PURPOSE: ALARM was a prospective, observational, cohort study designed to capture and analyze management and outcome variable patterns for ventilator-associated pneumonia (VAP).
METHODS: Investigators from 20 institutions within the United States identified 398 (60.8% male; mean age, 58.3 yrs) ICU patients with a diagnosis of VAP. Clinical, diagnostic, and treatment data were gathered on each patient for the duration of ICU stay. Escalation/de-escalation of therapy was defined as switching to or adding a drug class(es) with a broader/narrower spectrum, respectively, or additional/less coverage as defined by the following spectrum categories: carbapenem> cefepime> ureidopenicillin/monobactam> quinolone> other/none.
RESULTS: The mean duration of mechanical ventilation prior to VAP diagnosis was 7.3 + 6.9 days (range, 0-44 days). Tracheal aspirate cultures (58.3% of patients), bronchoalveolar lavage fluid cultures (33.7%), or both (1.8%) were used to identify major pathogens, which included methicillin-resistant Staphylococcus aureus (14.8%), Pseudomonas aeruginosa (14.3%), other Staphylococcus species (8.8%), Klebsiella pneumoniae (3.3%), Enterobacter (3.3%), E coli (3.0%), and Acinetobacter (2.0%). The most common initial treatment regimens consisted of (alone or with other agents): cefepime (30.4%), piperacillin/tazobactam (27.9%), and vancomycin (17.8%). Mean duration of therapy was 11.8 days. Patterns of antibiotic therapy changes for VAP are shown in Table 1. Escalation of therapy occurred in 14.8% of patients and de-escalation in 22.4%. Overall mortality was 25.1%. Comparative mortality was lower among patients whose therapies were de-escalated (16.9%) compared to both patients undergoing escalation (42.4%) and those for whom therapy was neither escalated nor de-escalated (31.6%). Mean change in CPIS score at 72 hours was significantly less among patients who died (-0.10) compared with survivors (-2.35) (p<.05).
CONCLUSION: This multicenter study confirms that VAP in the ICU setting is associated with unacceptably high mortality rates.
CLINICAL IMPLICATIONS: Choices regarding initial antibiotic regimens and subsequent escalation/de-escalation of therapy have significant implications for patient outcomes.
DISCLOSURE: Matin Kollef, Grant monies (from industry related sources), The ALARM Study has been funded by a grant from Elan Pharmaceuticals.
Marin H. Kollef MD * Kenneth V. Leeper MD Antonio Anzueto MD Lee E. Morrow MD Lisa Benz-Scott PhD Frank J. Rodino MS Michael S. Niederman MD Washington University School of Medicine, St. Louis, MO
COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group