2025 Southeastern Residency Conference

Evaluate safety of DPP4 inhibitors versus mealtime insulin for glycemic control in non-critically ill-hospitalized patients
Graylon Cross-Penn, Aayush Patel, McKenzie Hodges, Kelly Carter
Piedmont Columbus Regional Midtown – Columbus, GA


Background: The management of hyperglycemia in non-critically ill hospitalized patients with type 2 diabetes is often complex and requires a careful balance between glycemic control and the risk of hypoglycemia. Insulin is a common therapy in the inpatient setting, but it carries a risk of hypoglycemia, particularly in patients with lower baseline insulin requirements. Hypoglycemia leads to adverse clinical outcomes, including cardiovascular events, neurological impairment, and prolonged hospitalizations. These risks highlight the need for alternative approach which maintains glycemic control while minimizing the risk of hypoglycemia. Dipeptidyl peptidase-4 inhibitors (DPP-4i) have emerged as a potential alternative for specific populations due to their favorable safety profile, oral route of administration, and tolerability. However, evidence comparing the safety of DPP-4 inhibitors versus insulin in the inpatient setting is limited. This study aimed to address this gap by evaluating the incidence of hypoglycemia and other glycemic outcomes in non-critically ill hospitalized patients receiving either DPP-4 inhibitors or mealtime insulin. The findings may provide critical insights into optimizing glycemic management strategies for hospitalized patients, improving patient outcomes and resource utilization.
Methods: The study reviewed the medical records of non-critically ill patients admitted to Piedmont Hospitals between January 2023 and December 2023. The primary objective was to compare the incidence of hypoglycemia (defined as blood glucose <70 mg/dL) between patients receiving DPP-4 inhibitors and those treated with insulin. Secondary objectives included evaluating the incidence of severe hypoglycemia (BG <50 mg/dL), mean daily blood glucose levels, incidence of hyperglycemia (BG >180 mg/dL), length of hospital stay, total daily insulin requirements, and the number of injections per day.
Patients were included if they were aged 18 years or older, A1C <8% and on outpatient diabetes regimens involving diet alone, oral antidiabetic agents, or insulin therapy of less than 0.5 units/kg/day. Patients were excluded if they were admitted with diabetic ketoacidosis or hyperosmolar state, history of type 1 diabetes, recent glucocorticoid use, ICU admission, history of gallstones, cholecystitis, gallbladder cancer or pancreatitis, blood sugar greater than 400 or less than 70 on admission, patients that underwent surgery, with triglycerides >150 mg/dL, or if they were unable to take oral medications. Statistical analysis was conducted using independent t-tests for primary and secondary outcomes.


Results: In the insulin-only group, there were 14 recorded hypoglycemic events. In the linagliptin with or without insulin group, there were 13 events.
The statistical analysis yielded a p-value of 0.897, indicating no significant difference between the two groups in terms of hypoglycemia risk. The linagliptin and insulin group was much lower with the average daily insulin requirements at 3.94 vs 8.77. This was the only category of the original study that was statistically significant out of the secondary outcomes. The linagliptin group shows significantly higher drug costs—$142.84 compared to $14.57 in the insulin-only group—primarily driven by the high cost of linagliptin ($21/day). When analyzing the linagliptin group only (a subgroup from the linagliptin and correctional insulin) the linagliptin only group out performed the insulin only group, yielding a p-value of 0.029. When further analyzing the linagliptin only subgroup versus the insulin only, it was also statistically significant when comparing average daily blood glucose, number of injections and average daily insulin requirements.


Conclusion: These findings suggest that, in our study population, the addition of a DPP-4 inhibitor did not significantly impact hypoglycemia incidence compared to insulin alone. The total mean cost per patient is nearly 10 times higher in the linagliptin group, highlighting the importance of weighing clinical benefits against financial impact when considering DPP-4 inhibitors in inpatient settings. Standard deviations reflect variability in hospital length of stay. While this chart review provides valuable insights, there are several limitations to consider. First, the data does not represent the United States. Using only linagliptin can be limiting as different DPP4i may have varying efficacy, safety profiles, and patient responses. Additionally, there is differences in diabetes management, which differs across healthcare providers, hospitals, and standard of care groups. These differences in management strategies can impact the consistency of findings and should be considered when interpreting the results. In our analysis, the use of DPP-4 inhibitors was found to be statistically significant when evaluating daily insulin requirements. This suggests that patients on a DPP-4 inhibitor may have different insulin needs compared to those on insulin alone. However, from a cost-effectiveness standpoint, continuing insulin therapy remains a viable and practical option. While our findings provide insight into the potential impact of DPP-4 inhibitors, larger studies with more robust sample sizes are needed to further evaluate their role in non-critically ill hospitalized patients, especially with the linagliptin only group versus insulin only treatment arms.

