients with 3 or more type 6 or 7 stools in 24 hours without an active order for a Clostridioides difficile test were listed. Patients with a laxative in the previous 48 hours, tube feedings without fever or leukocytosis, or a known cause of diarrhea were excluded. The attending provider of the patients meeting criteria were notified with a recommendation to test for C. difficile or provide alternative reason for symptoms. Results: In total, 26 patients were tested for C. difficile using polymerase chain reaction testing. Of those tested, 5 (19.2%) tested positive for C. difficile. There were 13 HO-CDI cases for the healthcare system during this period, for an SIR of 0.351. If the early identified cases were not identified until after hospital day 3, the SIR had the potential to have been 35.6% greater at 0.476. Conclusions: We were able to identify 5 CDI cases prior to hospital day 4 using an early identification report during this 13-month period. Although these cases may have been identified without the use of the EHR report, we were able to obtain a timely CDI diagnosis, potentially limiting the spread of C. difficile and preventing an increase in the CDI SIR by 35.6%. An EHR report to identify patients meeting C. difficile testing criteria may be an effective way to identify CO-CDI prior to HD 4 and thus reduce CDI SIR

Disclosures: None

Objectives: The number of lung transplants is increasing year by year in China and globally. With the widespread use of donation after brainstem death (DBD) donor lungs and donation after circulatory death (DCD) donor lungs, donor-derived infection (DDI) poses a major challenge in lung transplantation. Using donor lungs infected or colonized with carbapenem-resistant Enterobacteriaceae (CRE) may have serious implications in lung-transplant recipients. Currently, traditional microbial culture along with antimicrobial susceptibility testing cannot fully meet the need for rapid and accurate diagnosis of CRE infection in a donor before organ harvest. Methods: The Xpert Carba-R device (Cepheid, Sunnyvale, CA) was used to detect and differentiate Klebsiella pneumoniae carbapenemase (KPC), New Delhi metallo-β-lactamase (NDM), Verona integron-encoded metallo-β-lactamase (VIM), active-on-imipenem (IMP), and OXA-48 carbapenemase genotypes in bronchial lavage fluid from donor lungs before organ harvest. Positive detection of 1 or more of these genotypes indicated a potentially CRE-infected donor lung, and these organs were removed from the lung transplantation cohort. Donor lungs negative for all KPC, NDM, VIM, IMP, and OXA-48 genotypes determined by the Xpert Carba-R device were used for lung transplantation. The incidence of CRE-associated DDI and infection-related complications were compared in the Xpert Carba-R screening group and an historic control group. Results: In this study, 21 donor lungs were tested with the Xpert Carba-R device to detect and differentiate carbapenemase genotypes. Among them, 4 were positive for 1 or more carbapenemase genotypes and were discarded, and the remaining 17 donor lungs showing no carbapenemase gene presence were used for lung transplantation. No CRE-associated DDI occurred in these 17 lung-transplant recipients. Conclusions: Rapid and accurate detection of the carbapenemase gene in donor lungs at the point of care before transplantation using the Xpert Carba-R device reduced the risk of CRE-associated DDI in lung-transplant recipients.

Background: Clostridioides difficile infection (CDI) is a serious healthcare-associated infection responsible for >12,000 US deaths annually. Overtesting can lead to antibiotic overuse and potential patient harm when patients are colonized with C. difficile, but not infected, yet treated. National guidelines recommend when testing is appropriate; occasionally, guideline-noncompliant testing (GNCT) may be warranted. A multidisciplinary group at UNC Medical Center (UNCMC) including the antimicrobial stewardship program (ASP) used a best-practice alert in 2020 to improve diagnostic stewardship, to no effect. Evidence supports use of hard stops for this purpose, though less is known about provider acceptance. Methods: Beginning in May 2022, UNCMC implemented a hard stop in its electronic medical record system (EMR) for C. difficile GNCT orders, with exceptions to be approved by an ASP attending physician. Requests were retrospectively reviewed May–November 2022 to monitor for adverse patient outcomes and provider hard-stop compliance. The team exported data from the EMR (Epic Systems) and generated descriptive statistics in Microsoft Excel. Results: There were 85 GNCT orders during the study period. Most tests (62%) were reviewed by the ASP, and 38% sought non-ASP or no approval. Of the tests reviewed by the ASP, 33 (62%) were approved and 20 (38%) were not. Among tests not approved by the ASP, no patients subsequently received CDI-directed antibiotics, and 1 patient (5%) warranted same-admission CDI testing (negative). Of tests that circumvented ASP review, 18 (56%) ordering providers received a follow-up email from an associate chief medical officer to determine the rationale. No single response type dominated: 3 (17%) were unaware of the ASP review requirement, 2 (11%) indicated their patient’s uncharted refusal of laxatives, 2 (11%) indicated another patient-specific reason. Provider avoidance of the ASP approval mechanism decreased 38%, from 53% of noncompliant tests in month 1 to 33% of tests in month 6. Total tests orders dropped 15.5% from 1,129 during the same period in 2021 to 954 during the study period (95% CI, 13.4%–17.7%). Compliance with the guideline component requiring at least a 48-hour laxative-free interval prior to CDI testing increased from 85% (95% CI, 83%–87%) to 95% (95% CI, 93%–96%). CDI incidence rates decreased from 0.52 per 1,000 patient days (95% CI, 0.41–0.65) to 0.41 (95% CI, 0.32–0.53), though the change was neither significant at P = .05 nor attributable to any 1 intervention. Conclusions: Over time and with feedback to providers circumventing the exception process, providers accepted and used the hard stop, improving diagnostic stewardship and avoiding unneeded treatment.

Disclosures: None

The standardized antimicrobial administration ratio (SAAR) is the metric for reporting antimicrobial use that hospitals will be mandated to use in 2024. We highlight limitations of the SAAR and caution against efforts to use it for public reporting and financial reimbursement. Before the SAAR is ready for public reporting, it needs to include patient-level risk adjustment and antimicrobial resistance data as well as improved hospital location options and revised antimicrobial agent groupings to appropriately reflect and incentivize important stewardship work.