Optimizing HAP Therapy: Launch of a New Antibiotic Protocol

Authors:
Marianna Frazee, Jennifer Beavers, Robel Beyene, Jill Streams, Elizabeth Krebs

Body of Abstract:
Background: Hospital-acquired pneumonia (HAP) is common in trauma patients, making appropriate empiric antibiotic selection and de-escalation important for outcomes and resource use. Evidence suggests that per oral (PO) linezolid offers similar MRSA coverage to intravenous vancomycin with potential reductions in cost and medication-related complications. In August 2025, our Level 1 trauma center revised our practice management guideline (PMG) for antibiotic choice for patients with presumed HAP, recommending PO linezolid in place of IV vancomycin for empiric MRSA coverage. The new protocol also provided designated antibiotic de-escalation strategies based on clinical improvement metrics and respiratory culture results.

Aim: This study aimed to evaluate both protocol implementation and the impact of this new guideline on antibiotic stewardship, hypothesizing that the use of vancomycin and cefepime (as a marker of antipseudomonal antibiotic use) would decrease following protocol change.

Methods: A multidisciplinary team including surgeons, advanced practice providers, and pharmacists developed the new PMG. Education was provided at team-wide meetings and daily morning reports, with the guideline publicly available on our trauma website per our center’s standard practice. This study retrospectively compared patients treated for presumed HAP in the 3 months before and after protocol implementation. Demographics, antibiotic data, and outcomes were obtained via chart review and institutional trauma and infection registries. Data were compared using Wilcoxon rank sum and Chi-square analysis.  Changes in vancomycin and cefepime utilization were also evaluated using a statistical process control (SPC) c-chart to detect variation over time and identify shifts following the intervention.

Results: A total of 82 trauma patients were treated for 102 episodes of presumed pneumonia, 55 before and 47 after guideline implementation. Protocol adherence exceeded 75% within three months of rollout. The SPC c-chart for vancomycin use demonstrated special cause variation after implementation, indicating a statistically meaningful shift in the process (Figure). Median days of vancomycin per pneumonia episode decreased in the post-implementation period (3 vs. 0, p<.01), while median days of cefepime prescribed were similar (4 vs. 5, p=0.18). There was no difference in rate of antibiotic de-escalation (54% vs. 63%, p=0.37.) Importantly, there were no adverse drug reactions noted during this early implementation period, and rates of antibiotic re-escalation remained similar (37% vs. 27%, p=0.43).  Conclusions: Transitioning empiric HAP therapy from IV vancomycin/cefepime to PO linezolid/cefepime resulted in sustained reduction in vancomycin use, while the new de-escalation protocol did not significantly impact cefepime prescribing or de-escalation rates. This change in antibiotic regimen led to cost savings of approximately $6,200 in drug and lab costs over a 3-month period.

Outcomes of Pediatric Patients with Perforated Appendicitis After Implementation of a Delayed Post-operative Imaging Protocol

Authors:
Jill Knepprath, Eiichi Miyaska

Body of Abstract:
Background:

Post-operative management of pediatric perforated appendicitis is highly variable with no clear guidelines for the timing of post-operative imaging. Some studies suggest that delayed imaging may decrease both unnecessary radiation and invasive procedures, while other authors found that delayed imaging was associated with prolonged hospital stays. In this study we investigated changes in outcomes at a single institution after the implementation of a protocol recommending imaging on post-operative day (POD) 7.  

Methods:

This was a single institution retrospective chart review of pediatric patients who underwent an appendectomy for perforated acute appendicitis between January 2021 and June 2025. A modification to the protocol occurred in January 2024, with post-operative imaging recommendations changing from POD 5 to POD 7. No other changes to the protocol were made between January 2021 and June 2025. Patients were divided into two cohorts, Group 1 included patients before protocol modifications (January 2021-December 2023) and Group 2 included patients after protocol modifications (January 2024-June 2025). Categorical data between both groups were compared using the Chi-Square test and medians were compared using the Mann-Whitney U test.