Contact: graylon.cross-penn@piedmont.org

Title: Chronic Opioid Use Post ICU Exposure in Opioid Naive Patients


Authors: Hillary Anne Reeves, PharmD; Caitlin Thomas, PharmD, BCCCP


Objective: To assess the prevalence of opioid prescribing and post-discharge opioid use in opioid-naive ICU patients who received scheduled opioids during invasive mechanical ventilation (IMV) weaning. This study aims to identify risk factors for prolonged opioid use and guide opioid stewardship interventions.


Self-Assessment Question: Which of the following best describes a concern associated with the use of opioids for weaning patients from IMV?


1. Opioids used during weaning are often underdosed, leading to withdrawal symptoms.

2. Opioid use during weaning may predispose previously opioid-naïve patients to new prescriptions at discharge and long-term use.

3. Use of opioids during IMV can lead to inadequate sedation and poor weaning outcomes.

4. Opioids are no longer recommended for use in ICU settings due to high misuse rates.

Background: Opioid prescribing in hospitalized patients, particularly those in intensive care units (ICUs), is an area of growing concern due to its potential impact on long-term opioid use. While opioids are commonly used to facilitate weaning from IMV by providing sedation and analgesia, there is limited understanding of their role in continued opioid use after discharge, especially in opioid-naive patients. This study aims to evaluate the practice patterns of opioid prescribing and post-discharge opioid use during IMV weaning. The findings from this study may help guide opioid stewardship efforts and inform strategies for optimizing pain management while minimizing opioid exposure in critically ill patients.


Methods: This is a single-center, retrospective cohort study that has been deemed a quality improvement project and exempt from IRB approval. The study evaluated opioid prescribing patterns in opioid-naive ICU patients who received scheduled opioids during IMV weaning at a large, tertiary level, community teaching hospital. The electronic medical record (EMR) and Prescription Drug Monitoring Program (PDMP) will be used to evaluate opioid prescribing patterns at discharge and outpatient use. Patients > 18 years old admitted to the ICU, mechanically ventilated for > 3 days, and receiving scheduled opioids during IMV weaning will be included. Patients are excluded if they have a documented history of chronic opioid use or opioid use disorder prior to ICU admission, patients discharged to hospice or palliative care, major surgery at any point during admission, or incomplete medical records that do not provide sufficient data for analysis. EMR data from January 2023 -December 2023 will assess opioid initiation in the ICU, with PDMP review for up to 1-year post-discharge to evaluate outpatient opioid use. This study will compare patients who received scheduled opioids during IMV weaning to those who did not, identifying factors associated with outpatient prescribing and long-term opioid use. 


Results: A total of 531 patients were screened, resulting in 58 patients who met inclusion criteria. Of these, 47 patients comprised the comparator group (no scheduled opioids during weaning), while 11 patients made up the study group (received scheduled opioids during weaning). The primary outcome showed no statistically significant difference (p=1.00) in the number of patients discharged with an opioid prescription between the two groups. Six patients in the comparator group were discharged with an opioid prescription, compared to one patient in the study group.


Conclusions: This study explored the relationship between scheduled opioid administration during invasive mechanical ventilation (IMV) weaning and subsequent outpatient opioid prescribing in opioid-naïve ICU patients. Although the study was underpowered, the results did not show a significant increase in discharge opioid prescribing among patients who received scheduled opioids during IMV weaning compared to those who did not. These findings suggest that current prescribing practices at our institution may reflect appropriate tapering or discontinuation of opioids prior to discharge, even among patients exposed to scheduled opioids during their ICU stay.

