Informing estimates of probability of Clostridioides difficile infection for testing and treatment: expert consensus from a modified-Delphi procedure.

Background: Clostridioides difficile infection (CDI) may be misdiagnosed if testing is performed in the absence of signs or symptoms of disease. This study sought to support appropriate testing by estimating the impact of signs, symptoms, and healthcare exposures on pre-test likelihood of CDI. Methods: A panel of fifteen experts in infectious diseases participated in a modified UCLA/RAND Delphi study to estimate likelihood of CDI. Consensus, defined as agreement by >70% of panelists, was assessed via a REDCap survey. Items without consensus were discussed in a virtual meeting followed by a second survey. Results: All fifteen panelists completed both surveys (100% response rate). In the initial survey, consensus was present on 6 of 15 (40%) items related to risk of CDI. After panel discussion and clarification of questions, consensus (>70% agreement) was reached on all remaining items in the second survey. Antibiotics were identified as the primary risk factor for CDI and grouped into three categories: high-risk (likelihood ratio [LR] 7, 93% agreement among panelists in first survey), low-risk (LR 3, 87% agreement in first survey), and minimal-risk (LR 1, 71% agreement in first survey). Other major factors included new or unexplained severe diarrhea (e.g., ≥ 10 liquid bowel movements per day; LR 5, 100% agreement in second survey) and severe immunosuppression (LR 5, 87% agreement in second survey). Conclusion: Infectious disease experts concurred on the importance of signs, symptoms, and healthcare exposures for diagnosing CDI. The resulting risk estimates can be used by clinicians to optimize CDI testing and treatment.

Effect of automated identification of antimicrobial stewardship opportunities for suspected urinary tract infections.

Objective: We aimed to determine whether automated identification of antibiotic targeting suspected urinary tract infection (UTI) shortened the time to antimicrobial stewardship (AS) intervention. Design: Retrospective before-and-after study. Setting: Tertiary and quaternary care academic medical center. Patients: Emergency department (ED) or admitted adult patients meeting best practice alert (BPA) criteria during pre- and post-BPA periods. Methods: We developed a BPA to alert AS pharmacists of potential ASB triggered by the following criteria: ED or admitted status, antibiotic order with genitourinary indication, and a preceding urinalysis with ≤ 10 WBC/hpf. We evaluated the median time from antibiotic order to AS intervention and overall percent of UTI-related interventions among patients in pre-BPA (01/2020-12/2020) and post-BPA (04/15/2021-04/30/2022) periods. Results: 774 antibiotic orders met inclusion criteria: 355 in the pre- and 419 in the post-BPA group. 43 (35 UTI-related) pre-BPA and 117 (94 UTI-related) post-BPA interventions were documented. The median time to intervention was 28 hours (IQR 18-65) in the pre-BPA group compared to 16 hours (IQR 2-34) in the post-BPA group (P < 0.01). Despite absent pyuria, there were six cases with gram-negative bacteremia presumably from a urinary source. Conclusions: Automated identification of antibiotics targeting UTI without pyuria on urinalysis reduced the time to stewardship intervention and increased the rate of UTI-specific interventions. Clinical decision support aided in the efficiency of AS review and syndrome-targeted impact, but cases still required AS clinical review.

Beyond Infection: Mortality and End-of-Life Care Associated With Infectious Disease Consultation in an Academic Health System.

BACKGROUND: Infectious diseases (ID) physicians are increasingly faced with the challenge of caring for patients with terminal illnesses or incurable infections. METHODS: This was a retrospective cohort of all patients with an ID consult within an academic health system from 1 January 2014 through 31 December 2023, including community, general, and transplant ID consult services. RESULTS: There were 60 820 inpatient ID consults (17 235 community, 29 999 general, and 13 586 transplant) involving 37 848 unique patients. The number of consults increased by 94% and the rate rose from 5.0 to 9.9 consults per 100 inpatients (P < .001). In total, 7.5% of patients receiving an ID consult died during admission and 1006 (2.6%) of patients were discharged to hospice. In-hospital mortality was 5.2% for community ID, 7.8% for general ID, and 10.7% for transplant ID patients (P < .001). Six-month mortality was 9% for all nonobstetric admissions versus 19% for community ID, 20.9% for general ID, and 22.3% for transplant ID. In total 2866 (7.6%) of all patients receiving ID consultation also received palliative care consultation during the same hospitalization. The index ID consult preceded any palliative consult in the majority (69.5%) of cases. A total of 16.3% of patients had a do-not-resuscitate order during the index hospitalization; 12.2% of all patients with a do-not-resuscitate order had this placed on the same day as the ID consult. CONCLUSIONS: Patients receiving ID consultation were increasingly complex and more likely to die soon after consultation. These results provide a framework for ID clinicians to consider their role in end-of-life care.

