Early Combination Antibiotic Therapy Yields Improved Survival Compared to Monotherapy in Septic Shock: a Propensity-Matched Analysis
Total Page:16
File Type:pdf, Size:1020Kb
Early combination antibiotic therapy yields improved survival compared to monotherapy in septic shock: A propensity-matched analysis Anand Kumar, MD; Ryan Zarychanski, MD; Bruce Light, MD; Joseph E. Parrillo, MD; Dennis Maki, MD; Dave Simon, MD; Denny Laporta, MD; Steve Lapinsky, MD; Paul Ellis, MD; Yazdan Mirzanejad, MD; Greg Martinka, MD; Sean Keenan, MD; Gordon Wood, MD; Yaseen Arabi, MD; Daniel Feinstein, MD; Aseem Kumar, PhD; Peter Dodek, MD; Laura Kravetsky, BSc; Steve Doucette, MSc; the Cooperative Antimicrobial Therapy of Septic Shock (CATSS) Database Research Group Background: Septic shock represents the major cause of infec- day mortality (444/1223 [36.3%] vs. 355/1223 [29.0%]; hazard The .(0002. ؍ tion-associated mortality in the intensive care unit. The possibility ratio, 0.77; 95% confidence interval, 0.67–0.88; p that combination antibiotic therapy of bacterial septic shock im- beneficial impact of combination therapy applied to both Gram- proves outcome is controversial. Current guidelines do not recom- positive and Gram-negative infections but was restricted to pa- mend combination therapy except for the express purpose of broadening tients treated with -lactams in combination with aminoglyco- coverage when resistant pathogens are a concern. sides, fluoroquinolones, or macrolides/clindamycin. Combination Objective: To evaluate the therapeutic benefit of early combi- therapy was also associated with significant reductions in inten- nation therapy comprising at least two antibiotics of different sive care unit (437/1223 [35.2%] vs. 352/1223 [28.8%]; odds ratio, and hospital (0006. ؍ mechanisms with in vitro activity for the pathogen in patients with 0.75; 95% confidence interval, 0.63–0.92; p bacterial septic shock. mortality (537/1223 [53.6%]vs. 424/1223 [46.4%]; odds ratio, Design: Retrospective, propensity matched, multicenter, co- 0.69; 95% confidence interval, 0.59–0.81; p < .0001). The use of hort study. combination therapy was associated with increased ventilator ؍ Setting: Intensive care units of 28 academic and community (median and [interquartile range], 10 [0–25] vs. 17 [0–26]; p hospitals in three countries between 1996 and 2006. .008) and pressor/inotrope-free days (median and [interquartile .up to 30 days (007. ؍ Subjects: A total of 4662 eligible cases of culture-positive, range], 23 [0–28] vs. 25 [0–28]; p bacterial septic shock treated with combination or monotherapy Conclusion: Early combination antibiotic therapy is associated from which 1223 propensity-matched pairs were generated. with decreased mortality in septic shock. Prospective randomized Measurements and Main Results: The primary outcome of trials are needed. (Crit Care Med 2010; 38:000–000) study was 28-day mortality. Using a Cox proportional hazards KEY WORDS: antibiotic; combination; monotherapy; mortality; model, combination therapy was associated with decreased 28- sepsis; septic shock From Section of Critical Care Medicine (AK, LK), Arabia; Moses H. Cone Memorial Hospital (DF), Foundation, and the Alfred Deacon Foundation. Addi- Health Sciences Centre/St. Boniface Hospital, Univer- Greensboro, North Carolina; Laurentian University (AK), tional support was provided by through unrestricted sity of Manitoba, Winnipeg, Manitoba, Canada; Cancer Biomolecular Sciences Program and Department of grants from Eli-Lilly, Pfizer, Astellas Pharma, Merck, Care Manitoba (RZ), University of Manitoba, Winnipeg, Chemistry and Biochemistry, Sudbury, Ontario, Can- Wyeth, Bayer, Bristol-Myers-Squibb, and Astra- Manitoba, Canada; Cooper Hospital/University Medical ada; Section of Critical Care Medicine (PD), St. Paul’s Zeneca. Funding sources had no role in the design and Center (BL), Robert Wood Johnson Medical School, Hospital, University of British Columbia, Vancouver, conduct of the study; collection, management, analy- UMDNJ, Camden, New Jersey; Section of Infectious British Columbia, Canada; Ottawa Health Research sis, and interpretation of the data; and preparation, Diseases (DM), University of Wisconsin, Madison, Mad- Centre (SD), Ottawa, Ontario, Canada. review, or approval of the manuscript. ison, WI; Section of Infectious Diseases (DS), Rush Dr. Kumar has received grants from Wyeth, Astra- Dr. Kumar had full access to all the data in the University, Chicago, Illinois; Section of Critical Care Zeneca, Pfizer, and Rocheu. Dr. Parrillo consulted with study is responsible for the integrity of the database Medicine (DL), Jewish General Hospital, McGill Univer- Sangart, Artisan, Philips, and Immunetrics. He received sity, Montreal, Quebec City, Canada; Section of Critical a grant from the Robert Wood Johnson Foundation. Dr. and the accuracy of the data