Title: Pharmacist Driven Penicillin Allergy De-escalation using the PEN-FAST Allergy Decision Rule
Authors: Ny'Asia Cleckley, Molly Thompson, Jessica Sutton
Background: 
β-lactam allergies, specifically allergies to penicillin, are the most commonly reported antibiotic medication allergies in the hospital setting. However, according to the American Academy of Asthma, Allergy, and Immunology (AAAAI) Joint Task Force, 90% of patients with a reported penicillin allergy are able to tolerate the penicillin drug class.
The PEN-FAST penicillin allergy assessment is a clinical decision-making tool designed to be a quick and low burden method to identify patients with low-risk penicillin allergies who would likely have a negative result if formal penicillin allergy testing was conducted. The mnemonic PEN-FAST describes the risk score of true penicillin allergy based on the following queries: five years or fewer since the reaction (2 points), anaphylaxis/angioedema or severe cutaneous adverse reaction (SCAR) (2 points), and treatment required for the reaction (1 point).  Trubiano et al (2020) validated the PEN-FAST assessment with results showing it to have a high negative predictive value of 96.3% in scores less than 3 points, coinciding with the cutoff for low risk penicillin allergies.  
Methods: A single-center, retrospective, cohort analysis conducted on adult (>18 years old) patients with a documented penicillin allergy admitted to Trident Medical Center from February 2025 to April 2025. The study is pending exempt review by the health system’s Institutional Review Board. Eligible patients included non-pregnant adults with a penicillin allergy documented in electronic medical record (EMR) who were screened during an admission within the defined timeframe using the PEN-FAST assessment. Patients were excluded from the study if there was a documented allergic reaction of anaphylaxis (throat/mouth swelling, trouble breathing) or SCAR to penicillin in the EMR or were unable/unwilling to participate in the interview. The primary outcome of this study is the incidence of patients with a PEN-FAST score less than 3 points correlating with low risk of true penicillin allergy. Secondary outcomes include penicillin allergy de-labeling, post-screening oral amoxicillin challenge, antibiotic selection, antibiotic-associated adverse events and the time required for patient interview. The patients’ electronic medical record will be used to collect data on patient demographics, PEN-FAST screening result, antibiotic selection, antibiotic-associated adverse events and use of a post-documentation amoxicillin oral challenge.
Results: In progress
Conclusion: In progress

Cefazolin Plus Metronidazole Compared to Cefoxitin Alone for Surgical Prophylaxis in Hysterectomies
Tristan Jernigan, Austin Roberts, Kendra Spilkin, Benjamin Casey
Background: Antibiotic prophylaxis is a critical component of surgical site infection prevention in hysterectomy procedures. The 2013 IDSA guidelines recommend cefazolin, cefoxitin, or ampicillin-sulbactam as first-line agents, administered within one hour prior to surgical incision. When choosing between the guideline recommendations it is important to consider reported efficacy and the logistical implications of different regimens. Cefazolin is redosed every four hour while cefoxitin is redosed every two hours. In a 2017 study by Till and colleagues, cefazolin combined with metronidazole reduced the incidence of post-operative infections compared to cefoxitin alone in hysterectomy patients. Based on these findings, our institution transitioned from cefoxitin to a cefazolin plus metronidazole regimen in May 2022. This study evaluated the impact of this change on compliance with surgical prophylactic antibiotic administration, focusing on whether this transition increased appropriate prophylaxis rates by leveraging cefazolin’s longer dosing interval to reduce the incidence of missed or delayed prophylactic doses.
Methods: This was a single-center, retrospective, pre and post implementation cohort study of adult patients who underwent a hysterectomy between November 15th, 2021 and November 14th, 2022. Patients were identified by a generated list of patients who underwent a hysterectomy during the designated time frame and were randomized. Patients were included if prophylactic antibiotics were administered. Patients were excluded based on extremes of age or if clindamycin and gentamicin were administered for surgical prophylaxis. The primary endpoint was the percentage of patients who received appropriate antimicrobial surgical prophylaxis. Appropriate prophylaxis was defined as preoperative administration within one hour of surgical incision and subsequent redosing within thirty minutes of the recommended frequency. The secondary endpoints included the incidence of post-operative infection, direct drug cost with each regimen, and a subgroup analysis of the primary outcome to evaluate reasons for inappropriate prophylaxis. 
Results: A total of 260 patients were screened, and 226 (n = 140 cefoxitin, n = 86 cefazolin plus metronidazole) met criteria for our study. Patients who received clindamycin plus gentamicin for surgical prophylaxis (n=34) were excluded from the primary and secondary endpoints. Baseline characteristics were similar between the two groups. Cefoxitin was associated with significantly higher appropriate prophylaxis rates when compared to cefazolin plus metronidazole (88.6% vs. 76.8%, p=0.009). There was no significant difference in post-operative infections between the two groups (0.4% vs. 0%, p=0.32). Subgroup analysis of the primary outcome found significantly more cases of inappropriate dosing in the cefazolin plus metronidazole group (4.7% vs 0%, p=0.02) and significantly more cases of redosing errors with cefoxitin (2.9% vs. 0%, p=0.04).
Conclusion: While cefoxitin was associated with higher rates of appropriate prophylaxis, the primary reason for inappropriate prophylaxis in the cefazolin plus metronidazole group was the use of cefazolin alone for prophylaxis, which is an appropriate, guideline-recommended prophylaxis regimen. Cefoxitin was associated with a higher rate of redosing errors when compared to cefazolin plus metronidazole. Overall rates of post-operative infection were low, with no significant difference being observed between the two groups. 
Disclaimer: This research was supported (in whole or in part) by HCA Healthcare and/or an HCA Healthcare affiliated entity. The views expressed in this publication represent those of the author(s) and do not necessarily represent the official views of HCA Healthcare or any of its affiliated entities.

