BACKGROUND: Sepsis is a leading cause of hospital morbidity and mortality. While current Surviving Sepsis Guidelines emphasize early broad-spectrum antibiotic administration, they do not address sequencing. Evidence suggests that delays in gram negative coverage, against Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa, are associated with increased mortality, and administration of broad-spectrum beta lactams prior to vancomycin may improve survival. In practice, initiation of antibiotic therapy is delayed due to lack of IV access, blood culture collection, medication verification, and absence of standardized order sets. These limitations are greater among spinal cord (SCI) and traumatic brain injury (TBI) patients who have atypical presentations complicating early sepsis recognition. The aim of this study is to assess the order of appropriate antibiotic administration, delays in administration from lack of intravenous (IV) access, and timing of blood culture collection to support creation of a sepsis order set.
METHODS: This is a single-center, retrospective, quality improvement project performed at a rehabilitation center. Patients (≥18 years) admitted between July 1, 2023, and July 31, 2025, who met systemic inflammatory response (SIRS) criteria and received IV vancomycin and one gram-negative agent for sepsis or neutropenic indications were included. Patients were excluded if antibiotic timing was inadequately documented.
Data collected included patient demographics, injury type (SCI or TBI), sepsis indicators, timing of sepsis recognition, antibiotic timing and sequencing, blood culture collection, lactate measurement, IV fluid resuscitation, and IV access placement.
The primary outcome was sequencing, including gram-negative coverage first followed by gram-positive coverage antibiotics. Secondary outcomes included time between gram-negative and gram-positive therapy, percentage of patients with cultures drawn before antibiotic administration, frequency of IV-related delays, and type of IV line. Data was analyzed using descriptive statistics.
RESULTS: A total of 65 patients meeting SIRS inclusion criteria were evaluated. The median age was 34 years old (SD 18.8), with males comprising 81.5% of the cohort (n=53). Injury classifications included SCI (n=24, 37%), TBI (n=26, 41.5%), and dual (n=14, 21.5%) diagnoses.
When vancomycin was administered prior to gram-negative therapy, the mean delay to gram-negative coverage was 141 minutes. Upon further analysis, vancomycin was administered first in 15.4% of encounters, while gram negative agents were more frequently administered first, including piperacillin-tazobactam (57%), cefepime (20%), and meropenem (7.6%).
In 20% of encounters, blood cultures were either obtained after the first antibiotic dose or not obtained. IV access was not established prior to antibiotic ordering in 46.2% of patients. Antibiotics were administered via peripheral IV (75%), midline (11%), or peripherally inserted central catheter (14%).
CONCLUSIONS: Administering vancomycin first resulted in an expected delay of more than two hours before effective gram-negative coverage. This delay was likely due to the standard vancomycin infusion time at this facility. Because delays in antibiotic administration increase mortality in sepsis, education on Y-site compatibility is essential to allow compatible agents to be administered simultaneously and facilitate faster, more efficient antibiotic delivery. Moreover, 20% of blood cultures were collected inappropriately, compromising diagnostic accuracy and limiting targeted antibiotic therapy, highlighting the need for reinforcement of proper culture collection.
Additionally, nearly half of patients meeting sepsis criteria at the time of antibiotic ordering did not have IV access, reflecting a challenge unique to rehabilitation settings, where the goal is early discontinuation of IV lines to enhance mobility, minimize line associated complications, and support functional recovery. However, for sepsis, a delay in IV access inadvertently leads to a delay in antibiotic administration.
These findings show inconsistent sepsis workflows and clinically relevant delays which can impact patient outcomes, highlighting the need for a standardized sepsis order set in rehabilitation hospitals to streamline IV access, culture collection, and appropriate antibiotic prioritization.
