This study examines the clinical outcomes and healthcare costs associated with gram negative bacteremia at Duke University Medical Center from 2002-2015.

Specific Aim 1: Define the clinical impact of gram negative bacterial (GNB) bloodstream infections (BSI). In this objective, the investigators will describe the prevalence, source of bacteremia, and location of acquisition (i.e., community-acquired, healthcare-associated, hospital-acquired) of GNB BSI. The investigators will also examine the outcomes associated with GNB BSI including length of hospital stay, length of ICU stay, in-house mortality, and hospital costs.

Specific Aim 2: Describe antibiotic resistance patterns in GNB BSI, and explore associations between antibiotic resistance and clinical and economic outcomes. In this objective, the investigators will describe the prevalence and antibiotic susceptibility profiles of multidrug resistant phenotypes including MDR bacteria, which are defined by resistance to three or more drug classes, ESBL-producers, and CRE. The investigators will examine the differences in mortality, resource use (length of hospital stay, length of ICU stay), and direct medical costs between cases with and without antibiotic resistance.

Inclusion Criteria:

Since January 2002, one investigator (VGF) has received daily reports from the clinical microbiology laboratory on all hospitalized patients at Duke University Medical Center with one or more blood culture(s) positive for gram negative rods. Patients were then evaluated within 36 hours of the detection of bacteremia for clinical evidence of infection

Exclusion Criteria:

Patients were excluded from the study for the following reasons: age less than 18 years, polymicrobial infection (blood culture positive for more than one pathogen), neutropenia (white blood cell count less than 1.0 x 109/L), or death prior to the return of positive blood cultures. In order to preserve the independence of observations, only the initial episode of gram negative bacteremia was included in the analysis (e.g., patients were only included once). The study was approved by the Duke University Medical Center Institutional Review Board.

The Duke Bloodstream Infection Biorepository (BSIB) contains prospectively collected clinical data on ~2000 unique inpatients at Duke with monomicrobial gram negative bacteria bloodstream infections.