Proposing the "Continuum of UTI" for a Nuanced Approach to Diagnosis and Management of Urinary Tract Infections

PURPOSE

Patients with suspected UTIs are categorized into 3 clinical phenotypes based on current guidelines: no UTI, asymptomatic bacteriuria (ASB), or UTI. However, all patients may not fit neatly into these groups. Our objective was to characterize clinical presentations of patients who receive urine tests using the "continuum of UTI" approach.

MATERIALS AND METHODS

This was a retrospective cohort study of a random sample of adult noncatheterized inpatient and emergency department encounters with paired urinalysis and urine cultures from 5 hospitals in 3 states between January 01, 2017, and December 31, 2019. Trained abstractors collected clinical (eg, symptom) and demographic data. A focus group discussion with multidisciplinary experts was conducted to define the continuum of UTI, a 5-level classification scheme that includes 2 new categories: lower urinary tract symptoms/other urologic symptoms and bacteriuria of unclear significance. The newly defined continuum of UTI categories were compared to the current UTI classification scheme.

RESULTS

Of 220,531 encounters, 3392 randomly selected encounters were reviewed. Based on the current classification scheme, 32.1% (n = 704) had ASB and 53% (n = 1614) did not have a UTI. When applying the continuum of UTI categories, 68% of patients (n = 478) with ASB were reclassified as bacteriuria of unclear significance and 29% of patients (n = 467) with "no UTI" were reclassified to lower urinary tract symptoms/other urologic symptoms.

CONCLUSIONS

Our data suggest the need to reframe our conceptual model of UTI vs ASB to reflect the full spectrum of clinical presentations, acknowledge the diagnostic uncertainty faced by frontline clinicians, and promote a nuanced approach to diagnosis and management of UTIs.

Antimicrobial stewardship knowledge, attitudes, and practices (KAP) among nurses

We performed a knowledge, attitudes, and practice (KAP) survey of bedside nurses to evaluate perceptions of antimicrobial use and aid in the design of nursing-based antimicrobial stewardship interventions. The survey highlighted discrepancies in knowledge and practice as well as opportunities to improve communication with nursing colleagues.

Teams in Transition: Increasing Role of Advanced Practice Providers in Antimicrobial Use and Infectious Diseases Consultation

Background

Advanced practice providers (APPs) have taken on increasing responsibilities as primary team members in acute care hospitals, but the impact of this practice shift on antimicrobial prescribing and infectious diseases (ID) consultation requests is unknown. Here we describe longitudinal trends in antimicrobial days of therapy (DOT) and ID consultation by attributed provider type in 3 hospitals.

Methods

We performed a retrospective time series analysis of antimicrobial use and ID consultation from July 2015 to June 2022 at a major university hospital and 2 community hospitals. We evaluated antimicrobial DOT and ID consultation over time and assessed attribution to 3 groups of providers: attending physicians, trainees, and APPs. We used multinomial logistic regression to measure changes in percentage of DOT and ID consultation across the clinician groups over time using physicians as the referent.

Results

Baseline distribution of antimicrobial DOT and ID consultation varied by practice setting, but all subgroups showed increases in the proportion attributable to APPs. Large increases were seen in the rate of ID consultation, increasing by >30% during the study period. At our university hospital, by study end >40% of new ID consults and restricted antimicrobial days were attributed to APPs.

Conclusions

Hospitals had differing baseline patterns of DOT attributed to provider groups, but all experienced increases in DOT attributed to APPs. Similar increases were seen in changes to ID consultation. APPs have increasing involvement in antimicrobial use decisions in the inpatient setting and should be engaged in future antimicrobial stewardship initiatives.

Racial disparities in Clostridioides difficile testing in three southeastern US hospitals

OBJECTIVE

To analyze Clostridioides difficile testing in 3 hospitals in central North Carolina to validate previous racial health-disparity findings.

METHODS

We completed a retrospective analysis of inpatient C. difficile tests from 2015 to 2021 at 3 university-affiliated hospitals in North Carolina. We calculated the number of C. difficile tests per 1,000 patient days stratified by race: White, Black, and non-White, non-Black (NWNB). We defined a unique C. difficile test as one that occurred in an inpatient unit with a matching laboratory accession ID and on differing calendar days. Tests were evaluated overall, by hospital, by year, and by positivity rate.

RESULTS

In total, 35,160 C. difficile tests and 2,571,850 patient days across all 3 hospitals from 2015 to 2021 were analyzed. The median number of C. difficile tests per 1,000 patient days was 13.85 (interquartile range [IQR], 9.88-16.07). Among all C. difficile tests, 5,225 (15%) were positive. White patients were administered more C. difficile tests (14.46 per 1,000 patient days) than Black patients (12.96; P < .0001) or NWNB race patients (10.27; P < .0001). Black patients were administered more tests than NWNB patients (P < .0001). White patients tested positive at a similar rate to Black patients (15% vs 15%; P = .3655) and higher than NWNB individuals (12%; P = .0061), and Black patients tested positive at a higher rate than NWNB patients (P = .0024).

CONCLUSION

White patients received more C. difficile tests than Black and NWNB patient groups when controlling for race patient days. Future studies should control for comorbidities and investigate community onset of C. difficile by race and ethnicity.

Using clinical decision support to improve urine testing and antibiotic utilization

OBJECTIVE

Urine cultures collected from catheterized patients have a high likelihood of false-positive results due to colonization. We examined the impact of a clinical decision support (CDS) tool that includes catheter information on test utilization and patient-level outcomes.

METHODS

This before-and-after intervention study was conducted at 3 hospitals in North Carolina. In March 2021, a CDS tool was incorporated into urine-culture order entry in the electronic health record, providing education about indications for culture and suggesting catheter removal or exchange prior to specimen collection for catheters present >7 days. We used an interrupted time-series analysis with Poisson regression to evaluate the impact of CDS implementation on utilization of urinalyses and urine cultures, antibiotic use, and other outcomes during the pre- and postintervention periods.

RESULTS

The CDS tool was prompted in 38,361 instances of urine cultures ordered in all patients, including 2,133 catheterized patients during the postintervention study period. There was significant decrease in urine culture orders (1.4% decrease per month; P < .001) and antibiotic use for UTI indications (2.3% decrease per month; P = .006), but there was no significant decline in CAUTI rates in the postintervention period. Clinicians opted for urinary catheter removal in 183 (8.5%) instances. Evaluation of the safety reporting system revealed no apparent increase in safety events related to catheter removal or reinsertion.

CONCLUSION

CDS tools can aid in optimizing urine culture collection practices and can serve as a reminder for removal or exchange of long-term indwelling urinary catheters at the time of urine-culture collection.

Evaluation of an Opt-Out Protocol for Antibiotic De-Escalation in Patients With Suspected Sepsis: A Multicenter, Randomized, Controlled Trial

BACKGROUND

Sepsis guidelines recommend daily review to de-escalate or stop antibiotics in appropriate patients. This randomized, controlled trial evaluated an opt-out protocol to decrease unnecessary antibiotics in patients with suspected sepsis.

METHODS

We evaluated non-intensive care adults on broad-spectrum antibiotics despite negative blood cultures at 10 US hospitals from September 2018 through May 2020. A 23-item safety check excluded patients with ongoing signs of systemic infection, concerning or inadequate microbiologic data, or high-risk conditions. Eligible patients were randomized to the opt-out protocol vs usual care. Primary outcome was post-enrollment antibacterial days of therapy (DOT). Clinicians caring for intervention patients were contacted to encourage antibiotic discontinuation using opt-out language. If continued, clinicians discussed the rationale for continuing antibiotics and de-escalation plans. To evaluate those with zero post-enrollment DOT, hurdle models provided 2 measures: odds ratio of antibiotic continuation and ratio of mean DOT among those who continued antibiotics.

RESULTS

Among 9606 patients screened, 767 (8%) were enrolled. Intervention patients had 32% lower odds of antibiotic continuation (79% vs 84%; odds ratio, 0.68; 95% confidence interval [CI], .47-.98). DOT among those who continued antibiotics were similar (ratio of means, 1.06; 95% CI, .88-1.26). Fewer intervention patients were exposed to extended-spectrum antibiotics (36% vs 44%). Common reasons for continuing antibiotics were treatment of localized infection (76%) and belief that stopping antibiotics was unsafe (31%). Thirty-day safety events were similar.

CONCLUSIONS

An antibiotic opt-out protocol that targeted patients with suspected sepsis resulted in more antibiotic discontinuations, similar DOT when antibiotics were continued, and no evidence of harm.

CLINICAL TRIALS REGISTRATION

NCT03517007.

Optimizing reflex urine cultures: Using a population-specific approach to diagnostic stewardship

BACKGROUND

Clinicians and laboratories routinely use urinalysis (UA) parameters to determine whether antimicrobial treatment and/or urine cultures are needed. Yet the performance of individual UA parameters and common thresholds for action are not well defined and may vary across different patient populations.

METHODS

In this retrospective cohort study, we included all encounters with UAs ordered 24 hours prior to a urine culture between 2015 and 2020 at 3 North Carolina hospitals. We evaluated the performance of relevant UA parameters as potential outcome predictors, including sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). We also combined 18 different UA criteria and used receiver operating curves to identify the 5 best-performing models for predicting significant bacteriuria (≥100,000 colony-forming units of bacteria/mL).

RESULTS

In 221,933 encounters during the 6-year study period, no single UA parameter had both high sensitivity and high specificity in predicting bacteriuria. Absence of leukocyte esterase and pyuria had a high NPV for significant bacteriuria. Combined UA parameters did not perform better than pyuria alone with regard to NPV. The high NPV ≥0.90 of pyuria was maintained among most patient subgroups except females aged ≥65 years and patients with indwelling catheters.

CONCLUSION

When used as a part of a diagnostic workup, UA parameters should be leveraged for their NPV instead of sensitivity. Because many laboratories and hospitals use reflex urine culture algorithms, their workflow should include clinical decision support and or education to target symptomatic patients and focus on populations where absence of pyuria has high NPV.

935. Antibiotic Use (AU) Adjustment by Infection-Related Patient Volume Across a Health System

Background

Benchmarking AU is important to identify opportunities and allocate resources within a health system. Patient level factors such as infection diagnosis codes further refine risk adjustments but have not been more widely adopted because of the burden of accurately collecting and submitting granular data. The goal of this study was to evaluate a novel metric to estimate facility-level infection burden as a potential factor to use in adjustment of AU.

Methods

We conducted a retrospective analysis of hospital administrative data (for calendar year 2020) from 8 hospitals in a single health system using a common electronic health record and coding department. We identified inpatient encounters with an infection-related primary ICD-10 code (I-PDX), based on the health system’s coding department determination and extracted the length of stay (LOS) for each encounter. For any encounter with an I-PDX, the entire LOS was classified as infection-related patient days (IPD). Overall AU in days of therapy (DOT) was adjusted using two novel infection diagnoses denominators. The first was based on proportion of total patient days (PD) attributable to I-PDX encounters (% I-PDX x PD). Since LOS tends to be longer in I-PDX, we also calculated DOT with adjustment for actual extracted IPDs. We then rank ordered study hospitals based on standard DOT / 1,000 PD, NHSN SAAR metrics, and our novel DOT / (% I-PDX x PD) and DOT / 1,000 IPD metrics.

