Susan L. Davis

Objectives: To compare outcomes of oral switch versus intravenous antibiotics for the treatment of carbapenem‐resistant Enterobacterales (CRE) and Pseudomonas aeruginosa (CRPA) infections at hospital discharge. Methods: Institutional review board approved, retrospective cohort of adults infected with CRE or CRPA who received oral switch or intravenous antibiotics at hospital discharge from January 1, 2017, to April 30, 2024. Patients were included if they were eligible for oral switch and infected with an isolate susceptible to one or more oral antibiotics; non‐bacteremic urinary tract infections were excluded. The primary outcome was 30‐day clinical success at end of therapy, defined as lack of infection‐related hospitalization, infection‐related recurrence, or change/escalation of therapy. Secondary outcomes included hospital length of stay (LOS) and 30‐day all‐cause mortality from end of therapy. Results: Fifty‐five patients were included; 51% received oral switch antibiotics and 49% received intravenous antibiotics. Thirty‐three percent of patients had CRE, 67% had CRPA, and 38% of cultures were polymicrobial. The most common infection types were pneumonia (33%), intra‐abdominal (26%), and bone/joint (22%). The median (interquartile range [IQR]) duration of outpatient therapy was 12 (6–25) days versus 20 (4–34) days for the oral switch and intravenous antibiotic groups, respectively ( p = 0.341). 30‐day clinical success was 61% in the oral switch and 48% in the intravenous antibiotic groups ( p = 0.349); the median (IQR) hospital LOS for the oral switch and intravenous antibiotic groups was 14 (9–25) days and 16 (9–49) days, respectively ( p = 0.165); 30‐day mortality was 4% in the oral switch group and 15% in the intravenous antibiotic group ( p = 0.193). Conclusion: A limited sample of patients who received oral switch antibiotics had similar outcomes to intravenous outpatient treatment of carbapenem‐resistant organisms, with a shorter hospital LOS.

Candidemia is associated with mortality rates exceeding 40%. However, prior studies indicate mortality may be reduced when antifungal therapy is initiated within 12 hours. The T2Candida Panel is a diagnostic assay that detects Candida species directly from a whole blood specimen within 3-5 hours (T2 Biosystems, Lexington, MA). The objective of this study is to identify predictors of 30-day mortality in patients with candidemia identified by T2Candida Panel. This is a retrospective, multicenter study of critically ill patients with candidemia identified by T2Candida Panel from January 2016 - December 2022. Critically ill patients were defined as those who developed candidemia during an intensive care unit (ICU) stay or within 72 hours of ICU admission or discharge. T2Candida sites were chosen across the United States based on T2Candida utilization. Exclusion criteria were patients < 18 years of age, those with prophylactic indications for antifungal therapy, prisoners and pregnant patients. Multivariate logistic regression was conducted to identify factors associated with 30-day mortality measured from the T2Candida draw time. There were 171 ICU patients from seven institutions with candidemia identified by T2Candida panel. The mean (standard deviation [SD]) age was 59.7 (14.8) years and 52.1% were male. Mean (SD) APACHE II and Charlson Comorbidity Index scores were 20.6 (7.1) and 4.9 (2.8), respectively. Empiric antifungal therapy was administered to 36.8% of patients and the majority received infectious diseases (ID) consult (92.4%). Echinocandins were the most common agents used for empiric (72.7%) and definitive therapy (62.6%). Overall, 30-day mortality occurred in 36.0% and was not associated with antifungal de-escalation. Administration of empiric therapy (aOR 0.457, 95% CI 0.199-1.054) and ID consult (aOR 0.225, 95% CI 0.056-0.913) were associated with reduced odds of 30-day mortality. Empiric antifungal administration and ID consult were independently associated with reduced odds of 30-day mortality in patients with candidemia identified by T2Candida Panel. Future studies are needed to evaluate the impact of the T2Candida panel on antifungal stewardship. Ryan K. Shields, PharmD, MS, Allergan: Advisor/Consultant|Cidara: Advisor/Consultant|Entasis: Advisor/Consultant|GSK: Advisor/Consultant|Melinta: Advisor/Consultant|Melinta: Grant/Research Support|Menarini: Advisor/Consultant|Merck: Advisor/Consultant|Merck: Grant/Research Support|Pfizer: Advisor/Consultant|Roche: Grant/Research Support|Shionogi: Advisor/Consultant|Shionogi: Grant/Research Support|Utility: Advisor/Consultant|Venatorx: Advisor/Consultant|Venatorx: Grant/Research Support Michael J. Rybak, PharmD, PhD, MPH, Abbvie, Merck, Paratek, Shionogi, Entasis, La Jolla, T2 Biosystems: Advisor/Consultant