Late surgical site infections among solid organ transplant recipients: an unrecognized clinical entity.

This study identified 26 late invasive primary surgical site infection (IP-SSI) within 4-12 months of transplantation among 2073 SOT recipients at Duke University Hospital over the period 2015-2019. Thoracic organ transplants accounted for 25 late IP-SSI. Surveillance for late IP-SSI should be maintained for at least one year following transplant.

Risk Factors for and Outcomes Following Early Acquisition of Mycobacterium abscessus Complex After Lung Transplantation.

Background: Lung transplant recipients are at increased risk of Mycobacterium abscessus complex (MABC) acquisition and invasive infection. We analyzed risk factors and outcomes of early post-lung transplant MABC acquisition. Methods: We conducted a retrospective matched case-control study of patients who underwent lung transplant from 1/1/2012 to 12/31/2021 at a single large tertiary care facility. Cases had de novo MABC isolation within 90 days post-transplant. Controls had no positive MABC cultures and were matched 3:1 with cases based on age and transplant date. Recipient demographics and pre-/peri-operative characteristics were analyzed, and a regression model was used to determine independent risk factors for MABC acquisition. We also assessed 1-year post-transplant outcomes, including mortality. Results: Among 1145 lung transplants, we identified 79 cases and 237 matched controls. Post-transplant mechanical ventilation for >48 hours was independently associated with MABC acquisition (adjusted odds ratio, 2.46; 95% CI, 1.29-4.72; P = .007). Compared with controls, cases required more days of hospitalization after the MABC index date (28 vs 12 days; P = .01) and had decreased 1-year post-transplant survival (78% vs 89%; log-rank P = .02). One-year mortality appeared highest for cases who acquired M. abscessus subsp. abscessus (31% mortality) or had extrapulmonary infections (43% mortality). Conclusions: In this large case-control study, prolonged post-transplant ventilator duration was associated with early post-lung transplant MABC acquisition, which in turn was associated with increased hospital-days and mortality. Further studies are needed to determine the best strategies for MABC prevention, surveillance, and management.

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.

The devil's in the defaults: An interrupted time-series analysis of the impact of default duration elimination on exposure to fluoroquinolone therapy.

OBJECTIVE: To determine whether removal of default duration, embedded in electronic prescription (e-script), influenced antibiotic days of therapy. DESIGN: Interrupted time-series analysis. SETTING: The study was conducted across 2 community hospitals, 1 academic hospital, 3 emergency departments, and 86 ambulatory clinics. PATIENTS: Adults prescribed a fluoroquinolone with a duration <31 days. INTERVENTIONS: Removal of standard 10-day fluoroquinolone default duration and addition of literature-based duration guidance in the order entry on December 19, 2017. The study period included data for 12 months before and after the intervention. RESULTS: The study included 35,609 fluoroquinolone e-scripts from the preintervention period and 31,303 fluoroquinolone e-scripts from the postintervention period, accounting for 520,388 cumulative fluoroquinolone DOT. Mean durations before and after the intervention were 7.8 (SD, 4.3) and 7.7 (SD, 4.5), a nonsignificant change. E-scripts with a 10-day duration decreased prior to and after the default removal. The inpatient setting showed a significant 8% drop in 10-day e-scripts after default removal and a reduced median duration by 1 day; 10-day scripts declined nonsignificantly in ED and ambulatory settings. In the ambulatory settings, both 7- and 14-day e-script durations increased after default removal. CONCLUSION: Removal of default 10-day antibiotic durations did not affect overall mean duration but did shift patterns in prescribing, depending on practice setting. Stewardship interventions must be studied in the context of practice setting. Ambulatory stewardship efforts separate from inpatient programs are needed because interventions cannot be assumed to have similar effects.

Risk factors, management, and clinical outcomes of invasive Mycoplasma and Ureaplasma infections after lung transplantation.