Results

The proportion of I-PDX was highly variable among hospitals, with a system-wide median of 37.27% (range 23.48 - 43.32) (Figure 1). Using DOT / 1,000 patient days for 1 year, Hospital A was the lowest in the system and hospital H was the highest (Figure 2). However, after adjusting for the proportion of patients with I-PDX encounters and IPDs, hospital rank changed considerably, i.e. Hospital H and C respectively ranked lowest and Hospital A was highest.

Conclusion

These novel infection diagnoses PD denominators more closely associated facility level infection burden with AU, for a more refined rank order within the health system. These metrics provide an example of a parsimonious adjustment using patient level data that is already collected at any facility. Next steps might include indirect standardization using PDX categories and other patient level factors readily collected.

Disclosures

Melissa D. Johnson, PharmD, Charles River Laboratories: Grant/Research Support|Entasis: Honoraria|Merck: Grant/Research Support|Pfizer: Grant/Research Support|Scynexis: Grant/Research Support|Theratechnologies: Grant/Research Support|UpToDate: Honoraria Rebekah W. Moehring, MD, MPH, FIDSA, FSHEA, UpToDate, Inc.: Author Royalties Angelina Davis, PharmD, M.S., Merck & Co.: Honoraria.

961. Pandemic Hits: Evaluation of an Antimicrobial Stewardship Program Website for Hospital Communication During the COVID-19 Pandemic

Background

Antibiotic Stewardship Programs (ASPs) assist front-line clinicians in synthesizing emerging data and establishing best practices. Our ASP team directly maintained and edited an internal web application, Duke CustomID^®, to disseminate updated guideline, policy, and drug information during COVID-19. We aimed to describe website engagement and maintenance during the dynamic pandemic period.

Methods

We performed a descriptive, time-series analysis using Google Analytics software to measure engagement with Duke CustomID^® during a 1-year pre-pandemic period through the Omicron surge: January 2019 to March 2022. We measured total page views (or “hits”), COVID-specific page hits, and days requiring COVID-specific page edits by week. Given fluctuations in hospitalization rates, we defined the primary outcome as the rate of hits divided by total hospitalizations. Weekly data were assessed graphically with positive COVID tests and COVID hospitalizations. We used negative binomial regression to quantify the association between COVID hospitalizations and hit rates and to trend engagement over time, adjusted for seasonality. We stratified data by COVID page and calculated a hit/edit ratio.

Results

Engagement with CustomID® increased during the pandemic period, especially during surges (Figure). Hits in the pre-pandemic period were median 1707 (range 1165-2354) per week, and hit rates median 1.95 per hospitalization (range 1.40-2.86). Peaks were observed in March 2020 (hit rate 4.59) and January 2022 (hit rate 3.87). On average, for every 100 COVID hospitalizations, the hit rate increased by 0.08 (0.004-0.16, p=0.04). Engagement slowly increased over the study period (relative rate week 1 versus 170: 1.15, 95% confidence interval 1.02-1.28, p=0.02). COVID page edits per week had a median of 2 (range 0-12). Adult Inpatient Guidelines and COVID Monoclonal Antibody pages had highest use (Table).

Duke CustomID Hits and Maintenance Efforts over the Pandemic

Top: COVID-specific CustomID hits per week (Green), Positive COVID tests per week (Blue) over time

Middle: Total custom ID page hits relative to total hospitalizations per week (teal), COVID hospitalizations (Red)

Bottom: Number of edits to COVID-specific CustomID pages per week, stratified by management pages and drug pages

Several dates of significance are highlighted including the Emergency Use Authorizations (EUA) for remdesivir, the COVID Vaccines, and Paxlovid

Duke CustomID COVID-19 Page Hits and Edits

COVID specific pages on Duke CustomID with total hits, edits, and ratio over the pandemic

Conclusion

Our ASP’s website was a highly utilized, practical tool for disseminating practice-changing information during the pandemic. Use increased over time and especially during surges. An electronic reference customized for local practice and rapidly updated by ASPs offers critical support for front-line clinicians.

Disclosures

Martha B. Adams, M.D., Custom Clinical Decision Support, Inc: Board Member|Custom Clinical Decision Support, Inc: Ownership Interest Richard H. Drew, PharmD MS, American College of Clinical Pharmacists: Publication royalties|Takeda: Advisor/Consultant|UpToDate: publication royalties Rebekah W. Moehring, MD, MPH, FIDSA, FSHEA, UpToDate, Inc.: Author Royalties Michael J. Smith, M.D., M.S.C.E, Merck: Grant/Research Support|Pfizer: Grant/Research Support.

1570. Cumulative Antibiotic Exposure and Risk for Candidemia

Background

Broad-spectrum antibiotic use is a known risk factor for candidemia, but the duration and type of antibiotic exposure associated with greatest risk are not well characterized. Measurements of antibiotic days may be useful to assess cumulative burden of selective pressure on the microbiome and risk for candidemia.

Methods

This retrospective cohort study aimed to quantify the effect of antibiotic exposure on risk for candidemia. The primary outcome was hazard for developing candidemia across an array of antibiotic agents, classes and spectra. We measured antibiotic days of therapy (DOT) for adults admitted to Duke University hospitals 1/1/2016–12/31/2021. We excluded patients with community-onset candidemia, defined as growth of Candida spp. in blood culture collected ≤ 48 hours of admission, because antibiotic exposure prior to arrival was not reliably accessible. Time-to-event analyses were performed using the Nelson-Aalen estimator for modeling cumulative hazard functions to compare the proportion of candidemia observed based on exposure to each antibiotic. Logrank tests were used to evaluate for differences between hazard functions with a pre-specified alpha level of 0.05, and exponential cumulative proportional hazards models were implemented to generalize hazard functions.

Results

During 164,185 encounters in 105,330 unique patients, we identified candidemia in 237 patients. Prior to developing candidemia, cases received a total of 9,604 antibacterial DOT distributed across 46 unique antibiotic agents (Fig. 1) Carbapenems were associated with increased hazard for candidemia compared to beta-lactams (p&lt; 0.005) (Figs. 2–3). There were 57 encounters for candidemia where meropenem was administered with a median of 10 DOT prior to onset of candidemia.

Days of Antibiotic Therapy by Agent and Outcome.

Conclusion

This work represents a novel approach to quantifying antibiotic exposure as a risk factor for candidemia. In an unadjusted model, we identified carbapenems as a high-risk class; additional analysis with adjusted regression models will help contextualize exposure risk with respect to comorbidity and illness severity. This work may serve as a reference for antibiotic stewards as they promote appropriate antibiotic use, including reducing overuse of broad-spectrum antibiotics.

Disclosures

Rebekah W. Moehring, MD, MPH, FIDSA, FSHEA, UpToDate, Inc.: Author Royalties Barbara D. Alexander, MD, Astellas: Advisor/Consultant|HealthtrackRx: Advisor/Consultant|HealthtrackRx: Grant/Research Support|Scynexis: Grant/Research Support|UpToDate: Advisor/Consultant Melissa D. Johnson, PharmD, Biomeme: Licensed Transcriptional Signature for Candidemia|Charles River Laboratories: Grant/Research Support|Entasis Therapeutics: Advisor/Consultant|Merck & Co. Inc: Advisor/Consultant|Merck & Co. Inc: Grant/Research Support|Pfizer, Inc.: Advisor/Consultant|Scynexis Inc.: Grant/Research Support|Theratechnologies: Advisor/Consultant.

1571. Hospital COVID-19 Burden Impact on Inpatient Antibiotic Use Rates

Background

COVID-19 shifted antibiotic stewardship program resources and changed antibiotic use (AU). Shifts in patient populations with COVID surges, including pauses to surgical procedures, and dynamic practice changes makes temporal associations difficult to interpret. Our analysis aimed to address the impact of COVID on AU after adjusting for other practice shifts.

Methods

We performed a longitudinal analysis of AU data from 30 Southeast US hospitals. Three pandemic phases (1: 3/20–6/20; 2: 7/20–10/20; 3: 11/20–2/21) were compared to baseline (1/2018–1/2020). AU (days of therapy (DOT)/1000 patient days (PD)) was collected for all antimicrobial agents and specific subgroups: broad spectrum (NHSN group for hospital-onset infections), CAP (ceftriaxone, azithromycin, levofloxacin, moxifloxacin, and doxycycline), and antifungal. Monthly COVID burden was defined as all PD attributed to a COVID admission. We fit negative binomial GEE models to AU including phase and interaction terms between COVID burden and phase to test the hypothesis that AU changes during the phases were related to COVID burden. Models included adjustment for Charlson comorbidity, surgical volume, time since 12/2017 and seasonality.

Results

Observed AU rates by subgroup varied over time; peaks were observed for different subgroups during distinct pandemic phases (Figure). Compared to baseline, we observed a significant increase in overall, broad spectrum, and CAP groups during phase 1 (Table). In phase 2, overall and CAP AU was significantly higher than baseline, but in phase 3, AU was similar to baseline. These phase changes were separate from effects of COVID burden, except in phase 1 where we observed significant effects on antifungal (increased) and CAP (decreased) AU (Table).

Conclusion

Changes in hospital AU observed during early phases of the COVID pandemic appeared unrelated to COVID burden and may have been due to indirect pandemic effects (e.g., case mix, healthcare resource shifts). By pandemic phase 3, these disruptive effects were not as apparent, potentially related to shifts in non-COVID patient populations or ASP resources, availability of COVID treatments, or increased learning, diagnostic certainty, and provider comfort with avoiding antibacterials in patients with suspected COVID over time.

Disclosures

Melissa D. Johnson, PharmD, Biomeme: Licensed Transcriptional Signature for Candidemia|Charles River Laboratories: Grant/Research Support|Entasis Therapeutics: Advisor/Consultant|Merck & Co. Inc: Advisor/Consultant|Merck & Co. Inc: Grant/Research Support|Pfizer, Inc.: Advisor/Consultant|Scynexis Inc.: Grant/Research Support|Theratechnologies: Advisor/Consultant Angelina Davis, PharmD, M.S., Merck & Co.: Honoraria Sonali D. Advani, MBBS, MPH, FIDSA, Locus Biosciences: Advisor/Consultant|Locus Biosciences: Honoraria|Sysmex America: Advisor/Consultant Ibukun Kalu, MD, Pfizer, Inc.: Institutional support for clinical trial Rebekah W. Moehring, MD, MPH, FIDSA, FSHEA, UpToDate, Inc.: Author Royalties.

190. Performance of urinalysis parameters in predicting significant bacteriuria: Making the case for a population-specific approach to diagnostic stewardship

Background

Clinicians and laboratories routinely use urinalysis (UA) results to help determine if urine cultures and/or antimicrobials are indicated. Yet, the performance of individual UA parameters and common clinical thresholds for action are not well defined and may vary across different patient populations. Our objective was to compare the performance of different UA parameters in predicting significant bacteriuria irrespective of symptoms, and to assess performance of pyuria based on age, sex, and presence of indwelling catheter.

Methods

This retrospective review of UA and urine culture data from the Duke University Health System included all UAs ordered within 24 hours of a urine culture between 2015 and 2020 (no reflex urine cultures included). We defined significant bacteriuria as a urine culture with ≥1 uropathogen growing at ≥100,000 colony forming units/mL. Then, we used this definition to evaluate the performance of relevant UA parameters and result thresholds including sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV). We also combined 18 different UA criteria (as shown in Figure) and used receiver operating characteristic (ROC) curves to identify the top 5 performing models for predicting significant bacteriuria (sensitivity and specificity).

