Opportunities to Improve Antibiotic Appropriateness in U.S. ICUs

A Multicenter Evaluation

Kavita K. Trivedi, MD; Rachel Bartash, MD; Alyssa R. Letourneau, MD, MPH; Lilian Abbo, MD; Jorge Fleisher, MD; Christina Gagliardo, MD; Shannon Kelley, MPA; Priya Nori, MD; Gunter K. Rieg, MD; Phyllis Silver, MEd; Arjun Srinivasan, MD; Jaclyn Vargas, MD; Belinda Ostrowsky, MD, MPH


Crit Care Med. 2020;48(7):968-976. 

In This Article

Abstract and Introduction


Objectives: To use a standardized tool for a multicenter assessment of antibiotic appropriateness in ICUs and identify local antibiotic stewardship improvement opportunities.

Design: Pilot point prevalence conducted on October 5, 2016; point prevalence survey conducted on March 1, 2017.

Setting: ICUs in 12 U.S. acute care hospitals with median bed size 563.

Patients: Receiving antibiotics on participating units on March 1, 2017.

Interventions: The Centers for Disease Control and Prevention tool for the Assessment of Appropriateness of Inpatient Antibiotics was made actionable by an expert antibiotic stewardship panel and implemented across hospitals. Data were collected by antibiotic stewardship program personnel at each hospital, deidentified and submitted in aggregate for benchmarking. hospital personnel identified most salient reasons for inappropriate use by category and agent.

Measurements and Main Results: Forty-seven ICUs participated. Most hospitals (83%) identified as teaching with median licensed ICU beds of 70. On March 1, 2017, 362 (54%) of 667 ICU patients were on antibiotics (range, 8–81 patients); of these, 112 (31%) were identified as inappropriate and administered greater than 72 hours among all 12 hospitals (range, 9–82%). Prophylactic antibiotic regimens and PICU patients demonstrated a statistically significant risk ratio of 1.76 and 1.90 for inappropriate treatment, respectively. Reasons for inappropriate use included unnecessarily broad spectrum (29%), no infection or nonbacterial syndrome (22%), and duration longer than necessary (21%). Of patients on inappropriate antibiotic therapy in surgical ICUs, a statistically significant risk ratio of 2.59 was calculated for noninfectious or nonbacterial reasons for inappropriate therapy.

Conclusions: In this multicenter point prevalence study, 31% of ICU antibiotic regimens were inappropriate; prophylactic regimens were often inappropriate across different ICU types, particularly in surgical ICUs. Engaging intensivists in antibiotic stewardship program efforts is crucial to sustain the efficacy of antibiotics and quality of infectious diseases care in critical care settings. This study underscores the value of standardized assessment tools and benchmarking to be shared with local leaders for targeted antibiotic stewardship program interventions.


Each year in the United States, approximately 2.8 million people develop multidrug-resistant bacterial infections and at least 35,000 people die of these infections.[1] Use of antibiotics is the single most important modifiable risk factor for the development of multidrug-resistant bacterial infections.[2] Studies suggest 30–50% of hospital antibiotic use is unnecessary or inappropriate in acute care and outpatient settings but standardized data regarding appropriateness in multicenter studies in the United States is lacking.[3] The ability to identify and stop unnecessary use is essential to improve patient outcomes, promote patient safety, and reduce pressure for the emergence of resistance. Although there have been various definitions and studies to attempt to quantify inappropriate use as a definitive standard, a practical and widely used method of assessment does not exist.[4]

Antibiotic use in the ICU is unique due to the severity of illness inherent to this patient population and often leads to broad-spectrum antibiotic use even when not indicated.[5,6] An assessment utilizing national proprietary administrative data from the Truven Health Market Scan Hospital Drug Database for adult and pediatric inpatients from January 1, 2006, to December 31, 2012, identified ICU antibiotic utilization at 1,092 days of therapy/1,000 days-present compared with noncritical locations (720 d of therapy/1,000 d-present; p < 0.001).[7] Due in part to large volume antibiotic use, ICUs are epicenters for antibiotic resistance worldwide. In addition, patients are critically ill, require invasive devices, and multiple services coordinate input on treatment, which may contribute to antibiotic use. The use of vasopressors, hemodialysis, and transfusions also produce variable volumes of distribution, affecting pharmacokinetics and pharmacodynamics complicating appropriate antibiotic dosing.[8,9] A 72-hour snapshot study conducted in 67 ICUs from 32 U.S. hospitals in 2011 reported 330 of 660 empiric antibiotic courses ordered in critically ill patients were inappropriately continued for at least 72 hours in absence of adjudicated therapy.[10] Experts have highlighted the urgent need for antibiotic stewardship programs (ASPs) to target education and interventions focused on improving antibiotic use in the ICU; studies have demonstrated that ASPs can improve antibiotic use,[11–13] be cost-effective,[11,14] and safe[15] in the ICU setting, however, published literature on identifying inappropriate use in the ICU and how to address it is limited.[16]

In 2016, the Partnership for Quality Care (PQC) developed the Antibiotic Stewardship Initiative with input and collaboration from the Centers for Disease Control and Prevention (CDC). PQC is a coalition of healthcare providers and healthcare workers dedicated to promoting affordable, high-quality healthcare. PQC includes public, private, religious, teaching and nonprofit hospitals, integrated health systems as well as more than 1 million U.S. healthcare workers.

PQC organizations involved in the Antibiotic Stewardship Initiative conducted a multicenter assessment of antibiotic appropriateness in ICUs using an actionable standardized CDC tool that they validated in order to identify local improvement opportunities. Results of this multicenter appropriateness assessment are described here.