Objectives: The Centers for Medicare and Medicaid Services Severe Sepsis and Septic Shock Management Bundle (SEP‐1) assesses antibiotic administration, lactate measurement, and blood culture collection within 3 h of severe sepsis onset. The impact of the SEP‐1 3‐hour bundle among patients with severe sepsis is not extensively described. This investigation aimed to describe the impact of 3‐hour bundle compliance on 28‐day in‐hospital mortality in patients with severe sepsis.Study Design: This was a retrospective, propensity adjusted, nested case‐control study assessing the impact of compliance with a 3‐hour sepsis bundle among patients with severe sepsis.Setting: This study was conducted at a large, academic, tertiary care medical center in Detroit, Michigan from July 1, 2017 to December 31, 2019.Patients: Cases were defined as those suffering 28‐day in‐hospital mortality. Controls were defined as those surviving at or discharged by 28 days. Patients were separated based on 3‐hour bundle compliance or noncompliance. Nested and overall cohorts were assessed. Severe sepsis time zero was manually validated. Patients with shock, requiring vasopressors within 8 h of time zero, or those not meeting SEP‐1 inclusion criteria were excluded.Intervention: The primary outcome was the propensity adjusted odds of 28‐day in‐hospital mortality among 3‐hour bundle compliant versus noncompliant patients. Secondary outcomes included mortality for individual bundle element compliance, progression to septic shock, and predictors of mortality according to logistic regression.Results: A total of 325 compliant and 325 noncompliant patients were included. The median Sequential Organ Failure Assessment (SOFA) score was three in each group. There was no difference in propensity adjusted odds of mortality among those compliant versus noncompliant with the 3‐hour bundle (odds‐ratio [OR] 1.039; 95% CI: 0.721–1.497; p = 0.838) or with individual bundle elements. SOFA score and female sex were predictors of mortality.Conclusions: Three‐hour bundle compliance did not impact 28‐day in‐hospital mortality in patients with severe sepsis. Further research is needed to understand the impact of 3‐hour bundle compliance on mortality in severe sepsis.

In the US, 1 in 5 hospitalized adults is prescribed an antimicrobial at the time of discharge, accounting for millions of antimicrobial-days each year.¹ Although prescribers face numerous patient-centered challenges during transitions of care (TOC), resources are sorely needed for antimicrobial stewardship throughout this period.² In a statewide collaborative effort, the Michigan Hospital Medicine Safety Consortium classified the appropriateness of more than 20 000 antimicrobials prescribed at discharge, and nearly half were considered overuse.³ The most common antimicrobial class prescribed at discharge, the fluoroquinolone, has a number of box warnings and safety concerns, yet as few as 25% of these orders are considered appropriate.^1,4,5 Prolonged duration of antimicrobial therapy is consistently the major contributor to inappropriate prescribing during TOC, and a mean of 40% of each patients’ total duration of therapy is administered post discharge.^1,3,6 The review of antimicrobial prescribing during TOC represents a crucial moment in a patient’s clinical course to ensure safe, effective, and guideline-concordant therapy.

