Current issues of ACP Journal Club are published in Annals of Internal Medicine


Protocol-directed weaning reduced time spent on mechanical ventilation

ACP J Club. 1997 Nov-Dec;127:68. doi:10.7326/ACPJC-1997-127-3-068

Source Citation

Kollef MH, Shapiro SD, Silver P, et al. A randomized, controlled trial of protocol-directed versus physician-directed weaning from mechanical ventilation. Crit Care Med. 1997 Apr;25:567-74.



To compare protocol-directed weaning from mechanical ventilation (MV) by nurses and respiratory therapists (RTs) with physician-directed weaning.


Randomized controlled trial (RCT) with follow-up to hospital discharge.


4 intensive care units (ICUs) in 2 U.S. university teaching hospitals.


357 patients (50% surgical, median age 66 y, 53% women) who were > 18 years of age and required MV. Exclusion criteria were head or facial burns or trauma, previous MV in another hospital, or MV needed for organ harvest.


179 patients were allocated to protocol-directed weaning, and 178 were allocated to physician-directed weaning. Protocol-directed weaning was done by the nurses and RTs when the underlying conditions for MV had resolved or substantially improved. Physicians were notified about the start of weaning and could change the criteria-based procedures. The criteria were a PaO2/FIO2 ratio > 200, positive end-expiratory pressure ≤ 5 cm water, heart rate < 140 beats/min, respiratory rate ≤ 35 breaths/min, awake and oriented mental status, and no requirement for vasoactive or inotropic agents. Standard criteria for weaning failure were also preestablished. 3 separate weaning protocols were designed to facilitate acceptance in the 4 units. All protocols used pressure-support ventilation with emphasis on spontaneous breathing trials, pressure support, or intermittent mandatory ventilation. Physician-directed weaning was at the physicians' discretion; nurses and RTs could not assess for weaning readiness or initiate or advance weaning in these patients without physician orders.

Main outcome measures

Duration of MV. Secondary outcomes were hospital mortality, need for MV > 7 days, reintubation, length of stay, and costs.

Main results

Patients in the protocol-directed weaning group were less severely ill at baseline (P = 0.03), had a shorter duration of MV before weaning (39.6 vs 58.3 h, P = 0.016), and had a shorter total time on MV (69.4 vs 102 h, P = 0.029) than did patients in the physician-directed group. The groups did not differ for proportion of patients who required MV for > 7 days (12% vs 17%, P = 0.13), reintubation (13% vs 10%, P = 0.42), hospital mortality (22% vs 24%, P = 0.78), hospital length of stay (12.7 vs 14.2 d, P = 0.51), or cost (U.S. $27 439 vs $27 680, P = 0.93).


Protocol-directed weaning from mechanical ventilation implemented by nurses and respiratory therapists reduced time until the start of weaning and total time patients spent on mechanical ventilation but did not differ for other outcomes when compared with physician-directed weaning.

Source of funding: In part, Barnes-Jewish-Christian Hospitals Innovation in Healthcare Program.

For article reprint: Dr. M.H. Kollef, Pulmonary and Critical Care Division, Washington University School of Medicine, Box 8052, 660 South Euclid Avenue, St. Louis, MO 63110, USA. FAX 314-362-1334.


Although the protocol-driven "team approach" using nonphysicians, such as RTs and nurses, has been avidly promoted (1), until recently no RCTs had been done to support such a technique. This study by Kollef and colleagues is the second RCT to show a reduction in duration of MV using protocol-directed compared with physician-directed weaning.

This study is important because both medical and surgical patients were included. Nurses and RTs started weaning sooner and completed the job more quickly than physicians without adversely affecting outcomes. Limitations include a higher severity of illness in the control group and 3 different weaning protocols to accommodate varying ICU practices. This variation indicates the reluctance that physicians may have in relinquishing control over what they believe is the "right way" to manage their patients.

The RCT we did (2) differed in several substantial aspects. The population was entirely medical, and 25% of patients had an acute coronary condition. Our study showed that a 2-step objective approach that included rapid daily screening by RTs in conjunction with spontaneous breathing trials and physician prompts reduced weaning time and duration of MV by 2 days. In addition, patients in the protocol group had 50% fewer complications (including re-intubation and prolonged MV > 21 d) and had a reduced cost of care in the ICU (reduction > U.S. $5000 per patient).

A practical problem experienced by Kollef and colleagues was achieving adequate compliance with the protocol. Implementing this approach will require a change of culture on the part of RTs, nurses, and physicians. Taken together, these 2 studies support using a systematic, objective "team approach" to decide how rapidly patients should be liberated from MV in both medical and surgical intensive care settings.

E. Wesley Ely, MD, MPH
Bowman Gray Medical CenterWinston-Salem, North Carolina, USA


1. Cohen IL. Weaning from mechanical ventilation—the team approach and beyond. Intensive Care Med. 1994;20:317-8.

2. 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;335:1864-9.