18 Different Combinations of UA Parameters for Predicting Significant Bacteriuria on Urine Cultures

Results

Of 240,195 encounters during the 6-year study period, 38% were outpatient and 62% were inpatient. Twenty-nine percent had a urine culture with significant bacteriuria; 30.7% had a negative urine culture. No single UA parameter had both - high sensitivity and high specificity in predicting bacteriuria. Trace leukocyte esterase and low-level pyuria had a high NPV for significant bacteriuria (Table 1A). Combined UA parameters did not perform better than pyuria alone (Table 1B). The high NPV >=0.90 of pyuria was maintained among most patient age and sex subgroups with the exception of females ≥65 and patients with indwelling catheters (Table 2). Table 1A:

Use of Contact Networks to Estimate Potential Pathogen Risk Exposure in Hospitals

IMPORTANCE

Person-to-person contact is important for the transmission of health care-associated pathogens. Quantifying these contact patterns is crucial for modeling disease transmission and understanding routes of potential transmission.

OBJECTIVE

To generate and analyze the mixing matrices of hospital patients based on their contacts within hospital units.

DESIGN, SETTING, AND PARTICIPANTS

In this quality improvement study, mixing matrices were created using a weighted contact network of connected hospital patients, in which contact was defined as occupying the same hospital unit for 1 day. Participants included hospitalized patients at 299 hospital units in 24 hospitals in the Southeastern United States that were part of the Duke Antimicrobial Stewardship Outreach Network between January 2015 and December 2017. Analysis was conducted between October 2021 and February 2022.

MAIN OUTCOMES AND MEASURES

The mixing matrices of patients for each hospital unit were assessed using age, Elixhauser Score, and a measure of antibiotic exposure.

RESULTS

Among 1 549 413 hospitalized patients (median [IQR] age, 44 [26-63] years; 883 580 [56.3%] women) in 299 hospital units, some units had highly similar patterns across multiple hospitals, although the number of patients varied to a great extent. For most of the adult inpatient units, frequent mixing was observed for older adult groups, while outpatient units (eg, emergency departments and behavioral health units) showed mixing between different age groups. Most units mixing patterns followed the marginal distribution of age; however, patients aged 90 years or older with longer lengths of stay created a secondary peak in some medical wards. From the mixing matrices by Elixhauser Score, mixing between patients with relatively higher comorbidity index was observed in intensive care units. Mixing matrices by antibiotic spectrum, a 4-point scale based on priority for antibiotic stewardship programs, resulted in 6 major distinct patterns owing to the variation of the type of antibiotics used in different units, namely those dominated by a single antibiotic spectrum (narrow, broad, or extended), 1 pattern spanning all antibiotic spectrum types and 2 forms of narrow- and extended-spectrum dominant exposure patterns (an emergency room where patients were exposed to one type of antibiotic or the other and a pediatric ward where patients were exposed to both types).

CONCLUSIONS AND RELEVANCE

This quality improvement study found that the mixing patterns of patients both within and between hospitals followed broadly expected patterns, although with a considerable amount of heterogeneity. These patterns could be used to inform mathematical models of health care-associated infections, assess the appropriateness of both models and policies for smaller community hospitals, and provide baseline information for the design of interventions that rely on altering patient contact patterns, such as practices for transferring patients within hospitals.

Curvularia alcornii Aortic Pseudoaneurysm Following Aortic Valve Replacement: Case Report and Review of the Literature

We report the first case of Curvularia alcornii aortic pseudoaneurysm following bioprosthetic aortic valve replacement in an immunocompetent host. Infection was complicated by septic emboli to multiple organs. Despite aggressive surgical intervention and antifungal therapy, infection progressed. We review the literature on invasive Curvularia infection to inform diagnosis and management.

Impact of Education and Data Feedback on Guideline-Concordant Prescribing for Urinary Tract Infections in the Outpatient Setting

Background

Urinary tract infections (UTIs) are the most common outpatient indication for antibiotics and an important target for antimicrobial stewardship (AS) activities. With The Joint Commission standards now requiring outpatient AS, data supporting effective strategies are needed.

Methods

We conducted a 2-phase, prospective, quasi-experimental study to estimate the effect of an outpatient AS intervention on guideline-concordant antibiotic prescribing in a primary care (PC) clinic and an urgent care (UC) clinic between August 2017 and July 2019. Phase 1 of the intervention included the development of clinic-specific antibiograms and UTI diagnosis and treatment guidelines, presented during educational sessions with clinic providers. Phase 2, consisting of routine clinic- and provider-specific feedback, began ~12 months after the initial education. The primary outcome was percentage of encounters with first- or second-line antibiotics prescribed according to clinic-specific guidelines and was assessed using an interrupted time series approach.

Results

Data were collected on 4724 distinct patients seen during 6318 UTI encounters. The percentage of guideline-concordant prescribing increased by 22% (95% CI, 12% to 32%) after Phase 1 education, but decreased by 0.5% every 2 weeks afterwards (95% CI, –0.9% to 0%). Following routine data feedback in Phase 2, guideline concordance stabilized, and significant further decline was not seen (–0.6%; 95% CI, –1.6% to 0.4%). This shift in prescribing patterns resulted in a 52% decrease in fluoroquinolone use.

Conclusions

Clinicians increased guideline-concordant prescribing, reduced UTI diagnoses, and limited use of high-collateral damage agents following this outpatient AS intervention. Routine data feedback was effective to maintain the response to the initial education.

Impact of Education and Data Feedback on Antibiotic Prescribing for Urinary Tract Infections in the Emergency Department: An Interrupted Time Series Analysis

BACKGROUND

Urinary tract infections (UTIs) are often misdiagnosed or treated with exceedingly broad-spectrum antibiotics, leading to negative downstream effects. We aimed to implement antimicrobial stewardship (AS) strategies targeting UTI prescribing in the emergency department (ED).

METHODS

We conducted a quasi-experimental prospective AS intervention outlining appropriate UTI diagnosis and management across three EDs, within an academic and two community hospitals, in North Carolina, United States. The study was divided into three phases, a baseline period and two intervention phases. Phase 1 included introduction of an ED-specific urine antibiogram and UTI guideline, education, and department-specific feedback on UTI diagnosis and antibiotic prescribing. Phase 2 included re-education and provider-specific feedback. Eligible patients included adults with an antibiotic prescription for UTI diagnosed in the ED from 11/13/18 to 3/1/21. Admitted patients were excluded. The primary outcome was guideline-concordant antibiotic use, assessed using an interrupted time series regression analysis with 2-week intervals.

RESULTS

Overall, 8,742 distinct patients with 10,426 patient encounters were included. Ninety-two percent of all encounters (n=9,583) were diagnosed with cystitis and 8.1% with pyelonephritis (n=843). There was an initial 15% increase in guideline-concordant antibiotic prescribing in Phase 1 compared to the pre-intervention period (incidence rate ratio [IRR] 1.15; 95% confidence interval [CI] 1.03 to 1.29). A significant increase of guideline-concordant prescriptions was seen with every two-week interval during Phase 2 (IRR 1.03; 95% CI 1.01 to 1.04).

CONCLUSIONS

This multifaceted AS intervention involving a guideline, education, and provider-specific feedback increased guideline-concordant antibiotic choices for treat-and-release patients in the ED.

378. Descriptive Analysis of SARS-CoV-2 Infections Among Health System and University Employees

Background

We aimed to describe SARS-CoV-2 (COVID-19) infections among employees in a large, academic institution.

Table 1. COVID-19 Attribution Definitions

Methods

We prospectively tracked and traced COVID-19 infections among employees across our health system and university. Each employee with a confirmed positive test and 3 presumed positive cases were interviewed with a standard contact tracing template that included descriptive variables such as high-risk behaviors and contacts, dates worked while infectious, and initial symptoms. Using this information, the most likely location of infection acquisition was adjudicated (Table 1). We compared behavior frequency between community and unknown, likely community and community and unknown cases using descriptive statistics.

Table 3. Risk Factors for Community, Likely Community, and Unknown Cases

Results

From 3/2020 to 4/2021 we identified 3,140 COVID-19 infections in 3,119 employees out of a total of 34,562 employees (9.0%) (Figure 1). Of those 3,119 employees 1,685 (54.0%) were clinical employees working in the health system, 916 (29.4%) were non-clinical employees working in the health system, and 518 (16.6%) were university employees. Descriptive characteristics for the COVID-19 infections and adjudications are outlined in Table 2. Severe disease among employees was significantly less frequent compared to patients in the health system (15.3% vs 2.2%, p&lt; 0.01). The frequency of travel within 14 days, masked gatherings and unmasked gatherings/activities was not significantly different between the community and unknown, likely community groups or the community and unknown groups (Table 3).

Conclusion

The majority of COVID-19 infections were linked to acquisition in the community, and few were attributed to workplace exposures. Employees with unknown sources of COVID-19 participated in higher-risk activities at approximately the same frequency as employees with community sources of COVID-19. The most frequently reported initial symptoms were mild and non-specific and rarely included fever. Despite a comprehensive testing and benefit program, a large proportion of COVID-positive employees worked with symptoms, highlighting ongoing challenges with presenteeism in healthcare.

Disclosures

Rebekah W. Moehring, MD, MPH, UpToDate, Inc. (Other Financial or Material Support, Author Royalties)

418. Low Frequency of Healthcare Worker Infections Following Occupational Exposures to COVID-19

Background

Data on occupational acquisition of COVID-19 in healthcare settings are limited. Contact tracing efforts are high resource investments.

Methods

Duke Health developed robust COVID-19 contact tracing methods as part of a comprehensive prevention program. We prospectively collected data on HCW exposures and monitored for development of symptomatic (SYX) and asymptomatic (ASYX) COVID-19 infection after documented high-, medium, and low-risk exposures. HCWs were required to self-report exposures or were identified through contact tracing as potentially exposed to COVID-19 positive HCWs, patients or visitors. Contact tracers interviewed exposed HCWs and assessed the risk of exposure as high-, medium-, or low-risk based on CDC guidance (Table 1). Testing was recommended at 6 days after high- or medium-risk exposures and was provided upon HCW request following low-risk exposures. Our vaccination campaign began in 12/2020.

Results

12,916 HCWs registered in the contact tracing database. From March 2020-May 2021, we identified 6,606 occupational exposures (0.51 exposures/HCW). The highest incidence of workplace exposures per number of HCWs in each job category was among respiratory therapists (RT) (0.95 exposures/RT), nursing assistants (NA) (0.79 exposures/NA), and physicians (0.64 exposures/physician). The most common exposure risk level was medium (51.4%), followed by low (35.5%), and then high (13.1%). A total of 260 (2%) HCW had positive tests/conversions; 28 (10.8%) were ASYX at the time of testing. High-risk exposures had a significantly greater number of post-exposure infections compared to medium- and low-risk exposures (12.5% vs. 4.2%, vs. 0.4%; p < 0.001). The rate of SYX infection following exposure to a fellow HCW (179/3,198; 5.6%) was higher than that following exposure to a patient (81/3,408; 2.4%; p< 0.001).

Conclusion

Conversion following exposure to COVID-19 in the healthcare setting with appropriate protective equipment was low. Incomplete testing of all exposed individuals was a limitation and our data may under-estimate the true conversion rate. Our findings support our local practice of not quarantining HCWs following non-household exposures. Limiting contact tracing to only high or medium risk exposures may best utilize limited personnel resources.

