For 10 years, Zahar et al prospectively observed 3,588 patients developing severe sepsis & septic shock who ended up in French ICUs. Their sample captured a broad array of infections acquired in the community, the hospital ward, or the ICU (about 1/3 each).
After multivariate assessment, they could not find an independent influence on mortality due to the infecting organism, nor where it was acquired. All that mattered was whether early and appropriate antibiotics were given. (In this database analysis, "early" was defined as receipt of antibiotics on the same day as the diagnosis of severe sepsis; "appropriate" meant the infecting organism was sensitive to the identified organism.) Crit Care Med 2011;39:1886-1895.
Their work adds to the mounting proof that time-to-antibiotics is critically important, perhaps the most important determinant, of survival in sepsis. The Surviving Sepsis folks offer extra advice here.