Abstract
Objective: To evaluate whether Intensive Care Unit (ICU) admission source predicts delay in the prescription of appropriate antimicrobial therapy in patients with septic shock.
Design, Setting and Participants: Single-centre, retrospective cohort study at a tertiary ICU in Wellington, New Zealand. All adult patients admitted with septic shock between 2003 and 2010 were included in the study.
Main outcome measures: Data including demographics, APACHE III score and diagnosis, source of admission, microbiological data, onset of hypotension, and the time to appropriate antimicrobial therapy (TAAT) were collected from the Wellington ICU database and the medical records. Outcome data including the duration of mechanical ventilation, ICU and in-hospital mortality, hospital length of stay, and vital status at 6 months were collected.
Results: 107 patients were identified. 93 had sufficient data to calculate TAAT (median 16 minutes, interquartile range [IQR] 0-105). Cox modelling demonstrated no significant difference in TAAT based on admission source (log rank test=0, df=2, p>0.5); however, when a sensitivity analysis was performed excluding 43 patients in whom antibiotics were administered prior to the onset of shock and one outlying patient with fungaemia, source of admission was found to be predictive of antimicrobial delay (log rank test=12.14, df=2, p<0.005). Among those patients who experienced any delay in antibiotic administration, the median TAAT (IQR) for patients admitted from the ED, ward and regional hospitals was 58 (40-280), 180 (65-337), and 145 (120-544) minutes respectively.
Conclusions: Our study suggests that ICU admission source is predictive of antimicrobial delay in patients with septic shock.
Jeremy Fernando, Takayoshi Ikeda, Alex Psirides, Paul Young
