2026 Southeastern Residency Conference

Authors: Leah Franks, Jeremy Ray, and Mickala Thompson
 
Background: Alcohol withdrawal syndrome (AWS) is a serious condition that progresses in severity. The most severe symptoms include withdrawal seizures and delirium tremens. Phenobarbital is a barbiturate currently classified as an alternative agent for severe AWS in the 2020 American Society of Addiction Medicine’s “Clinical Practice Guideline on Alcohol Withdrawal Management.” A 2023 meta-analysis conducted by Umar, et al. concluded that phenobarbital can be used safely and effectively for AWS within the ICU setting. Phenobarbital prescribing practices have not been analyzed thus far at our institution. Therefore, the purpose of this research is to assess the current utilization and dosing schemes of phenobarbital for AWS at our large community hospital.
 
Methods: A single-center, institutional review committee-approved pre-post analysis was conducted at Huntsville Hospital to assess the safety, efficacy, and outcomes of patients receiving phenobarbital after the implementation of a streamlined-dosing order set. Information from an initial chart review was utilized to develop an order set based on current guideline recommendations/literature while also aligning with current prescribing practices at our institution. After implementation of streamlined dosing, a post-hoc analysis will be conducted. Adult patients, greater than 18 years of age, who received phenobarbital for management of AWS were included in the study. Exclusion criteria included vulnerable patients and those who received phenobarbital for underlying seizure disorders. The primary endpoints are the utilization of a pre-determined, guideline-directed dosing scheme and benzodiazepine usage post-phenobarbital loading dose administration. Secondary endpoints include ICU admission, need for intubation, withdrawal seizure or delirium tremens occurrence, and adverse events.
 
Results: A total of 54 patients in the pre-implementation group and 29 patients in the post-implementation group were included in this study. Baseline characteristics were similar between pre-post data regarding age (mean ~ 50 years), sex (majority male), baseline liver dysfunction (~50% for both groups), and receipt of CNS depressants (26% vs 21%) prior to phenobarbital. In terms of the primary endpoints, 38% of the providers utilized the phenobarbital order set. Benzodiazepine usage (in lorazepam equivalents - mg, median [IQR]) post-phenobarbital was comparable (4 [7.8] vs 4 [7.25]). Lastly, variability amongst secondary outcomes occurred with the need for ICU admission and the need for intubation post-phenobarbital between the pre-implementation group (56% and 13%) and the post-implementation group (31% and 21%).
 
Conclusion: Overall, order set utilization was minimal, which hindered appropriate comparison between pre-post implementation data. Limitations included the variety of ordering providers, a single center study, as well as the order set not initially being in all alcohol withdrawal order sets. Future directions include implementing further education across all provider types and improving visibility of the order set to providers by adding it to the overall alcohol withdrawal admission and add-on order sets. 

EVALUATION OF A PHARMACY RESIDENT-LED VIRTUAL POPULATION HEALTH CLINIC 
Katrina Bitcon, Katie Sfirlea, Catie Harper, Madison Yates 

Background/Purpose: A value-based care delivery model aims to enhance patient care by reimbursing providers based on patient outcomes and the quality of care instead of the quantity of services delivered. In 2024, our health system established a new objective to expand value-based care throughout all areas of our system. Pharmacists in the health system's primary care clinics deliver chronic disease management services using a value-based care model; however, access to pharmacists working within these clinics has historically been limited to traditional business hours. To expand access to pharmacy services for patients experiencing barriers to this model, staffing for ambulatory care interested PGY1 pharmacy residents was structured to include virtual care outside of traditional business hours. During these visits, chronic disease states were assessed and documented, as well as any recommendations that the resident made for therapeutic management. The purpose of this study was to characterize this pharmacy resident-led virtual population health clinic and assess its impact on patient care. 


Methodology: Fifty patients were included in this single health system, retrospective, cohort study. Patients were eligible for inclusion if they completed at least one pharmacy resident-led virtual visit between October 1, 2024, and May 31, 2025.  The primary objective of this study was to determine the percentage of pharmacy resident recommendations approved by the primary care provider (PCP), which was defined as corresponding orders in the electronic medical record within 10 days of the visit. Secondary objectives included the percentage of patients achieving an A1C <8%, blood pressure (BP) <140/90 mm Hg, and LDL-C <70 mg/dL pre- and post-intervention, as well as the types of recommendations made by the pharmacy resident. 