Results:

There were 105 patients in Group 1 and 49 patients in Group 2. There was no difference in length of stay between groups. There was no difference in imaging use (CT or ultrasound) between perforated patients in either group. Group 2 perforated patients had imaging done on a later POD than Group 1. Neither group was more likely to alter their treatment plan based on imaging. There was no difference in post-operative interventions. There was no significant difference in overall complications or in post-operative abscess formation. However, Group 2 perforated patients were more likely to be readmitted (10.2% vs 5.7%, p= 0.0054) or have an emergency room visit after discharge (18.4% vs 5.7%, p= 0.0136). (Table) Of patients with emergency room visits in both groups, the proportion of patients with imaging during their index admission was no different.

Conclusions:

This study suggests that later imaging does not change length of stay, does not decrease imaging or interventions, and does not impact abscess rates. However, patients that undergo delayed imaging may be more likely to be readmitted or present to the emergency room after discharge.

Parabacteroides merdae: a rare Gram-negative anaerobic rod

Authors:
Aprill Park, Rebecca Kowalski, Sridha Gona, Aaron George, Stephen Kavic, Hugo Bonatti

Body of Abstract:
Introduction: Parabacteroides spp were previously listed under Bacteroides spp but refined microbiology testing led to the creation of the new genus with Parabacteroides distasonis being common than Parabacteroides merdae. They are Gram-negative obligate anaerobic rods and part of the normal microbiome. Isolation of the organisms within the microbiome has been associated with inflammatory bowel disease, metabolic syndrome, and most notable obesity. Parabacteroides merdae has been recovered from drained abscesses, wound infections, and blood. Resistance rates to clindamycin, cefoxitin, and moxifloxacin are high, which may lead to treatment failure.

Patients and Methods: Following an index case of a patient with perirectal abscess, in which Parabacteroides merdae was isolated, the institutional database of a rural hospital in the Appalachian region of Western Maryland was searched for all infections caused by the pathogen during a 4-year period. Also, all cases of Parabacteroides distasonis the more common relative were reviewed.

Results: Two patients (3 isolates) with surgical infections associated with Parabacteroides merdae were identified including a 50-year-old woman and a 74-year-old man. The morbidly obese female with poorly controlled DM underwent incision and drainage of an ischiorectal abscess which grew the pathogen together with Eggerthella lenta and Fusobacterium necrophorum and the male patient had an infected tunneled dialysis catheter, which was removed growing Parabacteroides merdae, Enterobacter cloacae and Peptostreptococcus anaerobius. Both infections were successfully managed with surgical intervention and antibiotics. Parabacteroides distasonis was isolated in 42 patients (54 isolates) and intraabdominal infections accounted for 74% of cases including diverticulitis in 11 and appendicitis in 9 patients. No sensitivity testing is done at our hospital for anaerobic bacteria such as Parabacteroides merdae or distasonis.

Conclusion: Parabacteroides merdae is an extremely rare pathogen and much less common than Parabacteroides distasonis. These infections are usually polymicrobial and whereas the exact role of these anaerobes is unknown, they should be considered true pathogens and not contaminants or innocent bystanders.

Patterns and Predictors of Delayed Antibiotic Administration in Open Fracture Care at CHUK

Authors:
Emile Musoni, Isabella Hung

Body of Abstract:
Background: Open fractures are among the most common cases received at emergency departments (ED) in low- and middle-income countries and early administration (within 60 minutes) of antibiotics is the cornerstone component in the management of open fractures. We aim to evaluate the risk factors associated with delays in time to antibiotics administration for patients with long bone open fractures at the University Teaching hospital of Kigali (CHUK) in Rwanda. 

Methods: This is a prospective study of all ages presenting to the emergency department of CHUK with an open long bone fracture. Those admitted from other health facilities or presented with infected wounds were excluded. Participants completed a structured questionnaire. Data were collected between December 2020 to April 2022 and analyzed using Stata version 13. Binary logistic regression was used to determine risk factors in >60 minutes antibiotic administration. 