Mollicute infections, caused by Mycoplasma and Ureaplasma species, are serious complications after lung transplantation; however, understanding of the epidemiology and outcomes of these infections remains limited. We conducted a single-center retrospective study of 1156 consecutive lung transplants performed from 2010-2019. We used log-binomial regression to identify risk factors for infection and analyzed clinical management and outcomes. In total, 27 (2.3%) recipients developed mollicute infection. Donor characteristics independently associated with recipient infection were age ≤40 years (prevalence rate ratio [PRR] 2.6, 95% CI 1.0-6.9), White race (PRR 3.1, 95% CI 1.1-8.8), and purulent secretions on donor bronchoscopy (PRR 2.3, 95% CI 1.1-5.0). Median time to diagnosis was 16 days posttransplant (IQR: 11-26 days). Mollicute-infected recipients were significantly more likely to require prolonged ventilatory support (66.7% vs 21.4%), undergo dialysis (44.4% vs 6.3%), and remain hospitalized ≥30 days (70.4% vs 27.4%) after transplant. One-year posttransplant mortality in mollicute-infected recipients was 12/27 (44%), compared to 148/1129 (13%) in those without infection (P <.0001). Hyperammonemia syndrome occurred in 5/27 (19%) mollicute-infected recipients, of whom 3 (60%) died within 10 weeks posttransplant. This study highlights the morbidity and mortality associated with mollicute infection after lung transplantation and the need for better screening and management protocols.

Disparities in Mpox Vaccination Among Priority Populations During the 2022 Outbreak.

Background: The 2022 mpox outbreak disproportionately affected men who have sex with men and persons living with HIV (PLWH). A 2-dose mpox vaccine series was deployed in mid-2022. Structural racism and insurance status may have affected equitable vaccination. Methods: We defined 3 cohorts: PLWH with at least 1 clinic visit between 1 July 2021 and 1 July 2022 (n = 2066), HIV preexposure prophylaxis (PrEP) recipients as of 1 January 2022 (n = 262), and all mpox-vaccinated patients in our health system between 1 July 2022 and 30 November 2022 (n = 807). We identified patients with prior diagnosed sexually transmitted infections (STIs) as having a positive test result for gonorrhea, chlamydia, or syphilis between 1 July 2021-1 July 2022. The primary outcome was receipt of at least 1 dose of mpox vaccine. Results: We identified 224 (10.8%) PLWH and 50 (19.0%) PrEP patients who received at least 1 dose of mpox vaccine. Among PLWH, White race (odds ratio [OR], 1.55; 95% CI, 1.11-2.16), private insurance (OR, 1.83; 95% CI, 1.01-3.34), prior STI (OR, 3.04; 95% CI, 2.16-4.27), prior COVID-19 vaccination (OR, 3.17; 95% CI, 1.93-5.20), and prior influenza vaccination (OR, 1.42; 95% CI, 1.30-1.96) independently predicted mpox vaccination. Within the PrEP cohort, prior COVID-19 vaccination and seasonal influenza vaccination predicted mpox vaccination. Uninsured patients were vaccinated later in the outbreak than patients with private insurance (median time to vaccination, 41 days in the privately insured group vs 83 days in the uninsured group; P < .0001). Conclusions: Race, insurance status, prior STI, and previous receipt of other vaccines influenced uptake of mpox vaccine. Addressing health disparities and vaccine acceptance will be essential in improving future outbreak response.

Comparing Natural Language Processing and Structured Medical Data to Develop a Computable Phenotype for Patients Hospitalized Due to COVID-19: Retrospective Analysis.

Background: Throughout the COVID-19 pandemic, many hospitals conducted routine testing of hospitalized patients for SARS-CoV-2 infection upon admission. Some of these patients are admitted for reasons unrelated to COVID-19 and incidentally test positive for the virus. Because COVID-19-related hospitalizations have become a critical public health indicator, it is important to identify patients who are hospitalized because of COVID-19 as opposed to those who are admitted for other indications. Objective: We compared the performance of different computable phenotype definitions for COVID-19 hospitalizations that use different types of data from electronic health records (EHRs), including structured EHR data elements, clinical notes, or a combination of both data types. Methods: We conducted a retrospective data analysis, using clinician chart review-based validation at a large academic medical center. We reviewed and analyzed the charts of 586 hospitalized individuals who tested positive for SARS-CoV-2 in January 2022. We used LASSO (least absolute shrinkage and selection operator) regression and random forests to fit classification algorithms that incorporated structured EHR data elements, clinical notes, or a combination of structured data and clinical notes. We used natural language processing to incorporate data from clinical notes. The performance of each model was evaluated based on the area under the receiver operator characteristic curve (AUROC) and an associated decision rule based on sensitivity and positive predictive value. We also identified top words and clinical indicators of COVID-19-specific hospitalization and assessed the impact of different phenotyping strategies on estimated hospital outcome metrics. Results: Based on a chart review, 38.2% (224/586) of patients were determined to have been hospitalized for reasons other than COVID-19, despite having tested positive for SARS-CoV-2. A computable phenotype that used clinical notes had significantly better discrimination than one that used structured EHR data elements (AUROC: 0.894 vs 0.841; P<.001) and performed similarly to a model that combined clinical notes with structured data elements (AUROC: 0.894 vs 0.893; P=.91). Assessments of hospital outcome metrics significantly differed based on whether the population included all hospitalized patients who tested positive for SARS-CoV-2 or those who were determined to have been hospitalized due to COVID-19. Conclusions: These findings highlight the importance of cause-specific phenotyping for COVID-19 hospitalizations. More generally, this work demonstrates the utility of natural language processing approaches for deriving information related to patient hospitalizations in cases where there may be multiple conditions that could serve as the primary indication for hospitalization.