analysis. Care Medicine (SL), Mount Sinai Hospital, University of Mirjanezad consulted for the advisory boards of Scher- This specific research concept, the septic shock Toronto, Toronto, Ontario Canada; Department of ing-Plough Corporation and Pfizer. He received hono- database, and manuscript were developed by Dr. Ku- Emergency Medicine (PE), University Health Network, raria/speaking fees from Merck, Schering-Plough Cor- mar. Dr. Kumar and Mr. Doucette were responsible for University of Toronto, Toronto, Ontario, Canada; Surrey poration, Bayer, and Wyeth. He also received grants the methodological design issues and data analysis. All Memorial Hospital (YM), Surrey, British Columbia, Can- from Par101, C.diff, and INC Research. All other au- authors assisted with data interpretation and manu- ada; Richmond General Hospital (GM), Vancouver, Brit- thors have no potential conflicts of interest to disclose. script revisions. ish Columbia, Canada; Royal Columbian Hospital (SK), None of the authors have financial or personal For information regarding this article, E-mail: Vancouver, British Columbia, Canada; Royal Jubilee relationships or affiliations that could influence (or [email protected] Hospital/Victoria General Hospital (GW), University of bias) the decision regarding the analysis or manuscript Copyright © 2010 by the Society of Critical Care British Columbia, Victoria, British Columbia, Canada; in any regard. Medicine and Lippincott Williams & Wilkins Intensive Care Department (YA), King Saud Bin Abdu- Funding for this work was provided by the Mani- laziz University for Health Sciences, Riyadh, Saudi toba Health Research Council, Health Sciences Centre DOI: 10.1097/CCM.0b013e3181eb3ccd Crit Care Med 2010 Vol. 38, No. 9 1 everal studies have shown that specific criteria for septic shock as described death if the patient survived Ͻ24 hrs after appropriate antimicrobial ther- by the 1991 Society of Critical Care Medicine/ hypotension documentation. The second apy, defined as the use of at American College of Chest Physicians Consen- agent had to be started within 24 hrs of the least one agent with in vitro sus Statement on Sepsis Definitions (20). The first antibiotic or within 24 hrs of the onset of Sactivity for the isolated pathogen, leads to process used to identify the final study popu- hypotension (if the first agent was initiated lower mortality rates in life-threatening lation is outlined in Figure 1. before hypotension was documented). The infections associated with sepsis (1–5). combination of two -lactams or a -lactam The incremental benefit of combination Data Elements and Definitions and a glycopepide was not considered to rep- as opposed to single-agent antimicrobial resent antibiotic combination therapy (be- Clinical infection definitions were adapted cause all are cell-wall–active agents with sim- therapy in such situations is controver- from previous recommendations or studies ilar mechanisms of action) for purposes of this sial (6–10). (21–23). To qualify as potential pathogens analysis. Typical permutations of antibiotics Some clinical studies of bacterial in- causing shock, isolates from anatomical sites meeting criteria for combination therapy in- fection, including endocarditis, Gram- and/or blood cultures were required to have clude any two of a cell-wall–active agent (i.e., negative bacteremia, and neutropenic been obtained within 48 hrs of onset of shock. a -lactam or glycopeptide), aminoglycoside, sepsis (11–13), and animal models of se- A priori criteria were developed to uni- fluoroquinolone, or macrolide/clindamycin. vere infection (14–16) have supported formly determine the primary pathogen/ Questionable cases or data elements were the possibility of clinically relevant anti- pathogens and to assess the appropriateness of microbial synergism with appropriate antimicrobial therapy across participating in- reviewed and adjudicated by the principal in- combinations of antibiotics. However, stitutions (Appendices 1 and 2). The first use vestigator. Cases of septic shock associated two separate meta-analyses have failed to of any appropriate antimicrobial therapy (i.e., with negative or absent cultures and those demonstrate any consistent benefit with with in vitro activity for the primary isolated caused by yeast/fungi, anaerobes, and atypical combination therapy in immunocompe- pathogen or pathogens) was determined for all pathogens, such as Mycobacterium tuberculo- sis and Legionella species, were excluded (Fig. tent patients with sepsis and/or Gram- cases. For the purposes of this study, antibi- 1). Patients who did not receive any appropri- negative bacteremia (17, 18). otic monotherapy was defined as the adminis- ate antimicrobial therapy before death were The beneficial effect of early appropri- tration of any single, appropriate, intravenous, also excluded. ate antimicrobial therapy