Results: A total of 204 resident visits among 50 patients were conducted with the pharmacy residents. Of the 77 recommendations provided by the pharmacy residents, 97.3% were accepted by the referring PCP, with the most common recommendations being an increase in dose or addition of medication. There was an increase in patients achieving the goal A1C by 22% (P < 0.01), blood pressure by 24.2% (P < 0.05), and LDL-C by 20.6% (P > 0.05). The most common disease state managed by the pharmacy residents was diabetes. After the intervention, 32% were helped with medication access, 16% gained access to CGM, and 4% of patients gained access to home blood pressure monitors. 


Conclusions: The results of this study show that the majority of pharmacy resident recommendations were approved and implemented by the PCP. Additionally, there was a significant reduction in A1C and blood pressure after pharmacy resident intervention. Alongside improvements in disease state management, patients were able to gain access to home monitoring devices and assisted with medication access. Based on the findings of this study, providing after-hours virtual visits with a pharmacy resident is an effective method for reaching patients who face barriers to traditional healthcare hours.

Authors: Isaac Sauvageau, Kerri Smith, Brittney Bright, Ryan Imel, Blake Sloan 

Background: Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are a therapeutic class that mimics the naturally occurring incretin hormone GLP-1. Through this mechanism, they enhance glucose-dependent insulin secretion, suppress glucagon release, delay gastric emptying, and promote satiety. These effects make GLP-1 RAs effective agents for the management of type 2 diabetes mellitus (T2DM) and chronic weight management. They have also demonstrated cardiovascular and renal protective benefits, leading to increased use in patients with cardio-renal comorbidities. Despite their broad acceptance in clinical practice, early clinical trials and observational studies suggest a potential increase in the risk of pancreatitis (0.4% versus 0.03% in the general population). Multiple studies identify an increased risk of pancreatitis across the class, though results remain inconsistent, and the underlying mechanisms are not fully established. 

Methods: This was a multi-center, matched case-control design that included adult patients with active GLP-1 RAs on their home medication lists from May 1, 2024 – October 1, 2025. The primary objective was to determine the incidence of pancreatitis in patients using GLP-1 RAs at North Carolina Baptist Hospital, Highpoint Medical Center, and Davie Medical Center. The secondary objective was to identify patient-specific factors associated with pancreatitis risk among GLP-1 RA users. Patients were excluded from the study if they initiated GLP-1 RA therapy within 30 days of data collection. The incidence of pancreatitis was calculated from the unadjusted population prior to case matching. Cases were defined as GLP-1 RA users admitted to the hospital with a primary diagnosis of pancreatitis. Controls consist of GLP-1 RA users without hospitalization for pancreatitis and were matched to cases in a 1:1 ratio by age category, sex, and residential zip code to minimize confounding variables.  

Results: A total of 6,109 GLP-1 RA users were identified, of which 34 were admitted to the hospital with a primary problem of pancreatitis. This corelates with an 0.56% incidence of pancreatitis. After matching, 50 patients were included; 25 pancreatitis cases and 25 matched controls. The median age was 54 years, and 60% of the cohort was male. The median BMI was 35 kg/m2. The majority of patients were obese (82%) and had type 2 diabetes (78%). Semaglutide (56%) and tirzepatide (32%) were the most common GLP‑1 RAs used. Significant differences between cases and controls were observed for GLP‑1 RA agent distribution (p = 0.004), presence of chronic kidney disease (CKD) (36% vs. 8%; p = 0.017), cholelithiasis/cholecystitis (32% vs. 0%; p = 0.004), and serum creatinine levels (0.99 vs. 0.81 mg/dL; p = 0.038). 

Univariable logistic regression indicated increased odds of pancreatitis associated with CKD (OR 6.47; 95% CI 1.43–46.3; p = 0.027) and higher serum creatinine (OR 5.32; 95% CI 1.32–42.7; p = 0.059). Tirzepatide use was associated with significantly lower odds of pancreatitis (OR 0.11; 95% CI 0.02–0.44; p = 0.003) compared with semaglutide. GLP‑1 RA dose category, HbA1c, triglycerides, alcohol use, smoking status, and baseline metabolic parameters were not significantly associated with pancreatitis. In multivariable regression adjusting for CKD and agent, CKD remained an independent predictor (OR 8.95; p = 0.028), while tirzepatide maintained a protective association (OR 0.10; p = 0.005). 