Results: A total of 120 participants were recruited, the majority were males (N= 100, 83.3[MSD2] %). 99.17% (N= 99) of participants presented to the emergency department >1 hour after their accident. Additionally, 42.5% (N= 51) of participants received antibiotics within the first 60 minutes. Nightshift admissions had 6 times the odds of receiving antibiotics after 60 minutes as compared to dayshift (OR=6.0, 95% CI: 2.58-13.93, p<0.001). Patients without medical insurance had 8.5 times the odds of receiving antibiotics after 60 minutes as compared to those with health insurance (OR=8.5, 95% CI: 1.04-58.51, p=0.045) and children who presented at the emergency department with open fractures had 3.8 times the odds of receiving antibiotics after 60 minutes compared to adults (OR=3.8, 95% CI: 1.02-14.07, p=0.031).  Conclusion: Delays in antibiotic administration were widespread amongst patients presenting to CHUK with open long bone fractures. Patients particularly at risk included pediatric patients, patients without health insurance, and those presenting overnight. Furthermore, delays in presentation resulted in nearly all patients receiving antibiotics greater than 60 minutes from time of injury. Providing antibiotics in the prehospital setting may help reduce these delays. Additionally, targeted educational programs for providers may help prevent delays in antibiotic administration after arrival.

Pediatric Surgical Patients with Sepsis at a National Referral Hospital in Uganda: Predominant Organisms and Antibiotic Resistance Patterns

Authors:
Julia Harrison, Brian Kasagga, Grace Nambatya, Janet Maseka, Michael Semanda, Nasser Kakembo, Jennifer Rickard

Body of Abstract:
Background

Sepsis remains a leading cause of morbidity and mortality among children in low- and middle-income countries (LMICs), yet its burden in pediatric surgical patients is poorly characterized. Additionally, there is limited data describing causative organisms or antibiotic resistance patterns in pediatric surgical patients in LMICs. 

Methods

This analysis is part of a larger prospective observational study evaluating the etiology, diagnosis, management, and outcomes of sepsis among children on a pediatric surgery ward in Uganda. Patients with clinical criteria for sepsis are eligible for blood culture collection. A single sample is routinely obtained and processed using the BD BACTEC system in Peds Plus Aerobic bottles. In addition to culture data, demographic information, clinical diagnoses, antibiotic use, operative details, and patient outcomes are recorded.

Results

Of 82 patients enrolled to date, 42 blood culture results were available. Pathologies leading to sepsis were predominantly intra-abdominal (67%), followed by urinary tract (10%), pulmonary (7%), and skin/soft tissue infection (6%), with 10% of patients having an unknown source. The median age at culture collection was 11 days (IQR 6-21). 27 of 42 cultures (64%) were positive and all isolated bacteria were multidrug resistant. Klebsiella pneumoniae accounted for 12/27 (44%) positive cultures, followed by Staphylococcus species (5/27, 18.5%), and Candida species (5/27, 18.5%). All patients were on antibiotics, while none were on antifungals, at the time of culture collection. Based on culture results, 1 of 27 (3.7%) patients was adequately covered by antimicrobials, 20/27 (74.1%) were not adequately covered, and adequate coverage was unable to be determined in 6/27 (22.2%) patients. Klebsiella pneumoniae isolates showed high rates of resistance across 10 of 11 tested antibiotic classes (figure 1), while most (10/12 isolates, 83%) retained sensitivity to carbapenems. In-hospital mortality among all enrolled was 58%. Mortality among those with positive cultures was 60%, with 73% mortality among those with Klebsiella bacteremia. 

Conclusions

These findings demonstrate a substantial burden of culture-positive sepsis among pediatric surgical patients in Uganda. There was a predominance of intra-abdominal, followed by urinary tract pathology and most patients were extremely young at presentation. All isolated bacteria demonstrated multidrug resistance and almost all isolated bacteria were resistant to empirically used antibiotics. Empiric antifungal coverage, while not currently in practice, appears to be needed. Klebsiella pneumoniae was the leading pathogen isolated and exhibited extensive multidrug resistance. The high overall mortality highlights the severity of sepsis in this population and underscores the urgent need for strengthened infection prevention, earlier recognition of sepsis, local antibiograms, and improved access to effective antibiotics and antifungals.