Disclosures

Rebekah W. Moehring, MD, MPH, UpToDate, Inc. (Other Financial or Material Support, Author Royalties)

384. SARS-CoV-2 Surveillance Testing Patterns among Hospitalized Pediatric Patients in a Single Academic Medical Center

Background

Children infected with SARS-CoV-2 often have mild or no symptoms, making symptom screening an ineffective tool for determining isolation precautions. As an infection control measure, universal pre-procedural and admission SARS-CoV-2 testing for pediatric patients was implemented in April and August 2020, respectively. Limited data exist on the utility screening programs in the pediatric population.

Methods

We performed a retrospective cohort study of pediatric patients (birth to 18 years) admitted to a tertiary care academic medical center from April 2020 to May 2021 that had one or more SARS-CoV-2 point-of-care or polymerase chain reaction tests performed. We describe demographic data, positivity rates and repeat testing trends observed in our cohort.

Results

A total of 2,579 SARS-CoV-2 tests were performed among 1,027 pediatric inpatients. Of these, 51 tests (2%) from 45 patients (4.3%) resulted positive. Community infection rates ranged from 4.5-60 cases/100,000 persons/day during the study period. Hispanic patients comprised 16% of the total children tested, but were disproportionately overrepresented (40%) among those testing positive (Figure1). Of 654 children with repeated tests, 7 (0.1%) converted to positive from a prior negative result. Median days between repeat tests was 12 (IQR 6-45), not necessarily performed during the same hospital stay. Five of these 7 patients had tests repeated < 3 days from a negative result, of which only 2 had no history of recent infection by testing performed at an outside facility. Pre-procedural tests accounted for 35% of repeat testing, of which 0.9% were positive. Repeated tests were most frequently ordered for patients in hematology/oncology (35%) and solid organ transplant/surgical (33%) wards, each with < 3% positive conversion rate. Notably, no hematopoietic stem cell transplant patients tested positive for SARS-CoV-2 during the study period.

Pediatric SARS-CoV-2 Testing Distributed by Race/Ethnicity

Conclusion

The positivity rate of universal pre-procedural and admission SARS-CoV-2 testing in pediatric patients was low in our inpatient cohort. Tests repeated < 3 days from a negative result were especially low yield, suggesting limited utility of this practice. Diagnostic testing stewardship in certain populations may be useful, especially as community infection rates decline.

Disclosures

Michael J. Smith, MD, M.S.C.E, Merck (Grant/Research Support)Pfizer (Grant/Research Support) Rebekah W. Moehring, MD, MPH, UpToDate, Inc. (Other Financial or Material Support, Author Royalties)

106. Pandemic Pinch: The Impact of COVID Response on Antimicrobial Stewardship Program (ASP) Resource Allocation

Background

The COVID-19 pandemic placed a strain on inpatient clinical and hospital programs due to increased patient volume and rapidly evolving data on best COVID-19 management strategies. However, the impact of the pandemic on ASPs has not been well described.

Methods

We performed a cross-sectional electronic survey of stewardship pharmacy and physician leaders in 37 hospitals within the Duke Antimicrobial Stewardship Outreach Network (DASON) (community) and Duke/UNC Health systems (academic) in April-May 2021. The survey included 60 questions related to staffing changes, use of COVID-targeted therapies, related restrictions, and medication shortages.

Results

Twenty-seven facilities responded (response rate of 73%). Pharmacy personnel was reduced in 17 (63%) facilities by an average of 16%. Impacted pharmacy personnel included the stewardship lead in 15/17 (88.2%) hospitals. Converting to remote work was rare and only reported in academic institutions (n=2, 7.4%). ASP personnel were reassigned to non-stewardship duties in 12 (44%) hospitals with only half returning to routine ASP work as of May 2021. Respondents estimated that 62% of routine ASP activities were diverted during the time of the pandemic. Non-traditional, pandemic-related ASP activities included managing multiple drug shortages, of which ventilator support medications (91%) were most common affecting patient care at 52% of facilities. Steroid and hydroxychloroquine shortages were less frequent (44% and 22%, respectively). Despite staff reductions, pharmacists often served as primary contact for remdesivir approvals either using a criteria-based checklist at dispensing or as part of a dedicated phone approval team (Figure). Most (77%) hospitals used a criteria-based pharmacist review strategy after remdesivir FDA approval. Restriction processes for other COVID-19 therapies such as tocilizumab, hydroxychloroquine, and ivermectin were reported in 64% of hospitals.

Remdesivir Allocation Strategy

Proportion of facilities implementing specific remdesivir allocation strategies from the time of the first US Food and Drug Administration (FDA) Emergency Use Authorization (EUA) through FDA approval

Conclusion

Pandemic response diverted routine ASP work and has not yet returned to baseline. Despite the reduction in pharmacy personnel due to the pandemic, the ASP pharmacy lead took on a novel and critical stewardship role throughout the pandemic exemplified by their involvement in novel treatment allocation for COVID patients.

Disclosures

Melissa D. Johnson, PharmD, MHS, Astellas (Consultant, Grant/Research Support)Charles River Laboratories (Grant/Research Support)Cidara (Consultant)Merck & Co (Consultant, Research Grant or Support)Paratek (Consultant)Pfizer (Consultant)Scynexis (Scientific Research Study Investigator)Theratechnologies (Consultant)UpToDate (Other Financial or Material Support, Author Royalties) Sonali D. Advani, MBBS, MPH, Nothing to disclose Rebekah W. Moehring, MD, MPH, UpToDate, Inc. (Other Financial or Material Support, Author Royalties)

14. Effects of an Opt-Out Protocol for Antibiotic De-escalation among Selected Patients with Suspected Sepsis: The DETOURS Trial

Background

Sepsis guidelines recommend daily review to de-escalate or stop antibiotics in appropriate patients. We conducted a randomized controlled trial (NCT03517007) of an opt-out protocol to decrease unnecessary antibiotics in selected patients with suspected sepsis.

Methods

We evaluated non-ICU adults remaining on broad-spectrum antibiotics with negative blood cultures at 48-96 hours at ten U.S. hospitals during September 2018-May 2020. A 23-item safety check excluded patients with ongoing signs of infection, concerning or inadequate microbiologic data, or high-risk conditions (Figure 1). Eligible patients were randomized to the opt-out protocol vs. usual care. The primary outcome was 30-day post-enrollment antibacterial days of therapy (DOT). Clinicians caring for intervention patients were contacted by a pharmacist or physician to encourage antibiotic discontinuation or de-escalation using opt-out language, discuss rationale for continuing antibiotics, working diagnosis, and de-escalation and duration plans. Hurdle models separately compared the odds of antibiotic continuation and DOT distributions among those who continued antibiotics.

Components of the De-Escalating Empiric Therapy: Opting-OUt of Rx in Selected patients with Suspected Sepsis (DETOURS) Trial Protocol

Results

Among 9606 screened, 767 (8%) were enrolled (Figure 2). Common reasons for exclusion were antibiotics given prior to blood culture (35%), positive culture from non-blood sites (26%), and increased oxygen requirement (21%). Intervention patients had 32% lower odds of antibiotic continuation (79% vs. 84%, OR 0.68, 95% confidence interval [0.47, 0.98]). DOT distributions among those who continued antibiotics were similar (ratio of means 1.06 [0.88-1.26], Figure 3). Fewer intervention patients were exposed to extended-spectrum agents (38% vs. 44%). Common reasons for continuing antibiotics were treatment of localized infection (76%) and belief that stopping antibiotics was not safe (31%). Safety outcomes such as mortality, readmission, sepsis relapse, C. difficile, and length of stay did not differ.

DETOURS Trial Flow Diagram

Flow of participants through the DETOURS Trial.

Observed Days of Antibiotic Therapy Among Intervention and Control Subjects in the DETOURS Trial

Post-enrollment days of antibiotic therapy among 767 DETOURS Trial participants in 10 US acute care hospitals within 30 days after enrollment. Dark pink color indicates percent overlap between intervention (purple) and control (light pink) groups.

Conclusion

In this patient-level randomized trial of a stewardship intervention, the opt-out de-escalation protocol targeting selected patients with suspected sepsis resulted in more antibiotic discontinuations but did not affect safety events.

Disclosures

Rebekah W. Moehring, MD, MPH, UpToDate, Inc. (Other Financial or Material Support, Author Royalties) Michael Z. David, MD PhD, GSK (Board Member) Michael Klompas, MD, MPH, UpToDate (Other Financial or Material Support, Chapter Author)

1413. Effect of Automated Identification of Antimicrobial Stewardship Opportunities for Urinary Tract Infections

Background

The treatment of asymptomatic bacteriuria (ASB) does not improve clinical outcomes in most patients and may be associated with an increased risk of adverse events such as Clostridioides difficile infection. A best practice alert (BPA) was created to identify patients with possible ASB for antimicrobial stewardship (AS) review. We aimed to determine whether automated identification of ASB improved the timing of stewardship intervention.

Methods

An electronic health record BPA message to inpatient AS pharmacists was activated on 01/19/2021. The BPA identified inpatients with a new antibiotic order with an associated genitourinary indication and a preceding urinalysis with 0 to 5 WBC/hpf. BPAs were reviewed by an AS pharmacist during weekdays and normal business hours. We retrospectively evaluated the impact of the BPA on time from order to stewardship intervention between a cohort of pre-BPA (01/2020 to 12/2020) and post-BPA (01/20/2021 to 04/10/2021) patients. Included patients met the BPA criteria and had an AS intervention within 7 days of the antibiotic order. We specified interventions that were UTI-related. The median time from antibiotic order entry to any AS intervention was compared pre- to post-BPA using the Mann Whitney U test. Rates of UTI-related interventions were compared with Fisher’s Exact test.

Results

327 antibiotic orders met BPA criteria and were analyzed: 245 and 82 in the pre- and post-BPA group, respectively. Groups had similar baseline characteristics (Table 1). A total of 33 (27 UTI-related) pre-BPA group and 24 (17 UTI-related) post-BPA group interventions were documented by the AS team. The median time to any intervention was 28 hours (IQR 18-64.5) in the pre-BPA group compared to 13.5 hours (IQR 3.5-28.75) in the post-BPA group (p = 0.03, Figure). The pre-BPA group had a lower rate of UTI-related interventions compared to the post-BPA group (11.0% vs 20.7%, p = .04).

Conclusion

Automated identification of antibiotics targeting UTI with urinalysis showing absence of pyuria reduced the time to stewardship intervention and increased rate of UTI-specific interventions. The use of clinical decision support may aid in efficiency of AS review and syndrome-targeted AS impact.

Disclosures

Rebekah W. Moehring, MD, MPH, UpToDate, Inc. (Other Financial or Material Support, Author Royalties)

The Impact of CMS SEP-1 Core Measure Implementation on Antibacterial Utilization: a retrospective multicenter longitudinal cohort study with interrupted time-series analysis

BACKGROUND

The impact of the U.S. Centers for Medicare and Medicaid Services (CMS) Severe Sepsis and Septic Shock: Management Bundle (SEP-1) Core Measure on overall antibacterial utilization is unknown.

METHODS

We performed a retrospective multicenter longitudinal cohort study with interrupted time series analysis to determine the impact of SEP-1 implementation on antibacterial utilization and patient outcomes. All adult patients admitted to 26 hospitals between October 1, 2014, and September 30, 2015 (the "SEP-1 preparation period") and between November 1, 2015, and October 31, 2016 (the "SEP-1 implementation period") were evaluated for inclusion.The primary outcome was total antibacterial utilization measured as days of therapy (DOT) per 1,000 patient days.

