Evaluation of Actinomyces Species Infection in A Surgical Population

Presenter: Margaret Routon

This retrospective single-center study evaluated management patterns of Actinomyces infections in surgical patients. Of 184 screened patients, 84 were included. Median age was 55 years [IQR 43–64], with a median hospital stay of 11 days [IQR 5–22.5], and 18% were readmitted within 90 days.

During hospitalization, 95.2% (80 patients) received antibiotics with a median duration of 6.7 days [IQR 4-13] active against Actinomyces, but only 80.9% (68 patients) were discharged on appropriate therapy, with a median duration of 20 days [IQR 10-86]. Gaps in antibiotic therapy occurred in 35.7% of patients, with a median interruption of 35 days. Longer median duration of 14 days [IQR 7-60] in patients without a gap in therapy vs 62 days [IQR 10.75-187] for patients with gaps in therapy was observed. Aminopenicillins (54.2%) were the most common antibiotics prescribed at discharge.

These findings highlight significant inconsistencies in post-discharge management, despite adequate inpatient care. Improved standardized protocols, discharge planning, and outpatient follow-up are essential to optimize outcomes in Actinomyces infections.

Comparison of 8 vs 15 Days of Antibiotics for Hospital-Acquired Pneumonia in Adult ICU Patients

Presenter: Johnson, Amanda

This retrospective non-inferiority study compared short-course (≤8 days) vs extended-course (9–15 days) antibiotics in ICU patients with hospital-acquired pneumonia (HAP). Of 1,420 screened, 180 patients were included (93 short, 87 extended), with similar severity (APACHE II: 15 vs 16).

30-day recurrence was similar (9.2% vs 10.8%; risk difference −1.6%, 95% CI: −10.3 to 7.2), meeting non-inferiority. Hospital stay was shorter with short-course therapy (9 vs 13 days, p=0.04), while ICU stay (7 vs 8 days, p=0.24) and mortality (10.8% vs 9.2%, p=0.73) were comparable.

Pathogen isolation was higher in the extended group (33% vs 14%). Among recurrent cases, identical organisms were seen in 20% (short) vs 100% (extended) (p=0.04). These findings support short-course therapy as effective and efficient, reducing hospitalization while aligning with antimicrobial stewardship goals.

Evaluation of Pseudomonas Aeruginosa Susceptibility Patterns and Associated Treatment Strategies

Presenter: Hudson, Lauren

This single-centre retrospective study evaluated minimum inhibitory concentration (MIC) distributions and treatment patterns in patients with cefepime-susceptible Pseudomonas aeruginosa. Of 60 charts reviewed, 29 patients were included (mean age 67.2 years; 51.7% ICU admissions). Common infection sources were pneumonia (37.9%) and genitourinary infections (21.1%).

Cefepime MIC distribution showed variability: <2 (55.2%), <4 (37.9%), and 8 (6.9%). Piperacillin–tazobactam demonstrated 100% susceptibility, with 96.6% having MIC <8, while meropenem resistance was observed in 2 isolates, with most others showing MIC <1 (96.3%). Definitive therapy included cefepime (34.5%), piperacillin–tazobactam (27.6%), and meropenem (17.2%), with variability due to co-infections. Recurrent infection occurred in 3 patients, and 4 isolates developed resistance, indicating treatment failure.

These findings highlight significant variability in cefepime MICs, emphasizing the need for MIC-guided dosing strategies to optimize efficacy and limit resistance.

Treatment of Pyelonephritis with or without a Dose of IV Antibiotics in the Emergency Department

Presenter: Fix, Dellani

This retrospective cohort study evaluated whether administering an initial intravenous (IV) antibiotic dose improves outcomes in patients with acute pyelonephritis discharged from the emergency department (ED), in line with current IDSA recommendations. A total of 4,398 ED charts were screened, with 179 eligible adult patients included based on ICD-10 codes or clinical criteria.

Among these, 79 patients received IV antibiotics prior to oral therapy, while 100 did not. The primary outcome of 14-day ED revisit occurred in 15.2% of the IV group versus 10% in the non-IV group (p=0.29). Hospital admissions were reported in 2.5% versus 1%, and recurrent urinary tract infections in 8.9% versus 6%, respectively. Subgroup analyses showed no significant differences based on antibiotic class or sex.

Overall, the study found no significant benefit of IV antibiotics in reducing short-term outcomes, including ED revisits, hospitalizations, or recurrent infections. These findings question the routine use of IV antibiotics prior to oral therapy in outpatient management of acute pyelonephritis and highlight the need for more selective, evidence-based use.