Conclusions: The incidence of pancreatitis in GLP-1 RA users was 0.56%, which correlates with a higher risk of pancreatitis when compared to 0.03% in the general population. CKD and elevated serum creatinine were significantly associated with increased odds of pancreatitis in GLP-1 RA users, whereas tirzepatide was associated with reduced risk compared with semaglutide.

Background: Alcohol withdrawal syndrome (AWS) results from the sudden cessation of heavy alcohol use, causing symptoms such as tremors, anxiety, and agitation. Benzodiazepines remain first-line for alcohol withdrawal management according to American Society of Addition Medicine (ASAM) guidelines. Phenobarbital has a half-life of about 80 to 120 hours which allows for adequate control of withdrawal symptoms, and its self-tapering effect reduces the risk of medication abuse seen with benzodiazepines. The initial dosing of phenobarbital varies across institutions, typically starting at 130 mg or 260 mg, based on recommendations derived from clinical practice and studies such as Hendey et al. However, weight-based dosing strategies (ex. 10 mg/kg) have emerged with aims of achieving therapeutic serum concentrations and optimizing symptom control, as opposed to repeated smaller doses. The purpose of this study was to compare outcomes between patients that received an initial phenobarbital dose higher than 260 mg versus those that got a lower dose of 260 mg or less.

Methods: This IRB approved, retrospective chart review was conducted at a 711-bed academic medical center from 1/1/2020 to 9/26/2025. Adult patients who received phenobarbital were identified through a report from the electronic medical record. Patients were included if they had alcohol withdrawal documented within the chart. Patients were excluded if they were pregnant, incarcerated, had phenobarbital listed as a home medication, had a urine drug screen positive for barbiturates, any history of seizures not related to alcohol withdrawal, or were mechanically ventilated prior to phenobarbital administration. The primary outcome was stable or improved CIWA and/or RASS scores from baseline within the first 24 hours, defined as two consecutive CIWA and/or RASS scores equal to or below baseline. The secondary outcomes were the change in CIWA and/or RASS score after the first dose, improved or stable CIWA and/or RASS scores from 24-48 hours and 48-72 hours, cumulative benzodiazepine requirements, adjunctive medication requirements, hospital and ICU length of stay, and escalation of care. Comparative analyses of the primary and secondary outcomes were conducted using t-tests. Other appropriate statistical tests were utilized as needed.

Results: Of the 395 patients screened, 187 met inclusion criteria and were enrolled in the final analysis. Patients were grouped by initial phenobarbital dose; 118 received low-dose phenobarbital (≤260 mg) and 69 received high-dose phenobarbital (>260 mg). The primary outcome was achieved in 81 (80%) patients in the low-dose group compared with 59 (90%) in the high-dose group (p=0.111). No significant differences were found between the low-dose group and high-dose group for stable or improved CIWA and/or RASS scores at 24-48 hours (87% vs 90%; p=0.53) or 48-72 hours (94% vs 97%; p=0.71). Cumulative benzodiazepine requirements, adjunctive medication requirements, hospital length of stay, ICU length of stay, and escalation of care were similar between groups. Patients in the high-dose group demonstrated statistically significant greater reductions in CIWA scores following the initial dose compared to the low-dose group at 6 hours (mean change: 6.24 vs 0; p=0.005) and 24 hours (mean change: 8.05 vs 2.25; p=0.002).

Conclusions: Based on the results of this study, both low-dose and high-dose phenobarbital appear to have similar outcomes in the management of alcohol withdrawal. This study also found that higher initial phenobarbital doses were associated with more rapid symptom control in the first 24 hours without resulting in any additional complications or safety concerns. Establishing evidence-based protocols could improve standardization of care and optimize outcomes for patients with alcohol withdrawal.

Title: Impact of Embedded Clinical Pharmacy Services on Glycemic Control in an Internal Medicine Clinic 

Authors: Kisara Thompson, Tacorya Adewodu, Andrew Bundeff, Molly Hinely, Danielle Baker, Michael DeWitt 

Objective: Evaluate the effect of clinical pharmacy services on A1c reduction in patients with type 2 diabetes and a baseline A1c of ≥ 8%, compared with patients who were managed by usual care, defined as a primary care provider acting as the sole responsible diabetes managing provider. 

