Lead/Presenter: Geneva Wilson, COIN - Hines
All Authors: Wilson GM (Center of Innovation for Complex Chronic Healthcare, Hines ; Northwestern University, Feinberg School of Medicine), Fitzpatrick, MA (Center of Innovation for Complex Chronic Healthcare, Hines) Suda, KJ (Center for Health Equity Research and Promotion, Pittsburgh; University of Pittsburgh, University of Pittsburgh School of Medicine) Smith, BM (Center of Innovation for Complex Chronic Healthcare, Hines; Northwestern University, Feinberg School of Medicine) Gonzalez, B (Center of Innovation for Complex Chronic Healthcare, Hines) Jones M (VA Salt Lake City Healthcare System, Salt Lake City; Department of Medicine, University of Utah) Schweizer, ML (Center for Access & Delivery Research and Evaluation, Iowa City VA Health Care System, Iowa City ; Carver College of Medicine, University of Iowa) Evans, M (Lexington VA Medical Center, Lexington) Evans CT (Center of Innovation for Complex Chronic Healthcare, Hines; Northwestern University, Feinberg School of Medicine)
Objectives:
Carbapenem-resistant Enterobacterales bloodstream infections (CRE-BSI) are associated with significant mortality and have limited treatment options. Guidelines recommending newer antibiotics such as ceftazidime/avibactam were only recently published in 2020. Because these infections are rare, there is a paucity of information on effectiveness and the impact on mortality of different treatments. This study examines treatment regimens and associated mortality risk for patients with CRE-BSI.
Methods:
This retrospective cohort study identified hospitalized patients within the VA between 2013-2018 with a positive CRE blood culture and antibiotic treatment within 5 days of culture date. Primary outcomes were in-hospital, 30-day and 1-year all-cause mortality. Secondary outcomes included health care costs at 30-days and one-year and Clostridioides difficile infection six weeks post-culture date. The propensity for receiving each treatment regimen was estimated and used in multivariable logistic regression to assess the association between treatment and outcomes.
Results:
393 hospitalized patients with CRE-BSI were predominantly male (97%) and elderly (mean age = 71.0 + SD12.1). Diabetes and renal disease were the predominant comorbidities (42.7% and 46.8%, respectively). Carbapenems were the most frequently prescribed antibiotic class (47%), while ceftazidime/avibactam use increased from 9% in 2015 to 46% in 2018. In-hospital, 30-day, and one-year mortality was 37%, 36%, and 64%, respectively. In unadjusted analysis, ceftazidime/avibactam was associated with a lower likelihood of 30-day (OR = 0.43, 95% CI = 0.19, 0.96) and one-year mortality (OR = 0.45, 95% CI = 0.23, 0.88). However, after propensity score adjustment, there was no difference in mortality associated with any antibiotic treatment. Only 6% of the cohort had a positive C. difficile test and no difference was found in C. difficile incidence at six weeks post-infection by treatment. Average total 30-day health care total costs ranged from $43,695.50 for aminoglycosides to $48,352.60 for carbapenems, however there was no statistical difference at 30-days or 1-year post culture date by any treatment.
Implications:
In hospitalized Veterans with CRE-BSI, carbapenems were used most frequently for treatment. Although ceftazidime/avibactam treatment trended towards lower 30-day and one-year all-cause mortality, after propensity score adjustment, this was not significant. During the study timeframe there were no clinical guidelines that specifically addressed treatments for CRE infections. Although the use of ceftazidime/avibactam was not statistically associated with decreased mortality in this study, this was heavily influenced by its low usage due to it only being approved halfway through the study timeframe. There was no difference in cost between any of the treatment regimens.
Impacts:
This study provides information on the treatment outcomes of CRE-BSI which is a serious concern for the VA and associated with high Veteran mortality. The evidence from our study shows that there was no difference in outcome or cost for any of the treatment options evaluated suggesting that treatment regimens can be selected based on clinical success. This information can be used to update the current VA Directive and future guidelines for treatment of CRE in the VA.