Phenotypic Clustering Reveals Distinct ICU Pneumonia Subgroups with Varying Severity and Outcomes

Presenter: Chilingarashvili, Giorgi

This study explored whether unsupervised clustering could identify distinct clinical phenotypes among ICU-admitted pneumonia patients, addressing the heterogeneity in comorbidities and disease severity that may affect outcomes. Using data from the MIMIC-IV database, 264 patients were analysed. Principal component analysis (PCA) was applied for dimensionality reduction, followed by KMeans clustering based on comorbidity profiles and severity indices (SOFA, CURB-65, PSI).

Three distinct clusters were identified. Cluster 0 (n=37) had the highest severity burden (SOFA 10.1, CURB-65 3.2, PSI 139) and 100% ICU mortality. Cluster 1 (n=128), despite the lowest severity scores (SOFA 2.5, PSI 63.7), showed a mortality rate of 96.9%. Cluster 2 (n=99) demonstrated intermediate severity (SOFA 4.1, PSI 104.6) with 96% mortality. The clustering model showed modest separation (silhouette score 0.195), and the first two PCA components explained 64% of the variance. Heatmap analysis revealed differing comorbidity burdens, with Cluster 0 characterized by complex multimorbidity.

These findings demonstrate marked heterogeneity among ICU pneumonia patients, with clustering identifying phenotypes associated with distinct mortality risks beyond conventional severity scores.

Concurrent MRSA and MSSA Bacteremia in Critical Care: Diagnostic and Treatment Challenges

Presenter: Nagy, Ahmed

This case report describes a 45-year-old female with asthma, hyperlipidemia, and type 2 diabetes mellitus, admitted with respiratory failure requiring intubation. Initial findings included respiratory acidosis, leukocytosis, and hyperglycemia. On day 2, she was diagnosed with Influenza B and treated with oseltamivir. By day 4, persistent fever and leukocytosis prompted blood cultures and empiric cefepime + vancomycin. PCR testing confirmed concurrent MRSA and MSSA bacteremia, later validated by cultures.

Therapy was adjusted to cefazolin (MSSA) + vancomycin (MRSA). Serial cultures showed dynamic shifts: MSSA on day 6, MRSA on day 10; sputum cultures mirrored this (MSSA days 4–6, MRSA days 8–14). Imaging revealed cavitary pneumonia. Despite therapy, she developed septic shock (day 8) requiring vasopressors, and antibiotics were escalated to cefepime + nafcillin + vancomycin. On day 8, she suffered cardiac arrest due to tension pneumothorax, requiring chest tubes. Her condition worsened, and she died on day 14 from cardiac arrest secondary to hemothorax from necrotizing pneumonia.

This case underscores the complexity of mixed MRSA/MSSA infections, where fluctuating pathogen profiles and antibiotic pressure complicate treatment, highlighting the need for rapid diagnostics, adaptive therapy, and aggressive management.

MBL-producing Pseudomonas: Predictors of 30-Day Mortality in ICU Patients

Presenter: Pérez-Jaen, Marcia

This retrospective cohort study evaluated predictors of 30-day mortality in critically ill patients with metallo-β-lactamase (MBL)–producing Pseudomonas aeruginosa infections, a setting with limited treatment options and high mortality. Conducted at a tertiary care center in Costa Rica (January 2020–December 2022), the study included 42 adult patients  (mean age 48.3 years; 52.4% male) with microbiologically confirmed infections.

The overall 30-day mortality rate was 47.6%. In bivariate analysis, several host and procedural factors were significantly associated with increased mortality, including septic shock (HR 6.11, 95% CI 1.40–26.7, p=0.016), mechanical ventilation (HR 4.90, 95% CI 1.39–26.7, p=0.016), prior hospitalization (HR 3.53, 95% CI 1.15–10.79, p=0.027), and need for enteral feeding (HR 3.53, p=0.027). Diabetes showed a trend toward significance (HR 2.53, p=0.051).

Importantly, antibiotic regimen—including Ceftazidime/Avibactam plus Aztreonam—was not independently associated with improved survival. Kaplan-Meier analysis further demonstrated reduced survival among patients with septic shock and those requiring invasive support.

These findings suggest that clinical outcomes in MBL-Pseudomonas infections are more strongly influenced by host condition and need for supportive care than by antibiotic choice alone.