Impact of Policy Change on Access to Medication for Opioid Use Disorder in Primary Care.

OBJECTIVES: The opioid overdose epidemic is escalating. Increasing access to medications for opioid use disorder in primary care is crucial. The impact of the US Department of Health and Human Services' policy change removing the buprenorphine waiver training requirement on primary care buprenorphine prescribing remains unclear. We aimed to investigate the impact of the policy change on primary care providers' likelihood of applying for a waiver and the current attitudes, practices, and barriers to buprenorphine prescribing in primary care. METHODS: We used a cross-sectional survey with embedded educational resources disseminated to primary care providers in a southern US academic health system. We used descriptive statistics to aggregate survey data, logistic regression models to evaluate whether buprenorphine interest and familiarity correlate with clinical characteristics, and a χ2 test to evaluate the effect of the educational intervention on screening. RESULTS: Of the 54 respondents, 70.4% reported seeing patients with opioid use disorder, but only 11.1% had a waiver to prescribe buprenorphine. Few nonwaivered providers were interested in prescribing, but perceiving buprenorphine to be beneficial to the patient population was associated with interest (adjusted odds ratio 34.7, P < 0.001). Two-thirds of nonwaivered respondents reported the policy change having no impact on their decision to obtain a waiver; however, among interested providers, it increased their likelihood of obtaining a waiver. Barriers to buprenorphine prescribing included lack of clinical experience, clinical capacity, and referral resources. Screening for opioid use disorder did not increase significantly after the survey. CONCLUSIONS: Although most primary care providers reported seeing patients with opioid use disorder, interest in prescribing buprenorphine was low and structural barriers remained the dominant obstacles. Providers with a preexisting interest in buprenorphine prescribing reported that removing the training requirement was helpful.

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 hospital blood culture utilization rates to identify opportunities for diagnostic stewardship.

OBJECTIVES: To evaluate the pattern of blood-culture utilization among a cohort of 6 hospitals to identify potential opportunities for diagnostic stewardship. METHODS: We completed a retrospective analysis of blood-culture utilization during adult inpatient or emergency department (ED) encounters in 6 hospitals from May 2019 to April 2020. We investigated 2 measures of blood-culture utilization rates (BCURs): the total number of blood cultures, defined as a unique accession number per 1,000 patient days (BCX) and a new metric of blood-culture events per 1,000 patient days to account for paired culture practices. We defined a blood-culture event as an initial blood culture and all subsequent samples for culture drawn within 12 hours for patients with an inpatient or ED encounter. Cultures were evaluated by unit type, positivity and contamination rates, and other markers evaluating the quality of blood-culture collection. RESULTS: In total, 111,520 blood cultures, 52,550 blood culture events, 165,456 inpatient admissions, and 568,928 patient days were analyzed. Overall, the mean BCUR was 196 blood cultures per 1,000 patient days, with 92 blood culture events per 1,000 patient days (range, 64-155 among hospitals). Furthermore, 7% of blood-culture events were single culture events, 55% began in the ED, and 77% occurred in the first 3 hospital days. Among all blood cultures, 7.7% grew a likely pathogen, 2.1% were contaminated, and 5.9% of first blood cultures were collected after the initiation of antibiotics. CONCLUSIONS: Blood-culture utilization varied by hospital and was heavily influenced by ED culture volumes. Hospital comparisons of blood-culture metrics can assist in identifying opportunities to optimize blood-culture collection practices.

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.

Measuring empiric antibiotic spectrum-A journey through space and time.