Authors: Akua Kuffour, Brianna Qualls, Christine Wong, Devon Tousignant 
Purpose: Erythropoiesis-stimulating agents (ESAs) are commonly utilized in the management of anemia associated with chronic kidney disease (CKD). Despite the established efficacy of these agents, the Kidney Disease: Improving Global Outcomes (KDIGO) Clinical Practice Guideline for Anemia in CKD recommends that the potential benefits of ESAs should be balanced against the potential risks of ESAs, which include hypertension and thromboembolism. Pharmacists can significantly contribute to the safe and effective use of ESAs through dose adjustments, order modifications, and continuous monitoring. This study aimed to assess the impact of a pharmacist-led ESA management protocol of hospitalized patients with CKD-related anemia. 
Methods: This study received an exemption from the Institutional Review Board. Using a quasi-experimental, retrospective comparative design, it assessed utilization of ESAs for anemia in CKD before and after the implementation of a pharmacist-driven management protocol. The pre-intervention group reflected nephrology-driven ESA management of anemia in CKD with data collected from November 01, 2023 to March 30, 2024. A locally approved protocol to provide guidance on pharmacy-led optimal initial ESA dosing, dose adjustments, monitoring parameters, and iron repletion was created with approval by nephrologists and the Pharmacy and Therapeutics Committee. All pharmacists and dialysis nurses were educated regarding the new process and protocol. The post-intervention period reflected pharmacy-driven ESA management of anemia in CKD from December 24, 2024 to February 28, 2025. Data was collected through retrospective chart review, capturing relevant patient information including ESA doses administered, transferrin saturation, ferritin levels, hemoglobin concentrations, hemodialysis status, adverse events, and pharmacist interventions. The primary outcome assessed the incidence of hemoglobin values within target range (10 g/dL – ≤ 11.5 g/dL). Secondary outcomes included the incidence of appropriate iron repletion when indicated, incidence of blood transfusion after ESA initiation and incidence of hemoglobin level > 11.5 g/dL.  
 Results: A total of 160 patients were included, 83 in the pre-intervention group and 77 in the post-intervention group. For the primary outcome, the incidence of hemoglobin values within the target range (10 g/dL – ≤ 11.5 g/dL) was 14.82% in the pre-intervention group and 21.45% in the post-intervention group (p < 0.001).  Compared with the pre-intervention group, the incidence of appropriate iron repletion when indicated was higher in the post-intervention group (17% vs. 57%; p= 0.003). The incidence of hemoglobin > 11.5 g/dL occurred in 5.51% in the pre-intervention group and 2.55% in the post intervention group (p= <0.001). Blood transfusion after ESA initiation occurred in 1.2% in the pre-intervention group and 3.9% in the post-intervention group (p= 0.276).  
Conclusion: The pharmacist-driven ESA management protocol for anemia in chronic kidney disease significantly improved the attainment of target hemoglobin levels and appropriate iron repletion. The protocol effectively prevented overcorrection of hemoglobin levels, reducing the risk of adverse cardiovascular events.  

