Early studies focused on septic shock pathophysiology (Am J Physiol 1988; Chest 1990), the role of endotoxemia (J Clin Invest 1989; J Exp Med 1989; Chest 1991; N Engl J Med 1993; Infect Immun 1996), and the efficacy of various anti-endotoxin therapies (Antimicrob Agents Chemother 1989; J Clin Invest 1987; Pharm Res 1990; JAMA 1993; J Infect Dis 1994). Nitric oxide (NO) was examined as a mediator of septic shock (Crit Care Med 1993; JAMA 1996). Non-selective NO synthase inhibitors were found to be toxic and never beneficial (J Exp Med 1992; Crit Care Med 1998; Am J Respir Crit Care Med 1998). Increased NO production in endotoxin challenged volunteers was blocked by ibuprofen, but blood pressure was unaffected, suggesting that other mechanisms compensated to maintain vasodilation (J Pharmacol Exp Ther 1999). Severity of illness (risk of death) influenced the therapeutic efficacy of anti-inflammatory agents in septic shock (Am J Respir Crit Care Med 2002). L-arginine in a canine model of septic shock was harmful (Crit Care Med 2006). L-arginine had been marketed as a component of immunonutrition formulas. A canine septic shock model was redeveloped to balance animal welfare and scientific relevance (Am J Physiol Heart Circ Physiol 2007). The influence of risk of death on hydrocortisone efficacy in septic shock was investigated in a mouse model of E. coli pneumonia (Intensive Care Med 2008). Intra-aortic balloon counterpulsation prolonged survival in a hypodynamic canine model of Staphylococcal pneumonia-induced septic shock (Crit Care Med 2009). The U.S. Critical Illness and Injury Trials Group (USCIITG; www.usciitg.org/) was founded to create a clinical research framework that can reduce the barriers to investigation (Crit Care Med 2009). A meta-analysis of bundled care for septic shock demonstrated consistent and significant improvements in survival strongly associated with early and appropriate antibiotics (Crit Care Med 2010). SB203580, a p38 inhibitor, improved cardiac function but worsened lung injury and survival during E. coli pneumonia in mice (J Trauma 2010). Fluids and vasopressors were harmful in a rat model of anthrax lethal toxin (LeTx)-induced shock (Crit Care Med 2009). In our canine model of septic shock, edema toxin increased mortality when added to lethal toxin challenges (J Infect Dis 2010). Heparin failed to improve lung injury or survival in a mouse model of E. coli pneumonia (Crit Care Med 2011). Corticosteroids consistently reverse hypotension in septic shock, but variably effect survival. In our canine model of S. aureus pneumonia-induced septic shock, mineralocorticoid was only beneficial if given prophylactically, while glucocorticoid was most beneficial when given at the onset of infection (Crit Care Med 2012). Stress dose corticosteroids were only beneficial in sepsis with a high risk for death (Intensive Care Med 2012). Tigecycline was found to be associated with increased mortality and non-cure (Clin Infect Dis 2012). Staphylococcal enterotoxin including B (SEB) can trigger a lethal superantigen-mediated cytokine storm. Using an aerosolized-toxin mouse model, gene-expression changes across multiple organs implicated a host-wide IFN-response in SEB-induced death (PLoS One 2014). Hypothalamic-pituitary-adrenal (HPA) function was characterized in 101 canines with severe S. aureus pneumonia. HPA axis unresponsiveness and high aldosterone levels identified a septic shock subpopulation with poor outcomes (Am J Physiol Endocrinol Metab 2014). Infections caused by antibiotic-resistant gram-negative organisms are an important emerging threat to critically ill patients worldwide. Intravenous colistin use identified a severely ill population with a high probability of culture-confirmed, extensively drug-resistant bacteria (Clin Infect Dis 2015). Meningoencephalitis was documented for the first time in a patient with Ebola virus disease at the NIH Clinical Center (Ann Intern Med 2016). The Surfactant, Positive Pressure, and Pulse Oximetry Randomized Trial (SUPPORT) consent forms suggested that the low oxygen saturation arm was a widely practiced subset of usual care. An analysis of usual care practices demonstrated that oxygen targets and other interventions in the trial differed substantially from routine management (PLoS One 2016). A survey of intensivists found that diagnosing sepsis is extremely subjective and variable (Critcal Care 2016). Toxoplasmosis encephalitis was complicated by immune-reconstitution inflammatory syndrome (IRIS), in an allogeneic stem cell transplant patient (Bone Marrow Transplant 2016). A clinical surveillance definition accurately identified septic shock hospitalizations and suggested that the incidence of septic shock has risen and mortality rates have fallen much more slowly than previously estimated based on ICD-9 codes (Chest 2017). In a propensity-matched patient cohort with necrotizing fasciitis and vasopressor-dependent shock from 130 US hospitals, adjunctive intravenous immunoglobulin (IVIG) was administered infrequently and had no apparent impact on mortality (Clin Infect Dis 2017). Using electronic clinical data from over 400 hospitals, sepsis incidence was stable between 2009-2014; hospital mortality decreased but the combined outcome of death or discharge to hospice was stable (JAMA 2017). Meta-analysis of available data in patients with cardiovascular disease demonstrated that a restrictive vs. liberal transfusion trigger was associated with an increased risk of death and acute coronary events (Transfusion Med 2018). Difficult to Treat Resistance (DTR) was defined as resistance to all high-efficacy, low-toxicity antibiotics (penicillins, cephalosporins, carbapenems, and quinolones). The prevalence and impact on mortality of DTR gram-negative bloodstream infections was examined using a multicenter repository of hospitalized patients (Clin Infect Dis 2018). A case series from the NIH Clinical Center reported our experience with low-dose, short course alteplase in submassive pulmonary embolism (Blood Coagulation and Fibrinolysis, in press). Procalcitonin-guided antibiotic discontinuation and mortality in critically ill adults was investigated by meta-analysis. Evidence quality was low with a high risk of bias and a survival benefit was not observed among studies restricted to septic patients or with high adherence to procalcitonin-only algorithms (submitted). The sensitivity of claims data for sepsis and organ dysfunction relative to clinical data was evaluated at 193 U.S. hospitals. Variation in the completeness and accuracy of claims data for identifying sepsis and organ dysfunction was found to limit their usefulness (submitted). Using propensity-matched tracer antibiotic algorithms, attributable mortality was estimated for extensively drug resistant (XDR) gram-negative infections. XDR attributable mortality estimates varied by comparator agents, site of infection, community or hospital onset, and the severity of sepsis. (submitted). Using inpatient encounters across 156 hospitals from 2000-2014 in the Cerner Healthfacts database, discordant antibiotic therapy detrimentally impacted survival in gram negative and S. aureus bloodstream infections (IDSA abstract 2017 and in preparation).
Showing the most recent 10 out of 31 publications
