UM Focus

Reduce cost and length of ventilation

Bypass intubation when safe

By Elgin K. Kennedy, MD

Editor, The Assertive Utilization and Quality Report

Mechanical ventilation remains the most common reason for admission to hospital intensive care units (ICUs). Unfortunately, mechanical ventilation is uncomfortable, making sedatives, analgesics, and muscle relaxants necessary. Furthermore, prolonged ventilation is accompanied by significant risks of costly complications, including pneumonia, pneumothorax, and other lung injuries.

A recent study showed how hospitals can substantially reduce the length of ICU ventilation, the number of ventilation complications, and the cost of ICU care.1 Simply notifying the attending physicians about their patients’ successful completion of a simple trial of unassisted breathing resulted in a shortening of the duration of mechanical ventilation by one and one-half days.

ICU respiratory therapists screened ICU ventilator patients to determine which ones might complete a two-hour trial of unassisted breathing. Protocol patients passing the screen were started on two-hour trials. Nothing out of the ordinary was done for the control patients. When a protocol patient passed the two-hour trial, the attending physician was notified, and the patient was typically extubated the same day.

The results of this relatively simple protocol were startling:

• Length of mechanical ventilation

Protocol Control

4.5 days 6.0 days

• Median interval between the time a patient met the screen and the discontinuation of mechanical ventilation

Protocol Control

1.0 days 3.0 days

• Total complication rate (patient extubation, reintubation, tracheostomy, mechanical ventilation more than 21 days)

Protocol Control

30 (20%) 62 (41%)

• Reintubations following extubation

Protocol Control

6 (4%) 15 (10%)

• Median costs for the intensive care unit

Protocol Control

$15,740 $20,890

• Total hospital costs

Protocol Control

$26,299 $29,048

• Hospital Mortality

Protocol Control

56 (38%) 60 (40%)

Three hundred randomized patients took part in the study; half were assigned to the protocol group. Unlike other investigations that have studied only postsurgery patients, this patient population also included medical patients and nonsurgical cardiac patients who had significant coexisting conditions. The hospital mortality for these patients was approximately 39%.

Each morning the ICU respiratory therapist performed the following five screens on all ventilator patients:

• ratio of PaO2 to FiO2 [>] 200;

• positive and end-expiratory pressure [] 5 cm of water;

• intact airway reflexes (i.e., adequate cough during suctioning);

• respiratory frequency/tidal volume [] 105 breaths/min./liter;

• no infusions of vasopressor agents or sedatives (however, dopamine could be given in doses not exceeding 5µg per kg of body weight per minute, and intermittent dosing of sedatives was allowed).

No complications occurred during the screening tests or during the trials of unassisted breathing.

Protocol patients who passed all five screens began a trial of unassisted breathing later the same morning. This was accomplished with either a T-tube or with a ventilatory circuit using "flow triggering" (rather than triggering by pressure). The patients were provided a continuous positive airway pressure of 5 cm of water.

The trial of unassisted breathing was monitored in the usual way by the ICU nurse and respiratory therapist, and was terminated immediately upon the identification of any of the following conditions:

• respiratory rate [>] 35 breaths per minute lasting five minutes or longer;

• arterial oxygen saturation below 90%;

• heart rate [>] 140/minute;

• sustained changes in heart rate of 20%;

• systolic BP [> 180 mm Hg or [• increased anxiety;

• diaphoresis.

When the two-hour trial of unassisted breathing was completed successfully, the physician was notified, and a message placed in the chart that stated: "Your patient has successfully completed a two-hour trial of spontaneous breathing and has an 85% chance of successfully staying off mechanical ventilation for 48 hours."

While patients in the control group also were screened, the results were not made available to their attending physicians, and, of course, the control patients did not undergo the two-hour trial of unassisted breathing unless their attending physicians independently chose to order it.

[Editor’s note: The preceding article was first published in The Assertive Utilization and Quality Report — $60 per year (12 issues). For more information, contact Elgin Kennedy, MD, at 204 Second Ave., No. 334, San Mateo, CA 94401. Telephone: (415) 348-3647.]


1. Ely EW, Baker AM, Dunagan DP, et al. Effect on the duration of mechanical ventilation of identifying patients capable of breathing spontaneously. N Engl J Med 1996; 335:1,864-1,869.