The impact of improper empirical usage of antipseudomonals on admission to an acute care hospital.

BACKGROUND:Many septic patients are receiving empirical antipseudomonal (or Gram-negative non-glucose fermenting [GNNGF]) coverage on admission to acute care hospitals, despite the fact that the indications are not scientifically established. Overuse of antipseudomonals might contribute to the burden of resistance. MATERIALS AND METHODS:Retrospective observational analyses of the characteristics of septic adult patients who received empirical antipseudomonals, along with its impact on outcomes, were executed at Shamir Medical Center, Zerifin, Israel (08-12/2016). Proper empirical antipseudomonal usage was defined by the following: (1) if the patient received the agents as per Infectious Disease Society of America (IDSA) guidelines; (2) if the patient had a positive multidrug-resistant organism (MDRO) test on his or her admission score (https://assafharofe.azurewebsites.net); or (3) if a GNNGF was the eventual causative pathogen. Risk factors and outcomes were queried by logistic and Cox regression. RESULTS:GNNGF was the causative pathogen in only 57 (3.7%) of 1536 patients with acute sepsis. There were 192 (13%) who received empirical antipseudomonals, of whom 161 (84%) were defined as proper. Patients who received empirical antipseudomonals were significantly older (P < 0.001), with higher indices of chronic and acute conditions, and higher rates of past MDRO carriage; 24 patients received empirical antipseudomonals only because of IDSA guidelines (15%), and that was an independent predictor for later acquisition (up to 90 days) of carbapenem-resistant A. baumannii (CRAB; odds ratio [aOR] = 7.1; P = 0.03). CONCLUSIONS:Improper empirical usage of antipseudomonals in acute care hospitals is common. Instituting empirical antipseudomonals solely due to IDSA guidelines was independently associated with later acquisition of CRAB. Empirical antipseudomonal usage should be based on scientifically established prediction tools and not on IDSA guidelines.