RESULTS

The study cohort included 701,055 eligible patient admissions and 4.2 million patient days. Overall antibacterial utilization increased 2% each month during SEP-1 preparation (RR=1.02 per month [95% CI 1.00-1.04]; p=0.02). Cumulatively, the mean monthly DOT/1,000 patient-days increased 24.4% [95% CI 18.0, 38.8] over the entire study period (October 2014-October 2016). The rate of sepsis diagnosis/1,000 patients increased 2% each month during SEP-1 preparation (RR=1.02 per month [95% CI 1.00-1.04]; p=0.04). The rate of all-cause mortality/1,000 patients decreased during the study period (SEP-1 preparation RR=0.95 [0.92-0.98]; p=0.001 and SEP-1 implementation RR=0.98 [95% CI 0.97-1.00]; p=0.01). Cumulatively, the monthly mean all-cause mortality/1,000 patients declined 38.5% [95% CI 25.9, 48.0] over the study period.

CONCLUSIONS

Announcement and implementation of the CMS SEP-1 process measure was associated with increased diagnosis of sepsis and antibacterial utilization and decreased mortality among hospitalized patients.

Impact of Antibiotic Stewardship Rounds in the Intensive Care Setting: a prospective cluster-randomized crossover study

BACKGROUND

Few groups have formally studied the effect of dedicated antibiotic stewardship rounds (ASRs) on antibiotic use (AU) in intensive care units (ICUs).

METHODS

We implemented weekly ASRs using a two-arm, cluster-randomized, crossover study in 5 ICUs at Duke University Hospital from 11/2017 to 6/2018. We excluded patients without an active antibiotic order, or if they had a marker of high complexity including an existing infectious disease consult, transplant, ventricular assist device, or ECMO. AU during and following ICU stay for patients with ASRs was compared to the controls. We recorded the number of reviews, recommendations delivered, and responses. We evaluated change in ICU-specific AU during and after the study.

RESULTS

Our analysis included 4,683 patients: 2330 intervention and 2353 controls. Teams performed 761 reviews during ASRs, which excluded 1569 patients: 60% of patients off antibiotics, and 8% complex patients. Exclusions affected 88% the cardiac surgery ICU (CTICU) patients. AU rate ratio (RR) was 0.97 (0.91-1.04). When CTICU was removed, the RR was 0.93 (0.89-0.98). AU in the post-study period decreased by 16% (95% CI 11-24%) compared to the AU in the baseline period. Change in AU was differential among units: largest in the neurology ICU (-28%) and smallest in the CTICU (-2%).

CONCLUSION

Weekly multi-disciplinary ASRs was a high-resource intervention associated with a small AU reduction. The noticeable ICU AU decline over time is possibly due to indirect effects of ASRs. Effects differed among specialty ICUs, emphasizing the importance of customizing ASRs to match unit-specific population, workflow, and culture.

Tap Water Avoidance Decreases Rates of Hospital-onset Pulmonary Nontuberculous Mycobacteria

We analyzed the impact of a hospital tap water avoidance protocol on respiratory isolation of nontuberculous mycobacteria (NTM). After protocol implementation, hospital-onset episodes of respiratory NTM isolation on high-risk units decreased from 41.0 to 9.9 episodes per 10 000 patient-days (incidence rate ratio, 0.24; 95% confidence interval, .17-.34; P < .0001).

Evaluation of a Pharmacist-Led Penicillin Allergy Assessment Program and Allergy Delabeling in a Tertiary Care Hospital

IMPORTANCE

Penicillin allergies are frequently mislabeled, which may contribute to use of less-preferred alternative antibiotics.

OBJECTIVE

To evaluate a pharmacist-led allergy assessment program's association with antimicrobial use and clinical outcomes.

DESIGN, SETTING, AND PARTICIPANTS

A pharmacist-led allergy assessment program was launched in 2 phases (June 1, 2015, and November 2, 2016) at a single-center tertiary referral hospital. The longitudinal cross-sectional study included all study period adult admissions; hospitalwide outcomes were assessed by segmented regression. Individual outcomes were assessed within an embedded propensity score-matched case-control study of inpatients undergoing comprehensive allergy assessment following self-report of penicillin allergy. Analysis occurred from March 1, 2020, to February 29, 2020.

EXPOSURES

The longitudinal study analyzed hospital-level outcomes over 3 periods: preintervention (15 months), phase 1 (structured allergy history alone, 16 months), and phase 2 (comprehensive assessment including penicillin skin testing, 52 months). The case-control study defined cases as individuals undergoing comprehensive allergy assessment.

MAIN OUTCOMES AND MEASURES

Hospital-level outcomes included antibiotic days of therapy per 1000 patient-days and hospital-acquired Clostridioides difficile infection (CDI) incidence per 10 000 patient-days. Individual outcomes included antibiotic selection, overall survival, and CDI-free survival.

RESULTS

Longitudinal analysis spanned 2014-2020 (median admissions, 46 416 per year; interquartile range [IQR], 46 001-50 091 per year). Hospitalwide, allergy histories were temporally associated with decreased use of nonpenicillin alternative antibiotics (rate ratio, 0.87; 95% CI, 0.79-0.97) and high-CDI-risk antibiotics (rate ratio, 0.91; 95% CI, 0.85-0.98). Penicillin skin testing was temporally associated with lower hospital-acquired CDI rates (rate ratio, 0.61; 95% CI, 0.43-0.86). The embedded case-control study included 272 cases and 819 controls. Median age was 63 years (interquartile range, 51-73 years), 553 (50.7%) patients were women, and 229 (21.0%) patients were Black. Allergy-assessed patients were less likely to receive high-CDI-risk antibiotics at discharge (odds ratio, 0.66; 95% CI, 0.44-0.98). Estimated reductions in mortality (hazard ratio, 0.77; 95% CI, 0.55-1.07) and hospital-acquired CDI risk (hazard ratio, 0.53; 95% CI, 0.18-1.55) were not statistically significant.

CONCLUSIONS AND RELEVANCE

Pharmacist-led allergy assessments may be associated with reduced high-CDI-risk antibiotic use at both hospitalwide and individual levels. Although individual reductions in mortality and CDI risk did not achieve significance, divergence of survival curves suggest longer-term benefits of allergy delabeling warrant future study.

Effects of a Collaborative, Community Hospital Network for Antimicrobial Stewardship Program Implementation

BACKGROUND

Individual hospitals may lack expertise, data resources, and educational tools to support antimicrobial stewardship programs (ASP).

METHODS

We established a collaborative, consultative network focused on hospital ASP implementation. Services included on-site expert consultation, shared database for routine feedback and benchmarking, and educational programs. We performed a retrospective, longitudinal analysis of antimicrobial use (AU) in 17 hospitals that participated for at least 36 months during 2013-2018. ASP practice was assessed using structured interviews. Segmented regression estimated change in facility-wide AU after a 1-year assessment, planning, and intervention initiation period. Year one AU trend (1 to 12 months) and AU trend following the first year (13 to 42 months) were compared using relative rates (RR). Monthly AU rates were measured in days of therapy (DOT) per 1,000 patient days for overall AU, specific agents, and agent groups.

RESULTS

Analyzed data included over 2.5 million DOT and almost 3 million patient-days. Participating hospitals increased ASP-focused activities over time. Network-wide overall AU trends were flat during the first 12 months after network entry but decreased thereafter (RR month 42 vs month 13, 0.95, 95% Confidence Interval (CI) 0.91-0.99.) Large variation was seen in hospital-specific AU. Fluoroquinolone use was stable during year one, then dropped significantly. Other agent groups demonstrated a non-significant downward trajectory after year one.

CONCLUSIONS

Network hospitals increased ASP activities and demonstrated decline in AU over a 42-month period. A collaborative, consultative network is a unique model in which hospitals can access ASP implementation expertise to support long-term program growth.

Development of a Machine Learning Model Using Electronic Health Record Data to Identify Antibiotic Use Among Hospitalized Patients

IMPORTANCE

Comparisons of antimicrobial use among hospitals are difficult to interpret owing to variations in patient case mix. Risk-adjustment strategies incorporating larger numbers of variables haves been proposed as a method to improve comparisons for antimicrobial stewardship assessments.

OBJECTIVE

To evaluate whether variables of varying complexity and feasibility of measurement, derived retrospectively from the electronic health records, accurately identify inpatient antimicrobial use.

DESIGN, SETTING, AND PARTICIPANTS

Retrospective cohort study, using a 2-stage random forests machine learning modeling analysis of electronic health record data. Data were split into training and testing sets to measure model performance using area under the curve and absolute error. All adult and pediatric inpatient encounters from October 1, 2015, to September 30, 2017, at 2 community hospitals and 1 academic medical center in the Duke University Health System were analyzed. A total of 204 candidate variables were categorized into 4 tiers based on feasibility of measurement from the electronic health records.

MAIN OUTCOMES AND MEASURES

Antimicrobial exposure was measured at the encounter level in 2 ways: binary (ever or never) and number of days of therapy. Analyses were stratified by age (pediatric or adult), unit type, and antibiotic group.

RESULTS

The data set included 170 294 encounters and 204 candidate variables from 3 hospitals during the 3-year study period. Antimicrobial exposure occurred in 80 190 encounters (47%); 64 998 (38%) received 1 to 6 days of therapy, and 15 192 (9%) received 7 or more days of therapy. Two-stage models identified antimicrobial use with high fidelity (mean area under the curve, 0.85; mean absolute error, 1.0 days of therapy). Addition of more complex variables increased accuracy, with largest improvements occurring with inclusion of diagnosis information. Accuracy varied based on location and antibiotic group. Models underestimated the number of days of therapy of encounters with long lengths of stay.

CONCLUSIONS AND RELEVANCE

Models using variables derived from electronic health records identified antimicrobial exposure accurately. Future risk-adjustment strategies incorporating encounter-level information may make comparisons of antimicrobial use more meaningful for hospital antimicrobial stewardship assessments.

45. Antimicrobial Stewardship for Urinary Tract Infection in Three Emergency Departments Across a Health System

Background

Broad spectrum antibiotics are often prescribed to patients presenting to the emergency department (ED) for evaluation of urinary tract infection and pyelonephritis (UTI). We evaluated the effect of a target-specific antibiogram, education, and feedback on UTI diagnosis and antibiotic prescribing in this setting.

Methods

We created a urine-specific antibiogram from patients seen and treated without admission at three ED locations (one academic and two community hospitals). We then provided a treatment algorithm and supplemental educational content to ED providers in November 2019. Educational content highlighted appropriate diagnosis, antibiotic selection, and treatment duration for UTI. Adult encounters with appropriate ICD-9/10 codes within twelve months prior to content delivery comprised the preintervention cohort. The postintervention cohort consisted of adult visits following educational intervention until April 17, 2020. During the postintervention phase (November 2019 to April 2020), summary data regarding UTI diagnoses and guideline-concordant prescriptions were fed back routinely to ED providers through email. Guideline-concordant prescriptions were defined as those adhering to first or second-line therapy in the treatment algorithm. The proportion of prescriptions meeting this definition fulfilled the primary outcome. An interrupted time series analysis measured changes in guideline concordance.

Results

Data from 6,713 distinct encounters were analyzed across the three sites. While guideline concordant prescribing increased following intervention at all locations (30.9% to 38.8%, 48.1% to 49.1%, and 48.2% to 59.6%), these increases were not statistically significant (Figures 1, 2, and 3). The proportion of all ED encounters with a UTI diagnosis did not differ following the intervention. Interestingly, guideline concordance was greater in the academic ED, compared to the community hospitals.

