June 1997

Prolonged sedation of critically ill patients with midazolam or propofol: impact on weaning and costs.

Barrientos-Vega R, Sanchez-Soria MM, Morales-Garcia C, Robas-Gomez A, Cuena-Boy R, Ayensa-Rincon A; Critical Care Medicine 1997;25:33-40.

[ see abstract below ]

Propofol has enjoyed widespread acceptance as a sedative-hypnotic given by continuous infusion in the intensive care unit (ICU). Although it lacks analgesic and amnestic effects, and is usually given in combination with an opioid such as fentanyl or morphine, its rapid titrateability, lack of accumulation, and reliably fast emergence have made propofol a desirable addition to our sedative regimens.

It is particularly popular among ICU nursing staffs, who appreciate its ability to render and keep their patients calm or insensible quickly, smoothly and predictably. Propofol's major limitation is its high acquisition cost, which can become formidable when infusions are given at a high dose for many days.

Midazolam, a benzodiazepine, is considerably less expensive than propofol, provides an equivalent quality of sedation, and has potent amnesic effects to boot. However, its main limitation is its propensity to accumulate when administered for longer than 24 hours. Sudden discontinuation may result in an acute withdrawal syndrome, and delayed emergence may occupy two or three days or more of costly ICU time.

In the January issue of Critical Care Medicine, Barrientos-Vega et al. shed a great deal of light on these issues. In a study on long term (mean 140 hours) sedation in the ICU, the authors compared a continuous infusion of propofol to that of midazolam, with intermittent morphine for analgesia. They concluded that although propofol is two to three times as expensive as midazolam, delayed emergence after prolonged midazolam infusion increased intubation and ICU time so that the overall ICU cost was equivalent.

The quality of sedation provided by the two agents was similar - in fact, the failure rate (defined as inadequate sedation at a ceiling infusion rate of 0.5 mg/kg/hr for midazolam and 6 mg/kg/hr for propofol) was a surprisingly high 20%. A further 10% of cases had propofol discontinued because of unacceptable hyperlipdemia.

The acquisition cost for propofol was about $1050 per patient and for midazolam about $380 per patient. However, the mean time from discontinuation of the drug to extubation was more than twice as long for midazolam as for propofol (about 98 versus 35 hours). When the cost of ICU care (including the ventilatory weaning time) was added to the drug acquisition cost, the overall cost for propofol-treated patients was less than midazolam, about $9,500 versus $10,800 per patient.

There is an important lesson to be drawn here, that could be equally applied to the operating room (OR). When we compute the cost saving accrued by acquiring a cheaper drug, we need to evaluate the impact on the total care of the patient. For example, the cost benefit of substituting pancuronium for vecuronium as the muscle relaxant of choice in the OR may be more than offset by an increased requirement for postoperative mechanical ventilation in patients whose renal function is even moderately impaired (creatinine clearance < 50 mL/min).

In our own surgical intensive care unit, the use of propofol infusion has been drastically curtailed as a cost-savings measure. The article by Barrientos-Vega et al. suggests that we may have thrown out the baby with the bath water!

Return to the Current Literature Review Front Page, or read the abstract:

ABSTRACT