Background: Uncontrolled type 2 diabetes results in serious health complications. Hemoglobin A1c is used to measure average blood glucose over the past three months and is used to guide optimal treatment plans. Studies have demonstrated an improvement in microvascular and macrovascular complications in patients with improved glycemic control in those with type 2 diabetes. 

The Winston-Salem East Clinic, a part of Atrium Health Wake Forest Baptist, is the primary study site. It is a medical resident-run internal medicine clinic that serves a large population of underserved and uninsured patients. Since March of 2023, a clinical pharmacist practitioner (CPP), has been integrated into the clinic two days of the week working under collaborate practice agreements with supervising physicians. Patients with uncontrolled type 2 diabetes are referred to the CPP to provide medication management.  

Methods: This single-center, retrospective, cohort study compared mean percent change of A1c managed by a primary care provider (PCP) versus A1c managed by a CPP. Patient data was collected between July 1st, 2023 and October 31st, 2025. Patients included were ≥ 18 years of age, diagnosed with type 2 diabetes, and those who had completed at least two follow-up visits with their PCP or clinical pharmacist. Patients excluded were the following: type 1 diabetes, managed by endocrinology, seen by non-network PCPs, pregnant, lost to follow-up – defined as no additional follow-up appointments or A1c labs drawn within the study time frame. Demographic information such as sex assigned at birth, age at encounter, race and ethnicity and primary payor were collected. The primary outcome assessed was mean percent change in A1c at 90- and 180- days with and without established clinical pharmacy management. Secondary outcomes included clinical interventions made, number of continuous glucose monitors (CGMs) prescribed, and incidence of diabetes related emergency department visits and/or hospitalizations during the study time frame. Patients were identified through a data sorting platform within the electronic health record.  

Results: A total of 144 patients were enrolled in the study, with 72 patients per study group. The primary endpoints were statistically analyzed by linear regression. Adjusting for baseline A1c, there is an average reduction of 1.2 percentage points in A1c measured at 90-days in patients managed by a pharmacist, compared to patients managed by usual care (95% CI: -2.0 – (-0.36), p-value: 0.005). Adjusting for baseline A1c, there is an average reduction of 1.3 percentage points in A1c measured at 180-days in patients managed by a pharmacist, compared to patients managed by usual care (95% CI: -2.1 – (-0.46), p-value: 0.003). Pharmacist managed patients were approximately eight times more likely to have their insulin therapy changed compared to patients who were managed by usual care (95% CI: 1.3 – 2.9, p-value: 0.001). There were no differences associated with rates of change in other medication therapy changes. During the study period, pharmacy managed patients had approximately a 53-point higher prescription rate for a CGM than those managed by usual care. There were no differences detected in the incidence rates between pharmacy managed and usual care patients in regard to ED and/or hospitalizations related to type 2 diabetes within the study period.

Conclusions: Among patients with type 2 diabetes and a baseline A1c ≥ 8%, clinical pharmacy management was associated with a statistically significant greater reduction in A1c compared with usual care management with mean reduction of 1.2 and 1.3 percentage points at 90-days and 180-days, respectively. 

Elena Galagan, Joy Peterson, Karen Barlow
Wellstar Kennestone Regional Medical Center, Marietta, GA

Background

Vancomycin has a narrow therapeutic index, requiring individualized pharmacokinetic dosing to maintain trough levels between 10–20 mcg/mL while minimizing the risk of acute kidney injury (AKI). Despite its common use, many institutions lack formal metrics to assess vancomycin dosing appropriateness. This quality assurance project at Wellstar Kennestone Regional Medical Center (WKRMC) aimed to create a sustainable, ongoing quality metric and real-time monitoring system for vancomycin dosing practices.