Piperacillin-Tazobactam Order Panel Updates on Operational Efficiency and Outcomes: Pre-Post Analysis

Presenter: Schultheis, Jennifer

This retrospective pre–post study evaluated the impact of an electronic order panel redesign for piperacillin–tazobactam (TZP) in ICU patients. Updates included embedded loading dose, extended infusion dosing, renal-adjusted frequency, and optimized sequencing with vancomycin. A total of 655 patients were analyzed (353 pre, 302 post).

Post-implementation, appropriate TZP orders improved significantly (88.4% vs 80.5%; p=0.008), driven by fewer order corrections. There was a marked increase in loading dose use (difference in proportions 0.76, 95% CI 0.7-0.81), timely subsequent dosing (difference in proportions of 0.39; 95% CI, 0.12-0.66), and correct sequencing with vancomycin (difference in proportions 0.09, 95% CI 0.01-0.16) .

In patients with gram-negative infections, mortality was unchanged, but modest improvements were seen in length of stay, vasopressor duration, and ventilation duration. These findings demonstrate that EHR-based order optimization improves operational efficiency and antimicrobial use, supporting stewardship and workflow improvements in critical care settings.

Real-World Outcomes of Alternative Cefiderocol Dosing in Adults with Carbapenem-Resistant Organisms

Presenter: Huang, Janice

This retrospective cohort study evaluated a pharmacodynamically optimized alternative dosing strategy for cefiderocol in adults with carbapenem-resistant gram-negative infections (2019–2024). The regimen (1 g every 6–12 hours based on renal function) was designed using Monte Carlo simulations to improve vial utilization and reduce costs. Patients were divided into Group A (CREDIBLE-CR ineligible) and Group B (eligible).

The primary composite outcome (28-day mortality or treatment failure) occurred in 33.3% (A) vs 42.9% (B) (p=0.445), with no significant difference. Treatment failure rates were also similar (23.3% vs 21.4%, p=0862). Most infections were pneumonia (38%) and bloodstream infections (34%), predominantly caused by Pseudomonas aeruginosa (52%) and Acinetobacter baumannii (14%).

These findings suggest that alternative cefiderocol dosing may maintain clinical effectiveness while improving cost efficiency, supporting its potential role in real-world antimicrobial stewardship.

Evaluation of Cefepime Resistance Patterns After Cefepime Exposure on A Burn Unit

Presenter: Buckley, Christopher

This retrospective study evaluated the impact of switching from intermittent infusion to IV push cefepime on antibiotic resistance in a burn unit. A total of 61 patients (median age 56, total body surface area (TBSA burn) 9.75%, total weight of 83 kg and admission creatinine of 1.1 mg/dL) receiving ≥3 doses of cefepime were included. The median dose of cefepime was 2000 mg, with 83% of patients receiving an every-8-hour regimen for a median duration of 5 days.

Cefepime resistance to gram negative (GN) bacteria after exposure developed in 29.5% of patients, with an additional 6.6% progressing from intermediate to resistant strains. The most common resistant organism was Pseudomonas aeruginosa (18% of patients). Median time to resistance was 14.5 days. Notably, all patients with cefepime resistance also showed resistance to other antipseudomonal agents.

These findings suggest a high burden of resistance, particularly in burn patients, and raise concerns that IV push administration may contribute to suboptimal pharmacokinetics and resistance development, emphasizing the need for optimized dosing strategies.

Effectiveness of Beta-Lactams Adjusted Based on Cystatin C for Treatment of Gram-Negative Infections

Presenter: Marcus, Sarah

This single-centre, cohort, retrospective cohort study evaluated the use of cystatin C–based eGFR for dosing beta-lactam antibiotics (cefepime, piperacillin–tazobactam, meropenem) in critically ill ICU patients with gram-negative infections. Of 337 screened, 19 patients met inclusion criteria. Patients were included if they used cystatin C eGFR for cefepime, piperacillin-tazobactam, or meropenem dose adjustments, had one positive culture for a gram-negative organism(s) or were treated for pneumonia based off clinical signs and symptoms.

Treatment failure occurred in 21.1% (4/19)—including 3 deaths before treatment completion and 1 antibiotic restart. ICU mortality was significantly higher in the failure group (100% vs 33%, p=0.033). Safety outcomes were favourable: 1 case of AKI (6.7%), no confirmed C. difficile infections, and no beta-lactam encephalopathy.

These findings suggest that cystatin C–guided dosing may provide a more reliable and safer alternative to serum creatinine-based adjustments in critically ill patients, where traditional renal markers may be inaccurate.

Critical Care Congress 2026, March 22-24, Chicago.