Objective: To determine the cost savings of replacing intravenous midazolam with enterally administered lorazepam in mechanically ventilated children who require long-term continuous sedation.
Design: Retrospective review of patients treated according to a preestablished pediatric intensive care unit (ICU) sedation protocol.
Setting: Twenty-six-bed pediatric ICU in a tertiary care children's hospital.
Patients: The records of 30 mechanically ventilated children were analyzed. The median age was 1.5 yrs and the median weight was 8.0 kg. Patients required continuous sedation for a total of 16 days (median).
Interventions: According to our pediatric ICU sedation protocol, midazolam infusion was continued until the hourly midazolam requirement was stable for at least 24 hrs. Thereafter, patients with a nasojejunal tube who were likely to require a minimum of three additional days of continuous sedation were transitioned from intravenous midazolam to enterally administered lorazepam. The goal in transitioning therapy was to titrate the lorazepam dose and reduce midazolam administration while maintaining an unchanged level of sedation.
Measurements and main results: The rate of midazolam administration was significantly (p<.05) reduced beginning on day 1 of lorazepam treatment. Midazolam was successfully discontinued in 24 (80%) patients in 3 days (median), and adequate and appropriate sedation was maintained with lorazepam monotherapy. Six patients in whom midazolam could not be discontinued experienced a 52% reduction in the rate of midazolam administration as a result of adding lorazepam. Total projected midazolam utilization was defined as the sum of midazolam administration before initiating lorazepam and the projected midazolam requirement after initiating lorazepam. Projected midazolam cost was calculated as the product of total projected midazolam utilization and midazolam acquisition cost. Actual expenditures for both midazolam and lorazepam were subtracted from the projected midazolam cost to calculate the estimated cost savings. Overall, midazolam utilization (in milligrams) was reduced by 46.7+/-27.6% (median 52). Total projected midazolam cost for the 30 patients was $90,771. The actual cost of midazolam and lorazepam combined was $47,867, resulting in a cost savings of $42,904.
Conclusions: Transitioning from intravenous midazolam to enterally administered lorazepam in critically ill children who require long-term sedation results in significant cost savings. The oral formulation of lorazepam was convenient to use, inexpensive, and effective in maintaining a continuous and appropriate level of sedation once midazolam was discontinued.