OBJECTIVES: The typical 5-day work week affects healthcare outcomes. Structured work hours have also been implicated in antimicrobial prescribing choice. We developed a visualization tool to aid in evaluating breadth of antibiotic use in various time (day of week and hour of day) and space (patient location) combinations. METHODS: We evaluated antibiotic administration data from a tertiary-care academic medical center between July 1, 2018, and July 1, 2020. We calculated a cumulative empiric antibiotic spectrum score by adapting a previously validated antibiotic spectrum index (ASI) and applying that score to empiric antibiotic use. We visualized these data as a heat map based on various day-of-week-time combinations and then compared the distribution of scores between weekday nights, weekend days, and weekend nights to the typical workweek hours (weekday days, weekday days) using the Mann-Whitney U nonparametric test with a Bonferroni correction. RESULTS: The analysis included 76,535 antibiotic starts across 53,900 unique patient admissions over 2 years. The mean cumulative ASI was higher in all 3 night and weekend combinations (weekday nights, 7.3; weekend days, 7.6; weekend nights, 7.5) compared to the weekday daytime hours (weekday days, 7.1) and the distribution of scores was different in all groups compared to the weekday daytime reference. The cumulative ASI was also higher in intensive care units. CONCLUSIONS: Empiric antibiotic prescribing patterns differed across space and time; broader antibiotic choices occurred in the intensive care units and on nights and weekends. Visualization of these patterns aids in antimicrobial prescribing pattern recognition and may assist in finding opportunities for additional antimicrobial stewardship interventions.

Curating a knowledge base for individuals with coinfection of HIV and SARS-CoV-2: a study protocol of EHR-based data mining and clinical implementation.

INTRODUCTION: Despite a higher risk of severe COVID-19 disease in individuals with HIV, the interactions between SARS-CoV-2 and HIV infections remain unclear. To delineate these interactions, multicentre Electronic Health Records (EHR) hold existing promise to provide full-spectrum and longitudinal clinical data, demographics and sociobehavioural data at individual level. Presently, a comprehensive EHR-based cohort for the HIV/SARS-CoV-2 coinfection has not been established; EHR integration and data mining methods tailored for studying the coinfection are urgently needed yet remain underdeveloped. METHODS AND ANALYSIS: The overarching goal of this exploratory/developmental study is to establish an EHR-based cohort for individuals with HIV/SARS-CoV-2 coinfection and perform large-scale EHR-based data mining to examine the interactions between HIV and SARS-CoV-2 infections and systematically identify and validate factors contributing to the severe clinical course of the coinfection. We will use a nationwide EHR database in the USA, namely, National COVID Cohort Collaborative (N3C). Ultimately, collected clinical evidence will be implemented and used to pilot test a clinical decision support prototype to assist providers in screening and referral of at-risk patients in real-world clinics. ETHICS AND DISSEMINATION: The study was approved by the institutional review boards at the University of South Carolina (Pro00121828) as non-human subject study. Study findings will be presented at academic conferences and published in peer-reviewed journals. This study will disseminate urgently needed clinical evidence for guiding clinical practice for individuals with the coinfection at Prisma Health, a healthcare system in collaboration.

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 1 AU trend (1-12 months) and AU trend following the first year (13-42 months) were compared using relative rates (RR). Monthly AU rates were measured in days of therapy (DOT) per 1000 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]: .91-.99). Large variation was seen in hospital-specific AU. Fluoroquinolone use was stable during year 1 and then dropped significantly. Other agent groups demonstrated a nonsignificant downward trajectory after year 1. 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.

A modified Delphi approach to develop a trial protocol for antibiotic de-escalation in patients with suspected sepsis.

Background: Early administration of antibiotics in sepsis is associated with improved patient outcomes, but safe and generalizable approaches to de-escalate or discontinue antibiotics after suspected sepsis events are unknown. Methods: We used a modified Delphi approach to identify safety criteria for an opt-out protocol to guide de-escalation or discontinuation of antibiotic therapy after 72 hours in non-ICU patients with suspected sepsis. An expert panel with expertise in antimicrobial stewardship and hospital epidemiology rated 48 unique criteria across 3 electronic survey rating tools. Criteria were rated primarily based on their impact on patient safety and feasibility for extraction from electronic health record review. The 48 unique criteria were rated by anonymous electronic survey tools, and the results were fed back to the expert panel participants. Consensus was achieved to either retain or remove each criterion. Results: After 3 rounds, 22 unique criteria remained as part of the opt-out safety checklist. These criteria included high-risk comorbidities, signs of severe illness, lack of cultures during sepsis work-up or antibiotic use prior to blood cultures, or ongoing signs and symptoms of infection. Conclusions: The modified Delphi approach is a useful method to achieve expert-level consensus in the absence of evidence suifficient to provide validated guidance. The Delphi approach allowed for flexibility in development of an opt-out trial protocol for sepsis antibiotic de-escalation. The utility of this protocol should be evaluated in a randomized controlled trial.