Background/Purpose:  

Venous Thromboembolism (VTE) is a major public health issue, affecting about 900,000 people annually in the United States, and is the leading preventable cause of in-hospital mortality. Current guidelines recommend direct oral anticoagulants (DOACs) over vitamin K antagonists, like warfarin, for VTE due to their favorable pharmacokinetic profile, fewer interactions, and lack of routine monitoring. However, the safety and efficacy of DOACs in patients with extreme body weight remain uncertain. Obesity, defined as body mass index (BMI) ≥ 30 kg/m2, is a known risk factor for VTE, with mechanistic links to venous stasis, endothelial dysfunction and hypercoagulability. Despite this, obese individuals have been underrepresented in major randomized controlled trials that assess the efficacy of DOACs in treating VTE, leading to uncertainties regarding optimal anticoagulation.  

 In 2016, the International Society of Thrombosis and Haemostasias (ISTH) recommended warfarin over DOACs for patients with a BMI ≥ 40 kg/m2 or weight ≥ 120 kg due to concerns of reduced drug exposure with fixed dose DOACs. Though pharmacokinetic data suggest that obese patients may achieve therapeutic levels with DOACs, more clinical evidence is needed for definitive recommendations in this patient population. In 2021, ISTH updated their stance to recommend DOACs in all patients regardless of BMI, however clinical evidence in severely obese populations remains limited. This study aimed to evaluate the safety and efficacy of DOACs in obese patients compared to non-obese patients being treated for VTE at Emory Decatur Hospital.   


Methods:  

This was a single-center, IRB-approved, retrospective study conducted from October 2022 through November 2024. Patients were included if they were ≥18 years of age, diagnosed with a VTE and treated with a DOAC. Patients were excluded if they were on anticoagulation for any other indication. Patients were stratified into morbidly obese and non-morbidly obese patient groups. Morbid obesity was defined as BMI ≥40 kg/m2 or weight ≥ 120 kg. The primary outcome was the incidence of recurrence of VTE within 12 months. Secondary outcomes included all-cause mortality, VTE-related mortality, and length of stay. Safety outcomes include major bleeding events, clinically relevant non-major bleeding events and adverse drug reactions.  

Results:

The incidence of recurrence of VTE within 12 months occurred in 7.3% of patients in the BMI <40 kg/m2 group (n=4) and 2% of patients in the BMI ≥40 kg/m2 group (n=1) (OR 0.26, 95% CI [0.027-2.46], p-value 0.367). There was one death from all-cause mortality in the BMI ≥40 kg/m2 group and no deaths occurred in the BMI <40 kg/m2 group (p-value 0.47). No VTE related mortality was observed. The median length of stay was 50.5 hours in BMI < 40 and was 51.8 hours in the BMI ≥40 kg/m2 group (p-value 0.44). Major bleeding events occurred in 5% of patients in the BMI <40 kg/m2 group (n=3), while 6% of patients had major bleeding events in the BMI ≥40 kg/m2 group (n=1) (OR 1.13, 95% CI [0.217-5.88], p-value 1). Non-major bleeding events occurred in 12% of patients in the BMI < 40 kg/m2 (n=7) and in 6% of patients in the BMI ≥40 kg/m2 (n=3). 

Conclusion: 

The use of DOACs in patients with morbid obesity may be a safe and effective option for VTE treatment, with no observed increase in recurrent VTE or bleeding events. However, larger randomized controlled trials are needed to fully evaluate their efficacy and safety in this population.