Conclusion

Although guideline concordant prescribing for UTI increased in all three ED settings with education and email correspondence feedback, these results were not statistically significant. A variety of methods may be required to realize improved antibiotic prescribing across a diverse group of clinicians.

Disclosures

Rebekah W. Moehring, MD, MPH, Agency for Healthcare Quality and Research (Grant/Research Support)Centers for Disease Control and Prevention (Grant/Research Support)

51. Development and Assessment of a Process to Describe the Timing of Antibiotic Changes in Adult Inpatients

Background

Hospital antimicrobial stewardship programs (ASP) perform prospective audit and feedback to optimize use of antimicrobials; however, workflow inefficiency continues to be a distinct challenge. We developed a method to describe the volume and timing of antimicrobial changes to inform decisions on optimal timing of ASP review and intervention.

Methods

This retrospective study was performed at Duke University Hospital using anonymized antibiotic administration records from the DASON central database. Eligible antibiotic courses were administered to inpatients ≥ 18 years of age and had received ≥ 2 antibiotics administrations for ≥ 24 hours of treatment. A 2-month exploratory cohort (September to October 2017) was used to develop an antibiotic spectrum ranking (Table 1) and decision algorithm which was applied to a 1-year cohort (November 2017 to October 2018) for analysis of total change in antibiotic orders by day of the week. For each interval, the sum of antibiotic ranks was calculated and applied using specified definitions (Table 2) to determine the type of change occurring. The primary outcome was the number of total antibiotic changes that occurred on each day of the week. Secondary outcomes included the number and type (initiations, discontinuations, de-escalations, and escalations) of change. Descriptive statistics were used to describe the outcomes by day of the week.

Table 1: Antibiotic Spectrum Ranking

Table 2: Key Definitions

Results

The ranking and decision algorithm were applied to 16,993 unique antibiotic courses. Total changes occurred most on Wednesday (14,971, 16.2% [95% CI 15.7–17.1%]) and Friday (14,349, 15.6% [95% CI 15.0–16.2%]). Compared to intervals on weekdays (0.407 mean changes per patients on antibiotics [95% CI 0.401–0.413]), weekends had a lower number of changes (0.363 mean changes per patients on antibiotics [95% CI 0.349–0.377]). Initiations occurred most frequently on Tuesday (3,078, 18.1% [95% CI 16.3–19.9%]), and discontinuations on Wednesday (3,179, 18.7% [95% CI 17.4–20.5%]) (Figure 1).

Figure 1: Types of Changes per Day

Conclusion

We developed and applied a method to characterize antimicrobial changes. In our institution, the reductions in the number of changes observed on weekends provide an opportunity for ASP involvement to be incorporated and help facilitate appropriate antimicrobial changes.

Disclosures

Rebekah W. Moehring, MD, MPH, Agency for Healthcare Quality and Research (Grant/Research Support)Centers for Disease Control and Prevention (Grant/Research Support)

224. Effect of Easing Overnight Restrictions on Antimicrobial Starts

Background

Some institutions allow administration of restricted antibiotics overnight until evaluation the following day (i.e. first dose free) to adapt to limitations in personnel resources. Whether this method results in higher number of overnight requests compared to strict 24/7 preauthorization has not been fully described.

Methods

In October 2019, Duke University Hospital (DUH) changed from strict preauthorization to allow initiation of two restricted agents (meropenem and micafungin) between the hours of 11pm to 7am. We performed an interrupted time series (ITS) analysis to evaluate the phase shift and change in trend in the number of new meropenem and micafungin orders per week before (Jan 2019-Oct 2019) and after (Oct 2019- Mar 2020) the process change. First antimicrobial orders for meropenem and micafungin were counted for unique patient encounters. We fit a Gaussian distribution function to the number of orders per hour of day to estimate the percent of orders initiated overnight (11p-7a) and during day/evening hours (7a-11p) before and after the process change.

Results

Hospital data included 1728 new meropenem and micafungin orders over a 61-week period (~28 per week). The total number of meropenem and micafungin orders was constant between Jan 2019 and October 2019 (+0.07 orders/week, 95% CI -0.13 to 0.27, Figure 1) and the phase shift during the first week of October was non-significant (-4.38 orders, 95% CI -12.34 to 3.58). The number of orders increased after October 2019 (+0.70 orders/week, 95% CI 0.13 to 1.25), however a sensitivity analysis removing the largest outlier eliminates significance. The percent of total orders between 11am to 7pm increased from 13.3% to 17.2% after the intervention (Figure 2). Overall antibiotic use remained similar through the study period.

Figure 1. Estimated Approvals per Week

Figure 2. Approvals by Hour of Day

Conclusion

There was no significant immediate change in overnight prescribing of meropenem and micafungin, however a trend towards increased number of orders appeared after removing overnight restriction requirements. Instead of “stealth dosing”, where providers wait to enter restricted antibiotic orders until evening hours, we observed a small increase in starts in early morning hours (1am-6am). Preauthorization approaches must adapt to personnel resources and quality of life for antimicrobial stewards.

Disclosures

Michael J. Smith, MD, MSC.E, Kentucky Medicaid (Grant/Research Support)Merck (Grant/Research Support) Rebekah W. Moehring, MD, MPH, Agency for Healthcare Quality and Research (Grant/Research Support)Centers for Disease Control and Prevention (Grant/Research Support)

203. Opportunities for Antimicrobial Stewardship in Febrile Neutropenia

Background

Emerging evidence suggests antibiotics may be safely discontinued before neutropenia resolves in patients without identifiable infection. We estimated the volume of encounters and antibiotic use for future stewardship interventions shortening FN treatment duration.

Methods

This retrospective cohort study used electronic health records from inpatient encounters on the hematologic malignancies ward at Duke University Hospital from 5/21/2018 to 12/31/2019 where patients received at least one antibiotic for an indication of “neutropenic fever.” The primary outcome was length of therapy (LOT) of broad Gram-negative (GN) agents, including cefepime, piperacillin-tazobactam, meropenem, or aztreonam. FN LOT was counted by calendar day, starting with the first day of administration of a broad GN agent and ending with antibiotic discontinuation or hospital discharge. Encounters with at least one positive blood culture (positive cohort) were compared to those with no positive blood cultures (negative cohort) to assess if culture positivity was associated with differences in FN LOT. We included the first FN LOT from each encounter in the negative cohort and the FN LOT associated with the first positive blood culture in the positive cohort. We used descriptive statistics and a Gaussian density function to calculate the percent of encounters exceeding FN LOT of 14 days and the percent of broad GN agent days.

Results

We evaluated 15,678 GN antibiotic administrations from 471 unique FN encounters. Blood culture results were available for 443 encounters— 122 (27.5%) in the positive cohort, and 321 (72.5%) encounters in the negative cohort. Thirty percent of encounters (36/122) in the positive cohort received more than one GN treatment course, compared to 10% (32/321) of those in the negative cohort. FN LOT was significantly longer in the positive cohort (median 10.5, IQR 13 days vs. 6, IQR 8 days, p < 0.001). Among encounters with negative cultures, 57 (17.8%) had a first FN LOT greater than 14 days, accounting for 44% of broad GN agent days within that population (Figure 1).

Gram-Negative Antibiotic Therapy in Blood Culture-Negative Febrile Neutropenia

Conclusion

Nearly 20% of blood culture-negative encounters received initial GN treatment courses exceeding 14 days, representing a sizeable target for antimicrobial stewardship interventions focused on FN treatment duration.

Disclosures

Rebekah W. Moehring, MD, MPH, Agency for Healthcare Quality and Research (Grant/Research Support)Centers for Disease Control and Prevention (Grant/Research Support)

Navigating reflex urine culture practices in community hospitals: Need for a validated approach

We performed a descriptive study of reflex urine culture (RUC) practices across 51 community hospitals in southeastern United States. We found that 26 unique reflexing criteria were used in 28 hospitals. Only 14% hospitals of hospitals that offered RUC restricted it to specific populations (eg, emergency room). Our data suggest that the current RUC approach in community hospitals warrants further validation of urinalysis criteria and identification of specific populations in which RUC performs best.

Impact of Location of Acquisition of Gram-Positive Bloodstream Infections on Clinical Outcomes Among Patients Admitted to Community Hospitals

Purpose

We investigated the association between location of acquisition (LOA) of gram-positive (GP) bloodstream infections (BSI) in community hospitals and clinical outcomes.

Methods

We performed a multicenter cohort study of adult inpatients with GP BSI in nine community hospitals from 2003 to 2006. LOA was defined by CDC criteria: 1) community-acquired (CA), 2) healthcare-associated (HCA) such as BSI <48 hours after admission plus hospitalization, surgery, dialysis, invasive device, or residence in a long-term care facility in the prior 12 months, and 3) hospital-acquired (HA) as BSI ≥48 hours after hospital admission.

Results

A total of 750 patients were included. Patients with HCA or HA GP BSI were significantly more likely to require assistance with ≥1 activity of daily living, have higher Charlson scores, and die during the hospitalization. Patients with HCA or HA GP BSI were more likely to have BSI due to a multidrug-resistant GP organism, but less likely to receive appropriate antibiotics within 24 hours of BSI presentation. Those with CA BSI were more likely to have a streptococcal BSI and to be discharged home following hospitalization. HA BSI was a risk factor for requiring a procedure for BSI and receiving inappropriate antibiotics within 24 hours of BSI. Both HA and HCA GP BSI were risk factors for in-hospital mortality.

Conclusion

LOA for patients with GP BSI in community hospitals was significantly associated with differences in clinical outcomes including receiving inappropriate antibiotics and in-hospital mortality. Distinguishing LOA in a patient presenting with suspected GP BSI is a critical assessment that should influence empiric treatment patterns.

Development of an electronic definition for de-escalation of antibiotics in hospitalized patients

BACKGROUND

Antimicrobial stewardship programs (ASPs) promote the principle of de-escalation: moving from broad to narrow spectrum agents and stopping antibiotics when no longer indicated. A standard, objective definition of de-escalation applied to electronic data could be useful for ASP assessments.

METHODS

We derived an electronic definition of antibiotic de-escalation and performed a retrospective study among five hospitals. Antibiotics were ranked into 4 categories: narrow spectrum, broad spectrum, extended spectrum, and agents targeted for protection. Eligible adult patients were cared for on inpatient units, had antibiotic therapy for at least 2 days, and were hospitalized for at least 3 days after starting antibiotics. Number of antibiotics and rank were assessed at two time points: day of antibiotic initiation and either day of discharge or day 5. De-escalation was defined as reduction in either the number of antibiotics or rank. Escalation was an increase in either number or rank. Unchanged was either no change or discordant directions of change. We summarized outcomes among hospitals, units, and diagnoses.

RESULTS

Among 39,226 eligible admissions, de-escalation occurred in 14,138 (36%), escalation in 5,129 (13%), and antibiotics were unchanged in 19,959 (51%). De-escalation varied among hospitals (median 37%, range 31-39%, p<.001). Diagnoses with lower de-escalation rates included intra-abdominal (23%) and skin and soft tissue (28%) infections. Critical care had higher rates of both de-escalation and escalation compared with wards.

CONCLUSIONS

Our electronic de-escalation metric demonstrated variation among hospitals, units, and diagnoses. This metric may be useful for assessing stewardship opportunities and impact.