Objectives/Methods

The project was conducted in two parts. The primary objective was to develop an ongoing quality metric that evaluates vancomycin dosing quality by monitoring therapeutic trough levels and AKI in adult patients. The secondary objective was to use this metric and Microsoft® Power BI dashboard to enhance decision-making and increase the proportion of patients achieving therapeutic targets.
Part 1 (establishing a baseline) was a retrospective quality assurance project including data from three fiscal years (July 1, 2021, through June 30, 2024). Adult patients who received intravenous vancomycin during an inpatient admission at WKRMC, had vancomycin dosed by a clinical pharmacist, and had at least one vancomycin trough level were included; pediatric patients under 18 years were excluded. The baseline percentage of patients with therapeutic vancomycin levels (10–20 mcg/mL) was calculated to establish the goal for the metric.
Baseline AKI incidence was determined by collecting up to three serum creatinine values around the time of trough collection and measuring the frequency of AKI as defined by Kidney Disease: Improving Global Outcome (KDIGO) criteria. A random sample of 123 patients was selected by choosing every 100th patient from each fiscal year for detailed review to validate the dashboard, estimate data capture performance, identify opportunities for improvement, and refine variables.
Part 2 (implementation) focused on developing a sustainable, ongoing quality metric and monitoring dashboard. Core metric components included the percentage of patients achieving therapeutic trough levels and the incidence of AKI, with stratification by key risk factors. Data elements were mapped to automated reports from Epic®, aggregated in a secure environment, and visualized in Microsoft® Power BI to create an interactive dashboard for clinical pharmacists. The dashboard was designed to support ongoing monitoring, early detection of trends, and continuous optimization of vancomycin dosing practices.
 
Results

The 123-patient review validated data accuracy, assessed variable performance, and guided refinement of the quality metric. This process led to adjustments in data capture, including the addition, modification, and removal of specific variables to enhance clinical relevance and reliability. Variables related to trough timing, serum creatinine trends, and AKI risk factors were refined to improve clarity and support clinical decision-making.
Findings from the review were used to validate automated Epic® data feeds, ensuring accurate representation of therapeutic trough attainment and AKI indicators. Based on these insights, targeted education was provided to clinical pharmacists on interpreting dashboard metrics and using the tool for ongoing dosing evaluation. The finalized dashboard enables monitoring of vancomycin dosing, supports trend and outlier identification, and provides a foundation for continuous quality improvement.

Conclusion

This project developed and validated a sustainable, ongoing quality metric for vancomycin dosing, supported by an interactive Power BI dashboard. Although this project was conducted around vancomycin troughs, the concept is also applicable to Area Under the Curve (AUC) targets. The goal of this initiative is to provide actionable information to clinical pharmacists to assess and improve vancomycin dosing accuracy, with the intent of increasing the proportion of trough levels within the therapeutic range while supporting ongoing renal safety surveillance.

Contact: [email protected]

Purpose/Background: Previously, Huntsville Hospital implemented several multimodal analgesia order sets intended to enhance postoperative outcomes by reducing opioid utilization, leading to reduced hospital length of stay, early ambulation, and improved overall outcomes. Although initial findings were favorable, the emergence of new evidence necessitates re-evaluation of the protocol to ensure continued alignment with current best practices. This project will compare the utilization and effectiveness of the previous order sets with the development and implementation of updated order sets incorporating current literature.
Methodology: This was a single-center, Institutional Review Committee (IRC)–approved, pre–post analysis conducted at Huntsville Hospital to evaluate existing multimodal analgesia protocols. The primary objective was to reduce opioid usage by optimizing multimodal regimen. The secondary objective was to characterize utilization patterns of the order sets by order set and prescriber type. Data was reviewed for patients receiving select order sets following surgery between February 1, 2025, and August 31, 2025.
Baseline demographic variables, including sex, age, gender, and related allergies, were obtained from the electronic health record. Literature review was conducted, and associated findings was utilized to update the respective order sets.  Opioid use in morphine equivalents and length of stay will be compared between the two groups. Descriptive analyses were performed to assess adherence to the existing multimodal order set and to identify discrepancies between current practice and recommendations described in the literature.
Results: Pre-intervention data analysis demonstrated that multimodal analgesia utilization varied across specialties, with acetaminophen being the most consistently used agent. Additionally, existing order sets were not aligned with current primary literature recommendations.
Conclusion: Guided by pre-intervention data analysis and literature review, recommendations for order set modifications were developed. Post-intervention data assessing primary and secondary outcomes are currently pending.

Background & Purpose:
Herpes zoster, commonly known as shingles, results from reactivation of latent varicella zoster virus and is associated with substantial morbidity, particularly among older adults and immunocompromised populations. Veterans receiving immunosuppressive therapy are at increased risk for severe disease, including postherpetic neuralgia and hospitalization. The recombinant herpes zoster vaccine (Shingrix) is a non-live, adjuvanted vaccine with efficacy exceeding 90% and is recommended for adults aged ≥50 years and immunocompromised adults aged ≥19 years. Despite these recommendations, vaccination rates among immunocompromised Veterans remain suboptimal. Within the Veterans Health Administration, Shingrix administration requires provider ordering and coordination of nurse clinic visits, which may introduce logistical barriers and contribute to delayed vaccination or incomplete series. Pharmacist-led interventions that incorporate patient education and care coordination may improve vaccine uptake by addressing hesitancy and streamlining the vaccination process.