Over half of antimicrobials ordered at hospital discharge are not optimized, many of which having longer than necessary durations or inappropriate for the indication. Unnecessary antimicrobial exposures increase the risks of adverse events, antibiotic resistance, and Clostridioides difficile infections. However, discharge prescribing often escapes the purview of inpatient antimicrobial stewardship programs and few interventions have been implemented to optimize antibiotic use during transitions of care (TOC). Herein, the aim was to highlight critical steps and challenges in the implementation of a TOC antimicrobial stewardship program designed to improve prescribing at hospital discharge. In a five‐hospital health system, a pharmacist‐led intervention was implemented during TOC, with the objective of improving oral antibiotic selection and duration. Among a multidisciplinary team of physicians, case managers, and nurses, the pharmacists engaged in three strategies: (a) early identification of patients to be discharged on oral antibiotics; (b) collaborative planning and communication regarding guideline‐recommended antibiotic selection and duration; and (c) facilitation of discharge antibiotic prescription with appropriate stop date. Barriers to completing the intervention on each patient in this experience included: timely identification of eligible patients prior to discharge, stewardship periods of reduced staffing during evenings and weekends, and onboarding of new staff and trainees to the process. Other major challenges that stewardship and pharmacy departments will face include adoption of best‐practice guidelines for discharge, obtaining support from local physician champions, extending the intervention to various service lines and departments with limited resources, and engaging stakeholders to support the program and intervention. This experience demonstrated that pharmacist‐led antimicrobial stewardship at discharge can be successful in both academic and community settings.

Resource-intensive interventions and education are susceptible to a lack of long-term sustainability and regression to the mean. The respiratory culture nudge changed reporting to “Commensal Respiratory Flora only: No S. aureus/MRSA or P. aeruginosa.” This study demonstrated sustained reduction in broad-spectrum antibiotic duration and long-term sustainability 3 years after implementation.

Trimethoprim-sulfamethoxazole (TMP-SMX) is a high-bioavailability antibiotic associated with potentially serious adverse drug events (ADE). The objective of this study was to evaluate the safety of intravenous (IV) and oral (PO) high-dose TMP-SMX. IRB-approved retrospective cohort of hospitalized patients from January 2016 to November 2020. Inclusion: ≥ 18 years old and > 72 hours of renally adjusted high-dose TMP-SMX defined as ≥ 5 mg/kg/day of TMP. Exclusion: prophylaxis. Endpoints during treatment: hyponatremia with sodium < 135 mmol/L, hyperkalemia with potassium > 5 mmol/L, serum creatinine increase of ≥ 0.3 mg/dL or 1.5-1.9 times from baseline, and fluid overload on physical exam. Descriptive and bivariate statistics were performed. Each group included 50 patients (Table 1). Intensive care unit patients comprised 82% IV TMP-SMX compared to 32% PO. Most common infection: respiratory tract 86% IV and 68.1% PO. Most common organisms were Stenotrophomonas maltophilia (52% IV and 18% PO) and Pneumocystis jiroveci (16.3% IV and 62% PO). Median (IQR) days of inpatient therapy: 6 (5-7.5) PO vs. 7.5 (6-11.3) IV. Median (IQR) days of total duration: 9 (6-21.5) PO vs. 12 (7.8-14) IV (p=0.93). IV group: 88% of patients received >1 liter of D5W daily. Median (IQR) liters of D5W daily was 1 (1-1.5). 56% had a diuretic added, and 38% had a diuretic dose increase. Majority of patients (78%) on IV were taking other oral medications. 100% patients experienced any adverse event with IV vs. 70% with PO (unAdjOR 2.43; 95% CI 1.89-3.13). Most common ADE in both groups: hyponatremia, hyperkalemia, and elevated creatinine. Hyponatremia: 92% with IV and 32% with PO (unAdjOR 24.44; 95% CI 7.50-79.68). Edema on physical exam, an ADE specific to IV TMP-SMX, was the third most common side effect in the IV group. Relative changes from baseline in sodium, potassium, and creatinine from those who experienced hyponatremia, hyperkalemia and elevated creatinine were listed in Table 2. Table 1. Baseline and Clinical Characteristics Table 2. Adverse Effects Patients on IV TMP-SMX therapy were more likely to experience an ADE compared to PO, likely driven by the high volume of free water. Most patients on IV TMP-SMX were on other PO medications, suggesting a missed stewardship opportunity for IV to PO conversion to reduce patient harm. Susan L. Davis, PharmD, Nothing to disclose Michael P. Veve, Pharm.D., Cumberland (Grant/Research Support)Paratek Pharmaceuticals (Research Grant or Support) Rachel Kenney, PharmD, Medtronic, Inc. (Other Financial or Material Support, spouse is an employee and shareholder)