Identifying predictors of antimicrobial exposure in hospitalized patients using a machine learning approach

AIMS

Analysis and tracking of antimicrobial utilization (AU) are crucial in antimicrobial stewardship efforts which are used to find effective interventions for controlling antimicrobial resistance. In antimicrobial stewardship, standard risk adjustment models are needed for benchmarking appropriate AU and for fair inter-facility comparison. In this study we identify patient- and facility-level predictors of antimicrobial usage in hospitalized patients using a machine learning approach, which can be used to inform a risk adjustment model to facilitate assessment of AU. To our knowledge, this is the first time machine learning has been applied for this purpose.

METHODS AND RESULTS

Patient admission records were retrieved from the Duke Antimicrobial Stewardship Outreach Network which include clinical data for 27 community hospitals in the southeastern United States. Candidate features (predictors) were then generated from these records. The number of features was reduced using a statistical approach, and missing values of the reduced feature set were imputed using bootstrapping and expectation-maximization algorithm. Finally, support vector regression (SVR) and cubist regression (CB) models were applied to find root-mean-square error values which were used to evaluate the selected feature set. The performance of the SVR and CB models was found to be better than that of linear null and negative binomial null models, thereby demonstrating the effectiveness of our selected features.

CONCLUSIONS

Relevant patient- and facility-level predictors of antimicrobial usage in days of therapy were obtained and evaluated. The potential predictor set can be used in risk adjustment strategies for benchmarking antimicrobial use.

SIGNIFICANCE AND IMPACT OF THE STUDY

One reason for the rapid emergence of antimicrobial resistance is inappropriate use of antibiotics in hospitalized patients. Identifying predictors of antimicrobial exposure using a machine learning technique can improve the use of AU, enhance patient health outcomes, and reduce the infection spread caused by antimicrobial-resistant organisms.

Essential Resources and Strategies for Antibiotic Stewardship Programs in the Acute Care Setting

Background

Antibiotic stewardship programs improve clinical outcomes and patient safety and help combat antibiotic resistance. Specific guidance on resources needed to structure stewardship programs is lacking. This manuscript describes results of a survey of US stewardship programs and resultant recommendations regarding potential staffing structures in the acute care setting.

Methods

A cross-sectional survey of members of 3 infectious diseases subspecialty societies actively involved in antibiotic stewardship was conducted. Survey responses were analyzed with descriptive statistics. Logistic regression models were used to investigate the relationship between stewardship program staffing levels and self-reported effectiveness and to determine which strategies mediate effectiveness.

Results

Two-hundred forty-four respondents from a variety of acute care settings completed the survey. Prior authorization for select antibiotics, antibiotic reviews with prospective audit and feedback, and guideline development were common strategies. Eighty-five percent of surveyed programs demonstrated effectiveness in at least 1 outcome in the prior 2 years. Each 0.50 increase in pharmacist and physician full-time equivalent (FTE) support predicted a 1.48-fold increase in the odds of demonstrating effectiveness. The effect was mediated by the ability to perform prospective audit and feedback. Most programs noted significant barriers to success.

Conclusions

Based on our survey's results, we propose an FTE-to-bed ratio that can be used as a starting point to guide discussions regarding necessary resources for antibiotic stewardship programs with executive leadership. Prospective audit and feedback should be the cornerstone of stewardship programs, and both physician leadership and pharmacists with expertise in stewardship are crucial for success.

2089. Effect of the Duke Antimicrobial Stewardship Outreach Network (DASON): A Multi-Center Time Series Analysis

Background

DASON is a 30-member, community hospital network in the southeastern United States that supports the development and growth of local antibiotic stewardship programs (ASPs). Collaborative activities include on-site visits from liaison clinical pharmacists, data sharing for routine feedback and benchmarking, and educational programs.

Methods

We performed a retrospective cohort analysis of antibiotic use (AU) in 17 hospitals that participated in DASON for a minimum of 42 months during 2013–2018. Segmented negative binomial regression models were used to estimate the change in facility-wide AU after an initial 1-year assessment, planning, and ASP intervention initiation period. Baseline AU trend (1 to 12 months) was compared against AU following the first year (13 to 42 months). Monthly AU rates were measured in days of therapy (DOT) per 1,000 patient-days (pd). Models assessed overall AU and specific antibiotic groups, as defined by the National Healthcare Safety Network AU option. The models controlled for hospital size, presence of a pre-existing formal ASP upon network entry, and year of network entry.

Results

Hospital data included a total of 2,988,930 pd over 5 years. Facility-wide AU was increasing during the first year of network entry and then began decreasing by 0.2% per month (P = 0.01, figure). Fluoroquinolone use was stagnant in year one and then decreased by 1.5% per month (P ≤ 0.001, figure). Antifungal agents were decreasing in year one and continued to decrease 0.7% per month thereafter (P = 0.03, figure). Agents predominantly used for resistant Gram-positive infections and broad-spectrum agents used for hospital-onset infections were increasing during year one and then attenuated afterward, though the slope change did not reach statistical significance. The presence of a pre-existing formal ASP was not a significant covariate in any model, while bed size and year of network entry significantly contributed to models of some antibiotic groups.

Conclusion

Participation in DASON was associated with a decline in total AU and fluoroquinolone use, and a trend toward attenuated use of other broad-spectrum agents in community hospitals. Collaborative network experiences can help local ASPs achieve reductions in AU.

Disclosures

All authors: No reported disclosures.

1013. Electronic Assessment of Empiric Antibiotic Prescribing Using Diagnosis Codes

Background

Antimicrobial stewardship programs (ASPs) must understand empiric choices for specific disease syndromes to assess adherence to local empiric treatment guidelines. Electronically-derived metrics to track empiric therapy choices would allow ASPs to target areas for intervention without significant data collection burden.

Methods

Admissions from 10 community hospitals between 7/2016 and December 2018 were reviewed to identify those with common infectious syndromes: pneumonia (PNA), urinary tract infection (UTI) and skin and soft-tissue infection (SSTI). Admissions with a syndrome of interest were identified using AHRQ clinical classifications software codes based on ICD-10 codes for infection at the time of discharge. Admissions were categorized as having the syndrome of interest with or without sepsis. Antibiotics received during the first 48 hours of inpatient admission were obtained from electronic medication administration records. The proportion of syndrome admissions receiving specific antibiotic agents was determined to evaluate initial treatment choices as compared with local empiric guidelines. Antibiotic categories were not mutually exclusive, admissions receiving combination therapy were included in the count for each individual agent as well as the combination group. The denominator was the count of admissions with the syndrome of interest. Distributions were tracked over time to observe the effects of ASP intervention.

Results

The analysis included 49,303 admissions. The most common diagnosis was UTI (30%) followed by PNA (23%). Empiric antibiotic use varied by syndrome (Figure 1). In general, patients with a targeted infectious diagnosis and sepsis received more broad-spectrum agents than those without sepsis. SSTI was an exception, but few patients admitted with SSTI did not also have presumed sepsis. Longitudinal analysis demonstrated shifts from less preferred agents to guideline-concordant choices. For example, for admissions with a diagnosis of PNA, we observed a steady year on year increase in ceftriaxone (preferred) while levofloxacin (avoided in local guidelines) declined. (Figure 2)

Conclusion

Syndrome-specific diagnosis codes were helpful in assessing empiric antibiotic selection and may assist ASPs in improving empiric guideline adherence.

Disclosures

All authors: No reported disclosures.

1878. Title: Impact of Antibiotic Stewardship Rounds in the Intensive Care Setting: A Prospective Cluster-Randomized Crossover Study

Background

The impact of formalized, interdisciplinary antimicrobial stewardship program (ASP) rounds in the intensive care unit (ICU) setting has not been well described.

Methods

We performed a two-arm, cluster-randomized, crossover quality improvement study over 8 months to compare the impact of weekly ICU rounds with an ASP team vs. usual care. The primary outcome was antibiotic use (AU) in days of therapy (DOT) per 1,000 days present during and following ICU exposure. Our cohort consisted of ICU patients in 5 ICUs in Duke University Hospital. The unit of randomization was rounding team, which corresponded to half of the ICU beds in each unit. Each team was randomized to the intervention for 4 months followed by usual care for 4 months (or vice versa). The intervention involved multidisciplinary review of eligible patients to discuss antibiotic optimization. Patients not on antibiotics, followed by infectious diseases, post-transplant, on ECMO, or with a ventricular assist device were excluded from review. Intervention impact was assessed with multivariable negative binomial regression rate ratios (RR). AU was assessed over time before and after the study period to assess global and unit-level trends.

Results

We had 4,683 ICU-exposed patients. Intervention effect was not significant for the primary outcome (table). The intervention order was not significant in the model. Eligible patients were lower in the cardiothoracic ICU (CTICU) compared with other units (table); the intervention led to a significant decrease in AU when the CTICU was removed (RR = 0.93 [0.89–0.98], P = 0.0025). Intervention impact was differential among ICUs, with the greatest effect in surgical and least in CTICU (table). nit-level AU decreased in all ICUs, driven by 4 of the 5 ICUs (table, figure).

Conclusion

The effect of ASP rounds on AU was mixed for different types of ICUs. The direct effect on AU (intervention vs. control) was small because the analysis addressed the whole ICU population and thus was subject to biases from exposures after an ICU stay, ineligible patients, and lack of blinding. However, we observed an overall decline in AU during the study period, which we believe represents indirect effects of increased ASP activity and awareness. Additional ASP resources to round more than weekly may result in greater effect.

Disclosures

All Authors: No reported Disclosures.

1018. Using prediction modeling to inform risk-adjustment strategy for hospital antimicrobial use: Can we predict who gets an inpatient antimicrobial?

Background

Hospital antimicrobial stewardship program (ASP) assessments based on comparisons of antimicrobial use (AU) among multiple hospitals are difficult to interpret without risk-adjustment for patient case-mix. We aimed to determine whether variables of varying complexity, derived retrospectively from the electronic health record (EHR), were predictive of inpatient antimicrobial exposures.

Methods

We performed a retrospective study of EHR-derived data from adult and pediatric inpatients within the Duke University Health System from October 2015 to September 2017. We used Random Forests machine learning models on two antimicrobial exposure outcomes at the encounter level: binary (ever/never) exposure and days of therapy (DOT). Antimicrobial groups were defined by the NHSN AU Option 2017 baseline. Analyses were stratified by pediatric/adult, location type (ICU/ward), and antimicrobial group. Candidate variables were categorized into four tiers based on feasibility of measurement from the EHR. Tier 1 (easy) included demographics, season, location, while Tier 4 (hard) included all variables from Tier 1–3 and laboratory results, vital signs, and culture data. Data were split into 80/20 training and testing sets to measure model performance using area under the curve (AUC) for the binary outcomes and absolute error for DOT.

Results

The analysis dataset included 170,294 encounters and 204 candidate variables from three hospitals. A total of 80,190 (47%) encounters had antimicrobial exposure; 64,998 (38%) had 1–6 DOT, and 15,192 (9%) had 7 or greater DOT. Models strongly predicted the binary outcome, with AUCs ranging from 0.70 to 0.95 depending on the stratum (Figure A, B). The addition of more complex variables increased accuracy (Figure Model Tiers 1–4). Model performance varied based on location and antimicrobial group. Models for infrequently used groups performed better (Figure C, D). Models underestimated DOTs of encounters with extremely long lengths of stay.