Methods:
A pharmacist-led telephone intervention was conducted at the Ralph H. Johnson Veterans Affairs Health Care System in Charleston, South Carolina. Veterans prescribed immunosuppressive therapy who lacked one or both doses of the recombinant zoster vaccine were identified using a National VA Rheumatology Immunization Population Management Tool. Eligible patients were those enrolled in specialty clinics and prescribed a biologic, conventional synthetic, or targeted synthetic disease-modifying antirheumatic drug or systemic glucocorticoid. A random sample of 100 patients was selected. A PGY1 pharmacy resident conducted structured telephone outreach using motivational interviewing techniques to assess vaccine acceptance, explore concerns, and provide tailored education regarding shingles risk and vaccine safety. For patients who agreed to vaccination, orders for Shingrix and nurse clinic appointment were placed for one or both doses, if applicable. Vaccination outcomes and documented reasons for acceptance or refusal were tracked through the electronic medical record.

Results:
Sixty-six of 100 patients contacted agreed to receive Shingrix following pharmacist intervention, 30 declined, and 4 were previously vaccinated without documentation. Thirty-nine patients (59.1%) completed the vaccination series during the project period. Completion rates were higher among patients requiring one dose compared to two doses (72.2% vs 54.2%, p = 0.18). No significant differences in completion were observed by sex or ethnicity. Among patients who did not complete vaccination, over half were due to appointment-related barriers, including cancelled visits or failure to schedule.
Vaccine acceptance was most commonly associated with improved awareness, perceived benefit and risk reduction, and provider engagement that addressed questions, misinformation, and resolved logistical barriers. Patients frequently cited increased understanding of the need to complete the two-dose series and elevated herpes zoster risk in the setting of immunosuppressive therapy as key drivers of acceptance. Reasons for vaccine decline included vaccine hesitancy, low perceived risk, desire for additional time, information, or provider input, and scheduling barriers.

Conclusion:
Pharmacist-led outreach improved willingness to receive Shingrix among immunocompromised Veterans; however, series completion remained variable and was frequently impacted by logistical barriers. These findings highlight the importance of coordinated workflows and multidisciplinary collaboration to support vaccination beyond initial patient agreement. Addressing patient-specific concerns and streamlining scheduling processes can improve vaccine uptake and series completion while reinforcing the pharmacist’s role in vaccine stewardship.

Background:  
A common complication of traumatic and non-traumatic neurological injuries is elevated intracranial pressure (ICP). Preventing elevated ICP is critical for minimizing secondary neurologic injuries. First line pharmacologic therapies include hyperosmolar agents, such as hypertonic saline and mannitol. In the 2020 Neurocritical Care Society (NCS) Guidelines for Acute Treatment of Cerebral Edema, hypertonic saline and mannitol are both considered viable options, and there is little data to support one therapy over another, with many patients receiving both in clinical practice. Hypertonic saline is a plasma volume expander and may lead to hypernatremia; while mannitol is an osmotic diuretic and may not be preferred in patients due to risk of acute kidney injury (AKI). 
Increased serum concentrations of mannitol may be associated with AKI due to renal vasoconstriction. While most hospitals do not directly measure mannitol serum concentrations, osmolar gap monitoring has emerged as a surrogate marker for serum levels of mannitol. The NCS Guidelines recommend monitoring osmolar gap over serum osmolarity to assess the risk of AKI; however, the osmolar gap value at which to withhold mannitol remains unclear. The aim of this study is to determine risk factors for AKI among patients who receive mannitol for elevated ICP.  
 
Methods:  
This was a multi-center, retrospective, observational cohort study to identify the risk factors for AKI in patients who receive mannitol to manage elevated ICP. Patients were included if admitted to the Neurocritical Care or Trauma ICU between March 1st, 2024, and November 1st, 2025, received at least 1 g/kg of mannitol, and at least 2 doses of mannitol during admission. Patients were excluded if they had end-stage renal disease or have a baseline dependence on renal replacement therapy.  The primary outcome was identification of risk factors associated with AKI. Univariate logistic regression analysis was used to determine which variables are associated with AKI. We included variables such as peak osmolar gap, total mannitol dose administered, age, comorbid conditions, and number of nephrotoxic medications.  The secondary outcome was to compare characteristics of patients with AKI to those without AKI. Other statistical analyses were descriptive and inferential in nature.  
 
