Establishment of SEP-1 national practice guidelines does not impact fluid administration for septic shock patients

The Centers for Medicare and Medicaid Services introduced the Early Management Bundle, Severe Sepsis/Septic Shock (SEP-1) as a national quality measure in October 2015. The purpose of SEP-1 is to facilitate the efficient, effective, and timely delivery of high-quality care to patients presenting along the spectrum of sepsis severity.

The primary aim of this study was to investigate whether provider practice surrounding emergency department (ED) fluid management of suspected septic shock patients was impacted by SEP-1.

The study was a retrospective observational analysis of 470,558 patient encounters at an urban academic center over a five-year period. The sample of suspected septic shock patients was defined by the following: blood cultures collected, antibiotics administered, and vasopressors initiated. Participants were divided into two cohorts based on date of presentation (Pre-SEP-1: May 1, 2013, - August 30, 2015, and Post-SEP-1: November 1, 2015, - February 28, 2018). The primary outcome was classified as a dichotomous variable based on whether the total volume of fluids administered equaled or exceeded the calculated weight-based (≥30 cc/kg) goal. Segmented logistic regression analyses were used to assess the immediate impact of SEP-1 as well as to compare the long-term trend of fluid volume administered between Pre-SEP-1 and Post-SEP-1 cohorts.

A total of 413 and 482 septic shock patients were included in the Pre-SEP-1 and Post-SEP-1 cohorts, respectively. There was no statistically significant change in weight-based fluid management between the cohorts. The odds of compliance with the weight-based goal decreased 22% immediately following dissemination of SEP-1, however, this was not statistically significant (log-odds = -0.25, p = 0.41). A positive trend in compliance was observed during both the Pre-SEP-1 and Post-SEP-1 periods with odds ratios increasing 0.005 and 0.018 each month, respectively, however, these findings were not statistically significant (log-odds = 0.005, p = 0.736, and log-odds = 0.018, p = 0.10, respectively).

Overall, there were no clinically or statistically meaningful changes in fluid volume resuscitation strategies for suspected septic shock patients following SEP-1. Broad mandates may not be effective tools for promoting practice change in the ED setting. Further research investigating barrier to changes in practice patterns surrounding fluid administration and other SEP-1 bundle elements is warranted.