Conclusion

Models utilizing EHR-derived variables strongly predicted antimicrobial exposure. Risk-adjustment strategies incorporating measures of patient mix may provide more informative benchmark comparisons for use in Antimicrobial Stewardship Program assessments.

Disclosures

All authors: No reported disclosures.

1986. Impact of Two-Step Testing on the Diagnosis and Management of Clostridium difficile in a Multi-Hospital Healthcare System

Background

Methods

Results

Conclusion

Disclosures

996. Impact of Penicillin Allergy Labels on Carbapenem Use in a Multi-Center Study

Background

Antibiotic allergy labels lead to excess exposure to broad-spectrum antibiotics and can result in patient harm. We aimed to describe the prevalence of penicillin allergy labels (PAL) across a variety of hospital settings and its association with carbapenem exposure.

Methods

We performed a retrospective cohort analysis of inpatient admissions from 14 hospitals in the Duke Antimicrobial Stewardship Outreach Network (DASON) and Duke Health System from 2016 to 2018. Data were collected from the DASON central database which is derived from electronic health record extracts. PAL was defined from drug allergy documentation indicating any reaction to penicillin or its related agents, but did not include labels for other β-lactam agents (e.g., cephalosporin). Carbapenem exposure was defined as a binary variable indicating receipt of at least one dose of meropenem, ertapenem, doripenem or imipenem on an inpatient unit. The association between PAL and carbapenem exposure was assessed using multivariable logistical regression with candidate covariates including age, gender, comorbidity score, and exposure to intensive care or hematology/oncology unit. Hospital-level PAL prevalence was defined as the percentage of inpatient admissions. Hospital-level carbapenem use rates were assessed as days of therapy (DOT) per 1000 patient-days and stratified by PAL to understand the portion of use associated with PAL.

Results

Of the 727,168 admissions included in this study, 84,033 (11.6%) patients had a PAL. The majority of admissions with documented PAL were in patients >65 years old (47.9%, n = 40,240) and female (57.8%, n = 418,472). PAL was associated with a 2-fold higher risk of receipt of carbapenem (adjusted odds ratio 2.13, 95% CI 0.89–2.40, P < 0.0001). PAL prevalence varied among hospitals (median 14%, range 5–20%). Hospitals with antibiotic allergy-focused stewardship programs (ASP) had a similar PAL prevalence (median 13.8 vs. 15.9%, P = 0.08), but the percent of carbapenem DOT used in patients with PAL was similar (median 23% vs. 24%, P = 0.6).

Conclusion

PAL was associated with increased carbapenem exposure on the patient level. Allergy-focused ASP activities may affect PAL but it is unclear whether it reduces carbapenem use based on these observational data.

Disclosures

All authors: No reported disclosures.

1107. Impact of Routine Education and Data Feedback on the Durability of an Antimicrobial Stewardship Intervention for Outpatient Urinary Tract Infections

Background

Achieving lasting, sustainable effects in outpatient AS interventions has been a challenge for many programs. Our group observed an initial benefit of an outpatient AS intervention focused on diagnosis and management of urinary tract infections (UTIs). However, prescribing habits trended back toward baseline over time. This study aimed to evaluate the impact of routine education and comparative data feedback on the durability of an outpatient AS intervention for UTIs.

Methods

We conducted a prospective quasi-experimental study at one primary care (PC) and one urgent care (UC) clinic to evaluate the durability of an outpatient AS intervention implemented in August 2017 and November 2017, respectively. Clinicians who treated adult patients with a diagnosis of acute UTI at either clinic participated in the study. The initial intervention (phase 1) included development of clinic-specific antibiograms and UTI diagnosis and treatment guidelines. Approximately 12 months after the initial intervention, routine education along with clinic- and comparative provider-specific feedback reports were emailed to clinicians at regular intervals (phase 2). The primary outcome was percent of encounters in which first- or second-line antibiotics were prescribed. Pre- and post-intervention phase and trend changes were assessed using an interrupted time-series approach.

Results

Data were collected on 792 and 3,720 UTI encounters at PC and UC, respectively. In the 12 months after the initial intervention, rates of guideline concordance were 73% at PC and 57% at UC (Figures 1 and 2). After routine data feedback was provided for approximately 7 months at PC and 5 months at UC, rates of guideline concordance remained relatively stable at 75% for PC and 61% at UC. An initial 37% relative reduction in fluoroquinolone (FQ) use was observed during phase 1 which was further reduced by an additional 18% during phase 2.

Conclusion

Routine provision of clinic-specific feedback and peer comparisons sustained rates of guideline-concordant prescribing at two outpatient clinics. This intervention required significant resources for data analysis and delivery, but it was successful in decreasing rates of FQ prescribing and maintaining clinician engagement.

Disclosures

All authors: No reported disclosures.

2087. Electronic Capture and Feedback of Standardized Antibiotic Clinical Indications Data Among Community Hospitals

Background

Antibiotic clinical indications allow stewardship programs to assess therapy appropriateness; however, many hospitals that require antibiotic indications upon order entry lack standardized mapping of indications leading to variability in entered values. Electronic capture and feedback of standardized antibiotic clinical indications data may allow hospitals to more effectively compare indication-specific prescribing trends among facilities.

Methods

We collected antibiotic indications from electronic medication orders for 6 DASON hospitals. These indications were mapped to a list of 15 standardized indication categories created by consensus of the DASON stewardship team. To demonstrate the feasibility and utility of standardized clinical indications mapping, we evaluated agents given for the indication C. difficile infection (CDI) in 2018. Differences between the hospitals were compared with highlight the added benefit of standardized indication data in evaluating antibiotic use and adoption of local guidelines.

Results

For 249,916 antibiotic days of therapy (DOT) with an indication available, a total of 125 unique indications were reported. Of note, 3 facilities allowed more than one indication to be entered at prescriber discretion. The distribution of antibiotic DOT mapped to the standardized indication list can be seen in Figure 1. The most common indication was the other category (19.5%). These were primarily other, no additional information (47%) or empiric therapy for an unknown source of infection (17%). Additional indications in the other category included chronic obstructive pulmonary disease exacerbations and sexually transmitted infections (< 5% each). Figure 2 depicts the agents used for CDI indication between facilities. Despite universal adoption of local guidelines where oral vancomycin is the drug of choice for treating CDI, there was variability seen in vancomycin CDI DOT (range: 60 – 80% of CDI DOT).

Conclusion

Stewardship programs can implement standardized antimicrobial indications to facilitate electronic capture, feedback, and comparison and efficiently identify stewardship targets. Additionally, hospitals may use these data to explore the appropriateness of antibiotic use.

Disclosures

All authors: No reported disclosures.

1019. Defining electronic patient phenotypes to inform risk-adjustment strategies in hospital antimicrobial use comparisons

Background

Methods

Results

Conclusion

Disclosures

1044. Impact of Interdisciplinary Rounds on Antimicrobial Use at a Community Hospital

Background

Antimicrobial stewardship (AS) implementation is challenging in resource-limited settings such as smaller community hospitals that may lack dedicated personnel resources or have limited access to infectious diseases experts with dedicated time for AS. Few studies have evaluated the impact of interdisciplinary rounds as a strategy to optimize antimicrobial use (AU) in the community hospital setting.

Methods

We evaluated the impact of interdisciplinary rounds in a 280-bed acute care nonteaching, community hospital with an established ASP. The primary outcome was facility-wide antibiotic utilization pre- and post-implementation. Rounds included key healthcare personnel (hospitalists, clinical pharmacists, case managers, nurses) reviewing all patients on inpatient wards Monday through Friday, with a discussion of diagnosis, antibiotic selection, dosing, duration, and anticipated discharge plans. AU was compared for a 7-month post-intervention period (June 1, 2018–December 31, 2018) vs. similar months in 2017 based on days of therapy (DOT)/1,000 patient-days and length of therapy (LOT) per antimicrobial use admission. In addition, trends in AU for the post-intervention period were compared with the previous 17 months (January 1, 2017–May 31, 2018) using segmented binomial regression.

Results

Interdisciplinary rounds incorporating AS principles was associated with a decrease in overall AU in this facility, with a significant decrease of 16.33% (P < 0.0001) in DOT/1,000 pd in the first month and was stable (decrease of 1.1% per month, P = 0.15) thereafter (Figure 1). There was no significant change in LOT/admission after the first month of the intervention, but the trend demonstrated a 2% per month decrease (P < 0.03) thereafter (Figure 2). Comparing 2018 intervention months with similar months of 2017, the use of antibacterial agents decreased on average by 191.3 (95% CI −128.2 to −254.4) DOT/1,000 patient-days (Figure 3) and 0.546 (95% CI: −0.28 to −0.81) days per admission (Figure 4).

Conclusion

In this community hospital with an existing antimicrobial stewardship program, implementation of interdisciplinary rounds was associated with a substantial decrease in antimicrobial use. This was sustained for at least a 7-month period.

Disclosures

All authors: No reported disclosures.

Epidemiologic Trends in Clostridioides difficile Infections in a Regional Community Hospital Network

IMPORTANCE

Clostridioides difficile infection (CDI) remains a leading cause of health care facility-associated infection. A greater understanding of the regional epidemiologic profile of CDI could inform targeted prevention strategies.

OBJECTIVES

To assess trends in incidence of health care facility-associated and community-acquired CDI among hospitalized patients over time and to conduct a subanalysis of trends in the NAP1 strain of CDI over time.

DESIGN, SETTING, AND PARTICIPANTS

This long-term multicenter cohort study reviewed records of patients (N = 2 025 678) admitted to a network of 43 regional community hospitals primarily in the southeastern United States from January 1, 2013, through December 31, 2017. Generalized linear mixed-effects models were used to adjust for potential clustering within facilities and changing test method (nucleic acid amplification testing or toxin enzyme immunoassay) over time.

MAIN OUTCOMES AND MEASURES

Clostridioides difficile infection incidence rates were counted as cases per 1000 admissions for community-acquired and total CDI cases or cases per 10 000 patient-days for health care facility-associated CDI. Long-term trends in the proportion of cases acquired in the community and in NAP1 strain incidence were also evaluated.

RESULTS

A total of 2 025 678 admissions and 21 254 CDI cases were included (12 678 [59.6%] female; median [interquartile range] age, 69 [55-80] years). Median (interquartile range) total CDI incidence increased slightly from 7.9 (3.5-12.4) cases per 1000 admissions in 2013 to 9.3 (4.9-13.7) cases per 1000 admissions in 2017. After adjustment, the overall incidence of health care facility-associated CDI declined (incidence rate ratio [IRR], 0.995; 95% CI, 0.990-0.999; P = .03), whereas insufficient evidence was found for either an increase or a decrease in community-acquired CDI (IRR, 1.004; 95% CI, 0.999-1.009; P = .14). The proportion of cases classified as community acquired increased over time from a mean (SD) of 0.49 (0.28) in 2013 to 0.61 (0.26) in 2017 (odds ratio, 1.010 per month; 95% CI, 1.006-1.015; P < .001). Rates of the NAP1 strain of CDI varied widely between facilities, with no statistically significant change in NAP1 strain incidence over time in the community setting (IRR, 1.007; 95% CI, 0.994-1.021) or health care facility setting (IRR, 1.011; 95% CI, 0.990-1.032).

CONCLUSIONS AND RELEVANCE

The findings suggest that, despite the modest improvement in health care facility-associated CDI rates, a better understanding of community-acquired CDI incidence is needed for future infection prevention efforts.