Objective: To measure the impact of a national propofol shortage on the duration of mechanical ventilation.
Design: Before-after study.
Setting: Three, noncardiac surgery, adult intensive care units at a 320-bed academic medical center.
Patients: Consecutive patients requiring mechanical ventilation ≥48 hrs, administered a continuously infused sedative ≥24 hrs, extubated, and successfully discharged from the intensive care unit were compared between before (December 1, 2008 to May 31, 2009) and after (December 1, 2009, to May 31, 2010) a propofol shortage.
Intervention: None.
Measurements and main results: Sedation drug use and common factors affecting time on mechanical ventilation were collected and if found either to differ significantly (p ≤ .10) between the two groups or to have an unadjusted significant association (p ≤ .10) with time on mechanical ventilation were included in a multivariable model. The unadjusted analyses revealed that the median (interquartile range) duration of mechanical ventilation increased from 6.7 (9.8; n = 153) to 9.6 (9.5; n = 128) days (p = .02). Fewer after-group patients received ≥24 hrs of continuously infused propofol (94% vs. 15%, p < .0001); more received ≥24 hrs of continuously infused lorazepam (7% vs. 15%, p = .037) and midazolam (30% vs. 81%, p < .0001). Compared with the before group, the after group was younger, had a higher admission Acute Physiology and Chronic Health Evaluation II score, was more likely to be admitted by a surgical service, have acute alcohol withdrawal, and be managed with pressure-controlled ventilation as the primary mode of mechanical ventilation. Of these five factors, only the Acute Physiology and Chronic Health Evaluation II score, admission service, and use of a pressure-controlled ventilation affected duration of mechanical ventilation across both groups. Although a regression model revealed that Acute Physiology and Chronic Health Evaluation II score (p < .0001), admission by a medical service (p = .009), and use of pressure-controlled ventilation (p = .02) each affected duration of mechanical ventilation in both groups, inclusion in either the before- or after-propofol shortage groups (i.e., high vs. low use of propofol) did not affect duration of mechanical ventilation (p = .35).
Conclusions: An 84% decrease in propofol use in the adult intensive care units at our academic institution as a result of a national shortage did not affect duration of mechanical ventilation.