Title: Pre/Post Analysis of Vancomycin Dosing Guidance in Patients with Obesity in the Hospital Setting


Authors: Courtney Feagin, Kelly Gamble, Hannah Schmoock


Background: Pharmacokinetic data for patients with obesity is not readily available for most medications, and pharmacokinetic profiles vary greatly between patients making dosing difficult to standardize. Supratherapeutic doses of vancomycin can lead to nephrotoxicity and ototoxicity, thus leading to the need for close monitoring of vancomycin levels. Currently, there is limited guidance on vancomycin dosing recommendations in this population. The purpose of this study is to evaluate the effects of a vancomycin dosing in obesity guidance document on dosing efficacy and patient outcomes.  


Methods: This institutional review board approved, retrospective, single-center pre-post cohort study utilized an electronic list of adult patients hospitalized at McLeod Regional Medical Center between March 2024 and February 2025. The pre-group patients were evaluated between March 2024 and September 2024, and the post-group patients were evaluated between November 2024 and February 2025. Patients were included if they had a body mass index (BMI) of ≥ 30 kg/m2 on admission, received at least one dose of intravenous vancomycin, and had at least one vancomycin level drawn at steady state. Patients were excluded if they had baseline severe renal impairment (including acute kidney injury (AKI), hemodialysis, continuous renal replacement therapy, or peritoneal dialysis), pregnant or breastfeeding, reported an allergy to vancomycin, or if vancomycin was initiated at an outside institution. An internal guidance document was created by the infectious disease pharmacy team to aid with dosing vancomycin in patients with obesity, and it was presented to inpatient pharmacists during October 2024. Data was collected via an electronic, password protected spreadsheet accessible only to the primary study investigators. The primary outcome measured was the incidence of therapeutic doses of vancomycin according to target trough levels. Secondary outcomes measured were the incidence of obese patients receiving therapeutic doses of vancomycin according to target area under the curve (AUC) values, incidence of acute kidney injury (AKI), non-therapeutic trough levels at steady state, and non-therapeutic AUC values at steady state.    


Results: Overall, 905 patients were screened for inclusion: 597 patients in the pre-guidance document group and 308 patients in the post-guidance document group. In the pre-group, 550 patients were excluded, and 45 patients were included. In the post-group, 283 patients were excluded, and 23 patients were included.  The overall population is reflective of 61.8% (n=42/68) male, median age of 56 years old (interquartile range [IQR] 46 to 63), median BMI of 35.9 kg/m2 (IQR 32.3 to 41.4), and median Charleson Comorbidity Score of 2 (IQR 1 to 4). The primary outcome was seen in 26.7% (n=12/45) of patients in the pre-group and 34.8% (n=8/23) of patients in the post-group (p=0.49). Secondary outcomes in both groups are as follows: patients with AKI at 72 hours post discontinuation of vancomycin (17.8% vs. 21.7%, p=0.69), patients with a therapeutic AUC (60% vs. 56.5%, p=0.78), subtherapeutic trough level (35.6% vs. 30.4%, p=0.67), supratherapeutic trough level (37.8% vs. 34.8%, p=0.81), subtherapeutic AUC (17.8% vs. 17.4%, p=1.0), and supratherapeutic AUC (17.8% vs. 21.7%, p=0.72).  


Conclusion: This retrospective cohort study did not find that there was a statistically significant difference between patients whose vancomycin was dosed with or without implementation of a pharmacy dosing guidance document. However, this could have been impacted by many different factors that would need to be reevaluated in future studies. These strategies may include requiring use of the dosing guidance document, re-educating pharmacists on timing of troughs and using the guidance document, extending the study timeframe, and conducting the study after the national fluid shortage resolves.

Title: Impact of an Antibiotic Guide Paired with Audit and Feedback on Antipseudomonal Antibiotic Use in a Community Hospital


Authors: Olivia Bray, Linda Johnson


Objective: The purpose of this retrospective review is to evaluate the impact of the antimicrobial stewardship intervention in reducing the inappropriate use of piperacillin/tazobactam and cefepime when possible. 


Self Assessment Question: True / False - It is appropriate to use antipseudomonal antibiotic coverage in all patients with community-aquired gram-negative infections. 


