Saturday, March 20, 2010: 2:45 PM
Centennial III-IV (Hyatt Regency Atlanta)
Jon P. Furuno, PhD
,
University of Maryland School of Medicine, Baltimore, MD
Eli N. Perencevich, MD, MS
,
University of Maryland School of Medicine, Baltimore, MD
Anthony Harris, MD, MPH
,
University of Maryland School of Medicine, Baltimore, MD
Background: As of 2007, the
National Healthcare Safety Network (NHSN) reported that 32% of nosocomial
infections in the US were caused by Gram-negative bacilli (GNB), and the trend
is expected to increase. High rates of anti-microbial resistance and person-to-person
transmission of GNB have highlighted the need for effective infection control
interventions.
Objective: We sought to provide
a means of estimating the incidence rate reduction and financial impact
expected from a GNB infection control intervention prior to its implementation.
Methods: We conducted an
investigation of published literature to assess the practicality of either
hospital-wide or intensive care unit (ICU)-specific GNB infection control
interventions. After an initial PubMed search yielding articles related to GNB
infection control, we eliminated all those not meeting our pre-defined
inclusion criteria. We considered only quasi-experimental GNB-intervention
studies conducted in hospitals or adult ICUs in which the authors provided pre-
and post-intervention infection rates as infections and/or positive clinical
cultures per 1,000 patient-days. We then systematically reviewed and extracted
data from all relevant studies and computed the infection rate-reduction as a
percentage of the pre-intervention rate for each study. After aggregating these
reductions into a summary statistic, we conducted a business-case analysis from
a hospital's perspective to estimate cost thresholds for which yearly intervention
costs would be offset by cost savings from bloodstream infections (BSIs) prevented
per year, utilizing a pre-estimated cost of $18,432 per BSI. We then performed
two-way sensitivity analysis varying the bed size and pre-intervention
infection rates.
Results: We reviewed 304
studies produced by the key search. 13 were included in the final review based
on inclusion criteria. The mean reduction in GNB rates attributable to the
interventions was 61% (95% CI: 41% - 80%), corresponding to an overall
infection rate reduction of 20% (95% CI: 13% - 26%). Using the NHSN mean BSI rate
for ICUs of 1.1 per 1,000 patient days, a GNB intervention in a 15-bed ICU
resulting in a 20% rate reduction could be expected to save between $14,000 and
$29,000 (95% CI) over a one-year period (Figure). A 700-bed hospital with a BSI
rate of 0.33 (NHSN mean) could expect to save between $200,000 and $400,000.
Conclusions: These data
suggest that a strong business case can be made for implementing GNB infection
control interventions in most situations, especially in larger ICUs and
hospitals with high rates of GNB infection.