Results:   
A total of 230 patients who received mannitol between March 1st, 2024, and November 1st, 2025, were screened for inclusion. Of these, 102 patients met inclusion criteria and were included in the study cohort. The most common reason patients were not included was receiving less than 1 g/kg of mannitol and/or fewer than two doses of mannitol during admission. 
Baseline characteristics included patients that were predominantly male (56%) and White (53%), with a median age of 58 years. Most patients were admitted with a nontraumatic neurologic injury (94%) and were managed in the Neurocritical ICU (95%). Baseline serum creatinine was higher in patients who developed AKI compared to those who did not (1.02 mg/dL vs 0.79 mg/dL (p = 0.001). 
Among the 102 included patients, 28 developed AKI as defined by the Stage 1 AKI KDIGO Criteria. A univariate logistic regression analysis was conducted to evaluate differences between patients with and without AKI. Baseline serum creatinine (p = 0.010) and APACHE-II score (p = 0.032) were significantly associated with AKI. The median [IQR] peak serum creatinine among patients with AKI was 1.66 [1.41 - 2.46] mg/dL. 

Conclusions: 
In a univariate logistic regression analysis, higher baseline serum creatinine and APACHE II scores were associated with an increased risk of AKI. Factors associated with administration of mannitol, such as peak osmolar gap and total mannitol dose administered, were not associated with an increased risk of AKI.  

COMPARING TIME-TO-TARGET MEAN ARTERIAL PRESSURE WITH WEIGHT BASED VS. NON-WEIGHT-BASED NOREPINEPHRINE
Logan Turner, PharmD; Brian Hairston, PharmD, MBA; Allison Jolley, PharmD, BCCCP; Andrew B. Watkins, PharmD, BCIDP

FMOL Health | St. Dominic- Jackson, MS

Background/Purpose: Septic shock is a life-threatening manifestation of sepsis characterized by persistent hypotension requiring vasopressor therapy to maintain a mean arterial pressure (MAP) ≥ 65 mmHg. Norepinephrine is the first-line vasopressor recommended in current guidelines; however, there is no standardized dosing strategy, and institutions utilize either weight-based (mcg/kg/min) or fixed (mcg/min) dosing protocols. At FMOL Health | St. Dominic, practice transitioned from fixed dosing to weight-based dosing in 2022. The purpose of this study is to compare time to goal MAP between weight-based and fixed norepinephrine dosing strategies in critically ill adults with septic shock.


Methodology: This quasi-experimental, retrospective cohort study will evaluate adult ICU patients treated for septic shock during two time periods: January 1, 2019 to December 31, 2019 (fixed dosing cohort) and January 1, 2023 to December 31, 2023 (weight-based dosing cohort). Patients will be identified using electronic health record data. Eligible patients will include adults ≥ 18 years of age admitted to the ICU during the study periods with a diagnosis of septic shock requiring initiation of norepinephrine infusion to maintain a goal MAP ≥ 65 mmHg. Septic shock will be defined using clinical documentation and qSOFA criteria. Patients will be excluded if norepinephrine was initiated for indications other than septic shock, if vasopressors were started prior to ICU admission, if they are pregnant or incarcerated, or if key norepinephrine infusion data are incomplete or missing.


Results: A total of 100 patients were included, with 50 patients in each group. There was no significant difference in median time to target MAP between the weight-based and non-weight-based groups (41 vs 46 minutes; p = 0.73). Weight-based dosing was associated with significantly higher initial and maximum norepinephrine infusion rates. No statistically significant differences were observed in secondary outcomes, including AKI, mortality, or ICU length of stay.


Conclusions: Weight-based norepinephrine dosing did not improve time to target MAP and resulted in higher infusion rates without clear clinical benefit, suggesting no advantage over non-weight-based dosing in patients with septic shock. Although the overall incidence of acute kidney injury did not differ between groups, patients who developed AKI in the weight-based group received higher maximum norepinephrine doses, highlighting a potential safety concern.