Background: Overuse of broad spectrum antibiotics such as piperacillin/tazobactam and cefepime can lead to increased resistance and difficult to treat organisms. Our institution uses more piperacillin/tazobactam and cefepime than expected, with the majority of inappropriate use during the empiric phase of treatment. Based on this data, guidelines for appropriate use of piperacillin/tazobactam and cefepime were developed and distributed to providers. Furthermore, the antimicrobial stewardship pharmacist performed focused reviews and direct provider interventions Monday thru Friday to reduce inappropriate use. 


Methods: Adult patients admitted to the hosptial initiated on antibiotics for gram negative rod infections for at least 48 hours during November 2023 (pre-period) and November 2024 (post-period) were included. Patients were excluded if not being managed by the hospitalist service or admitted to intensive care units, adult step down units, or hematology/oncology units. Patients being managed by the infectious disease services were also excluded.


Results: In progress


Conclusion: In progress

Title: Impact of Oral Antibiotics on Length of Stay in Patients with Gram-Negative Rod Bacteremia
 
Authors: Katlyn Womble, Marcus Mize, Serina Tart


Objective: To compare the length of hospital stay (LOS) in days in patients with gram-negative rod (GNR) bacteremia who are transitioned to oral antibiotics to complete treatment versus patients who receive intravenous (IV) antibiotics for the duration of therapy.


Methods: Eligible patients were those aged ≥18 years old, admitted to a hospital within the health system from January 2023 – December 2023 with documented GNR bacteremia, and received antibiotic therapy for ≥48 hours. Patients with polymicrobial infections, multi-drug resistant (MDR) or extended-spectrum beta-lactamase (ESBL) organisms, were unable to tolerate oral medications, had an infection or predisposing condition that would require IV therapy, or received >14 days of IV therapy were excluded. Data was collected through chart review on the electronic health record system. The primary outcome was the difference in LOS between patients who were transitioned to oral therapy compared to those who remained on IV therapy for the duration of treatment. Secondary outcomes included percentage of treatment failure in each group, the difference in total duration of antibiotic therapy between groups, and the difference in LOS between patients who were transitioned to oral therapy early (defined as ≤4 days from start of appropriate antibiotics) compared to those who were transitioned to oral therapy late.


Results: Of the 388 patients reviewed, 176 patients were excluded with the most common reason being polymicrobial infection. A total of 208 patients were included, with 168 being transitioned to oral antibiotics and 40 receiving IV antibiotics for the duration of therapy. Of those included, 65.4% were female and the average age was 64.9 years in the oral group, compared to 71.5 years in the IV group. Of those in the oral group, 11.9% had an indwelling device and 6% were immunocompromised, compared to 15% and 10% in the IV group, respectively. The majority of patients had a urinary infection source, with the most common isolated organism being Escherichia coli. The most common parenteral antibiotic administered was ceftriaxone, with most patients in the oral group being transitioned to cefdinir. The average length of stay was 18.7 (±17.3) days in the IV group, compared to 6.3 (±12.0) days in the oral group (95% CI -18.2 to -6.6, p < 0.0001). Average duration of therapy in the IV group was 11.4 (±3.0) days, compared to 13.0 (±3.0) days in the oral group (95% CI 0.53 to 2.65, p=0.0039). Among patients in the oral group, the average length of stay was 5.6 (±17.8) days in those transitioned to oral therapy early, compared to 6.9 (±4.0) days in those who were not (95% CI -5.6 to 3.0, p =0.5606).  Treatment failure was low in both groups, occurring in 4.2% patients in the oral group versus 10% in the full IV group (p = 0.1708).


Conclusions: Among patients with GNR bacteremia, those who were transitioned to oral antibiotics had a significantly shorter LOS compared to those remaining on IV therapy, although duration of therapy was significantly longer. Within the oral therapy group, LOS was significantly shorter in those transitioned to oral antibiotics early. Although the results of this study are descriptive, it further supports current evidence demonstrating that oral antibiotics may be effective in patients with uncomplicated GNR bacteremia and may shorten hospital LOS. Limitations of this study include the small number of patients who were immunocompromised or had indwelling devices, limiting application to these populations, and the small rate of treatment failure which prevents the outcome from being analyzed for statistical significance. More studies should be conducted to include broader patient populations and further analyze the effect that transitioning to oral antibiotics may have on treatment failure rates.