Our antimicrobial stewardship program identified high rates of suboptimal metronidazole prescribing for Clostridioides difficile infection (CDI) within ambulatory clinics. An outpatient best practice advisory (BPA) was implemented to notify prescribers “Vancomycin or fidaxomicin are preferred over metronidazole for C.difficile infection” when metronidazole was prescribed to a patient with CDI. We conducted an IRB approved quasi-experiment before and after implementation of the BPA on June 3, 2020. Inclusion: Adult patients diagnosed with and treated for a first episode of symptomatic CDI at an ambulatory clinic between 11/1/2019 and 11/30/2020. Exclusion: fulminant CDI. Primary endpoint: guideline-concordant CDI therapy, defined as oral vancomycin or fidaxomicin. Oral metronidazole was considered guideline-concordant if prescribed due to cost barrier. Secondary endpoints: reasons for alternative CDI therapy, patient outcomes, prescriber response to the BPA. Descriptive and bivariate analyses were completed. 189 patients were included in the study, 92 before and 97 after the BPA. Median age: 59 years, 31% male, 75% Caucasian, 30% with CDI-related comorbidities, 35% with healthcare exposure, 65% with antibiotic exposure, 44% with gastric acid suppression therapy within 90 days of CDI diagnosis. The BPA was accepted 23 out of 26 times and optimized the therapy of 16 patients in six months. Guideline-concordant therapy increased after implementation of the BPA (72% vs. 91%, p=0.001) (Figure 1). Vancomycin prescribing increased and metronidazole prescribing decreased after the BPA (Figure 2). Reasons for alternative CDI therapy included medication cost, lack of insurance coverage, and non-CDI infection. There was no difference in clinical response or unplanned encounter within 14 days after treatment initiation. Fewer patients after the BPA had CDI recurrence within 14-56 days of the initial episode (27% vs. 7%, p< 0.001). Figure 1. Guideline-concordant CDI therapy Figure 2. Specific CDI therapy Clinical decision support increased prescribing of guideline-concordant CDI therapy in the outpatient setting. A targeted BPA is an effective stewardship intervention and may be especially useful in settings with limited antimicrobial stewardship resources. Susan L. Davis, PharmD, Nothing to disclose Rachel Kenney, PharmD, Medtronic, Inc. (Other Financial or Material Support, spouse is an employee and shareholder)

Patients with a reported β-lactam allergy (BLA) are often given alternative perioperative antibiotic prophylaxis, increasing risk of surgical site infections (SSIs), acute kidney injury (AKI), and Clostridioides difficile infection (CDI). The purpose of this study was to implement and evaluate a pharmacist-led BLA clarification interview service in the preoperative setting. A pharmacist performed BLA clarification telephone interviews before elective procedures from November 2018 to March 2019. On the basis of allergy history and a decision algorithm, first-line preoperative antibiotics, alternative antibiotics, or allergy testing referral was recommended. The pharmacist intervention (PI) group was compared to a standard of care (SOC) group who underwent surgery from November 2017 to March 2018. Eighty-seven patients were included, with 50 (57%) and 37 (43%) in the SOC and PI groups, respectively. The most common surgeries included orthopedic surgery in 41 patients (47%) and neurosurgery in 17 patients (20%). In the PI group, all BLA labels were updated after interview. Twenty-three patients were referred for allergy testing, 12 of the 23 (52%) completed BLA testing, and penicillin allergies were removed for 9 of the 12 patients. Overall, 28 of the 37 (76%) pharmacy antibiotic recommendations were accepted. Cefazolin use significantly increased from 28% to 65% after the intervention (P = 0.001). SSI occurred in 5 (10%) patients in the SOC group and no patients in the PI group (P = 0.051). All of these SSIs were associated with alternative antibiotics. Incidence of AKI and CDI was similar between the groups. No allergic reactions occurred in either group. Implementation of a pharmacy-driven BLA reconciliation significantly increased β-lactam preoperative use without negative safety outcomes.