Clinical parameters predicting failure of empirical antibacterial therapy in early onset neonatal sepsis, identified by classification and regression tree analysis

Table 2 Distribution of patients with empiric antibiotic treatment failure

Failure of early empirical AB regimen - N (% of all study patients)	32 (11.6)
I Proven EOS with etiologic pathogen(s) resistant to initial empirical AB therapy - n (% of treatment failure)	10 (31.3)
died within 7 days	4
II clinical EOS and death (with autopsy confirming the diagnosis of congenital infection) and/or major sepsis-related complication (IVH III-IV) within the first week of life - n (% of treatment failure)	11^a (34.4)
died within 7 days	10
III clinical EOS and change of initial empirical AB therapy within 72 h with new regimen continued for more than 72 h - n (% of treatment failure)	11 (34.4)

All cases are represented once - in the "highest" applicable category, i.e. a neonate with culture positive sepsis with a pathogen, resistant to empiric regimen is included only in proven EOS category, independent of whether he died within 7 days or not. ^a includes an ELBW neonate with clinical EOS and IVH IV by the third day of life who died on Day 8, autopsy confirmed the diagnosis of EOS. AB - antibiotic; EOS - early onset sepsis; IVH - intraventricular haemorrhage

ISSN: 1471-2431