SCCM Resource Library-38-The Association Between Antibiotic Timing in Sepsis and Antibiotic Use for Potential Infection

Video Transcription

Hello, my name is Sarah Seeley, and today I'm going to talk about the association between antibiotic timing and sepsis and antibiotic use for potential infection.  First, I want to acknowledge that I have no financial conflicts to disclose and that this  talk does not represent the views of the U.S. government or the Department of Veterans Affairs. In the figure in the middle of this slide are data on antibiotic treatment times for  sepsis in the VA healthcare system from 2013 to 2018. Here we can see that in the VA healthcare system, patients with severe sepsis began  receiving faster antibiotic treatment in 2017 and 2018 compared to earlier years. In 2013 and 2014, represented by the solid black line, the distribution per time to first  antibiotics is more spread out with a median time to first antibiotics of 5.8 hours. In 2013 and 2014, 14% of sepsis patients had received their first antibiotics within two  hours of ED arrival. However, over the next four years, we can see antibiotic timing of patients continually  speeding up, first in 2015 and 2016, represented by the thick dotted line, and again in 2017 and 2018, represented by the smaller dots.  By these last two years, more than 20% of sepsis patients had received a first treatment in two hours compared to 14% in 2013 and 2014.  In 2017 and 2018, the median time to first antibiotics also sped up to 4.8 hours compared to the 5.8 hours in 2013 and 2014.  A variety of programs have incentivized rapid antibiotic treatment for severe sepsis such as this.  But at the same time, there's been growing concern that earlier antibiotic timing for sepsis may result in increased antibiotic treatment for more patients overall, including  those without infection. However, to date, there is little evidence to support this claim.  Therefore, in this study, we ask the question, is hospital-level acceleration in antibiotic timing for sepsis associated with increasing antibiotic use among all patients overall  who are hospitalized with potential infection? The data in our study come from the Nationwide Veterans Affairs Healthcare System electronic  health records from 2013 to 2018. Our cohort consists of approximately 1.1 million hospitalizations across 132 VA hospitals.  All patients in our cohort were admitted through the emergency department with a potential infection, which we define as two or more positive search criteria for white blood cells,  heart rate, body temperature, and respiratory rate. Among the 1.1 million hospitalizations, over 600,000, or about 55%, received antibiotics  within 48 hours of ED arrival, and over 117,000, or nearly 11% of the cohort, met criteria for sepsis, which we define using the CDC's adult sepsis event definition.  Our methodological approach proceeds in three steps. First, we fit multi-level linear models to estimate hospital-specific temporal changes  in antibiotic timing in severe sepsis, adjusting for patient case mix and illness severity. Then we categorize VA hospitals by their change in antibiotic timing from 2013 to 2018.  Here we use the hospital-specific temporal slopes from models in step one to classify all 132 VA hospitals into one of three categories of temporal change in antibiotic timing.  We refer to these categories of hospitals as rapid accelerators, slow accelerators, or flat accelerators, depending on the extent of their change over time.  Finally, in step three, we seek to understand the association between antibiotic timing in sepsis and antibiotic use for potential infection in all patients.  In this step, we use the hospital-level exposure variable from steps one and two, those classifications  of rapid, slow, and flat accelerators, to measure the temporal change in the percentage of all potential infection patients receiving antibiotics within 48 hours of ED arrival.  Again, in these models, we adjust for patient case mix and illness severity. At the bottom of this slide are slopes for each VA hospital from models in step one.  The rapid accelerator hospitals on the left are those that show the greatest decrease in time-to-first antibiotics for severe sepsis.  From the chart at the top, we can see that rapid accelerators have a median time-to-antibiotics of 6.4 hours in 2013, with a decrease to 4.5 hours in 2018.  The figure in the middle shows the slopes for the slow accelerator hospitals. As we can see, these are hospitals with less pronounced slopes than the rapid accelerators.  The decline in time-to-first antibiotics for sepsis from 2013 to 2018 for the slow accelerator hospitals dropped more steadily, from a median of 5.6 to 4.6 hours.  Finally, we can see the flat accelerator hospitals in the last set of slopes on the right. Compared to the slow and rapid accelerator hospitals, these hospitals had little change  in antibiotic prescribing for sepsis from 2013 to 2018. While there was a slight overall decrease in time-to-first antibiotics from 5.3 to 5.2  hours during this period, there is much less temporal change for this group of hospitals. We can also see that while these hospitals had more rapid antibiotic timing at baseline  compared to the other two categories, they had slower antibiotic timing for sepsis in 2018 compared to the rapid and slow accelerators.  In 2018, the median time-to-antibiotics was 5.2 hours compared to 4.5 and 4.6 hours for the rapid and slow accelerators in 2018.  In our final analytic step, we investigate the association between changes in antibiotic timing for sepsis and antibiotic use for all patients with potential infection, which  is roughly 1.1 million hospitalizations in our cohort. We find that in 2013, approximately 57% of patients with potential infection are treated  with antibiotics in the first 48 hours after ED arrival, and by 2018, approximately 55% of all patients are treated.  In the zoomed out figure on the left, the three types of accelerator hospitals are nearly indistinguishable in their changeover time.  When we zoom in, we can see some subtle differences in the rate of change for the rapid accelerator  hospitals compared to the slow and flat accelerators, with the rapid accelerators treating a slightly larger share of potential infection patients with antibiotics in 2018.  However, in statistical tests, we find no difference in antibiotic prescribing between these three groups.  In summary, we find that there was marked acceleration in time to antibiotics for sepsis across VA hospitals from 2013 to 2018, with the median time to antibiotics for sepsis  declining from 5.8 to 4.8 hours. Despite this acceleration in antibiotic timing for sepsis, we found no evidence that changes  in time to antibiotics for sepsis were associated with rising antibiotic use among all hospitalizations for potential sepsis.  I'd like to thank my co-authors, Dr. Prescott, Liu, and Donnelly, as well as Shirley Wong, Daniel Moley, and Connor Hogan.  And finally, if you would like to email me, here's my contact information, and I'm looking forward to taking any of your questions and also want to thank you for attending today's  session.  Thank you.

