Inappropriate antibiotic use due to decreased compliance with a ventilator-associated pneumonia computerized clinical pathway: implications for continuing education and prospective feedback.

STUDY OBJECTIVE

To assess the impact of noncompliance with a ventilator-associated pneumonia (VAP) computerized clinical pathway (CCP) on antibiotic use after removal of prospective antibiotic stewardship resources.

DESIGN

Retrospective, observational, quasi-experimental study.

SETTING

Three intensive care units (medical, surgical, and neurotrauma) in a large, tertiary care hospital.

PATIENTS

A total of 136 patients with culture-positive VAP; 72 were treated from September 2006-August 2007 (period 1), during which use of the CCP was mandatory along with aggressive stewardship support, and 64 were treated from September 2009-April 2010 (period 2), during which use of the CCP was voluntary.

MEASUREMENTS AND MAIN RESULTS

Compliance with use of the CCP was 100% during period 1 and 44% (28/64 patients) during period 2. For the 36 patients (56%) whose antibiotic selection did not comply with the CCP, empiric antibiotics were selected by provider discretion. Most patients had late-onset VAP and were similar with respect to age, sex, and comorbidities between the two periods. Staphylococcus aureus (11-17% methicillin-resistant S. aureus) and Pseudomonas aeruginosa were the most common pathogens during both periods. The proportion of patients with appropriate antibiotics within 24 hours of VAP identification was not significantly different between period 1 (70.8%) and period 2 (56.3%, p=0.112). During period 2, patients who were treated according to the CCP were more likely to receive appropriate antibiotic therapy compared with patients treated according to provider discretion (82.1% vs 36.1%, p ≤ 0.001). Time to appropriate therapy was also shorter for patients treated according to the CCP (mean ± SD 0.43 ± 1.14 vs 1.29 ± 1.36 days, p=0.003). Treatment with the CCP was the only variable significantly associated with appropriate antibiotic therapy (odds ratio 4.8, 95% confidence interval 2.1-10.9). Mortality was not significantly different between period 1 and period 2, and only Acute Physiology and Chronic Health Evaluation II score and admission with a head injury were predictive of death. Finally, a greater proportion of patients treated with the CCP were de-escalated from anti- Pseudomonas β-lactams (85.0% vs 33.3%, p=0.006) when they were not necessary.

CONCLUSION

These data highlight the importance of continued stewardship resources after CCP implementation to ensure compliance and to maximize antibiotic stewardship outcomes.