Video Summary

The study analyzed data from the U.S. Veterans Affairs Healthcare System to evaluate the association between accelerated antibiotic timing for sepsis and antibiotic use for potential infection. Researchers found that antibiotic treatment for sepsis has become faster over the years, with the median time to antibiotics declining. However, there was no evidence to suggest that the faster antibiotic timing for sepsis led to increased antibiotic use among all hospitalized patients with potential infection. The study concluded that efforts to accelerate antibiotic treatment for sepsis did not result in overall increased antibiotic use.

Asset Subtitle

Sepsis, Pharmacology, 2021

Asset Caption

The Society of Critical Care Medicine's Critical Care Congress features internationally renowned faculty and content sessions highlighting the most up-to-date, evidence-based developments in critical care medicine. This is a presentation from the 2021 Critical Care Congress held virtually from January 31-February 12, 2021.

Sarah Seelye

Introduction/Hypothesis: There are growing concerns that accelerating time-to-antibiotics in sepsis may result in treating more patients with antibiotics, including those without infection, and thereby contribute to antimicrobial resistance. To date, there is little evidence to support this claim. We sought to investigate whether hospital-level acceleration in antibiotic timing for sepsis is associated with increasing antibiotic use among patients hospitalized with potential infection.

Methods: We identified patients hospitalized at 132 Veterans Affairs (VA) hospitals (2013-2018) who were admitted via the emergency department (ED) with potential infection (defined as 2+ SIRS criteria) and sepsis (according to CDC's adult sepsis event definition). We fit multilevel linear models to estimate hospital-specific temporal changes in antibiotic timing among sepsis patients, adjusting for patient case-mix. Using slopes from the model, we classified hospitals into tertiles of antibiotic acceleration for sepsis. We also classified hospitals into tertiles of baseline antibiotic timing to account for floor and ceiling effects. In a second multilevel model, we measured temporal changes in the percentage of all potential infection patients receiving antibiotics within 48 hours of ED presentation. Using this model, we tested whether temporal changes in antibiotic use for potential infection differed by tertile of antibiotic acceleration for sepsis.

Results: Among 1,101,239 hospitalizations for potential infection in 132 VA hospitals, 2013-2018, 608,128 (55.2%) received antibiotics within 48 hours of ED presentation, and 117,435 (10.7%) met criteria for sepsis. The median time-to-antibiotics among sepsis hospitalizations declined from 5.2 hrs (IQR:3,8) in 2013 to 3.5 hrs (IQR:2,6) in 2018, with a majority of hospitals 111 (84.1%) accelerating time-to-antibiotics for sepsis. We found that the temporal change in antibiotic use in potential infection does not differ across the tertiles, Ï‡2=5.98 (p=0.650).

Conclusions: There was marked acceleration in time-to-antibiotics for sepsis across VA hospitals from 2013-2018. However, we found no evidence that antibiotic acceleration in sepsis was associated with rising antibiotic use among hospitalizations for potential infection.

Meta Tag

Content Type Presentation

Knowledge Area Sepsis

Knowledge Area Pharmacology

Knowledge Level Intermediate

Knowledge Level Advanced

Membership Level Select

Tag Antibiotics

Tag Infection

Year 2021

Keywords

U.S. Veterans Affairs Healthcare System

accelerated antibiotic timing

sepsis

antibiotic use

potential infection

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.