Using Protocols to Wean Patients from Ventilatory Support
By Robert M. Kacmarek, PhD, RRT, FCCM, FCCP
Over the past few years, heightened interest has been directed toward weaning from ventilatory support. Randomized, controlled, prospective trials have compared various approaches to weaning, with different findings.1,2 However, recently, two trials comparing protocolized therapist/nurse-directed weaning to individual physician ordering practices have shown more rapid weaning with similar or fewer complications with protocols than with individual physician orders.3,4
These trials have established the methodology to develop weaning protocols for all mechanically ventilated patients.
The Brochard and Esteban Trials
Much debate has occurred regarding the results of the prospective, randomized, controlled, multicenter weaning studies of Brochard et al1 and Esteban et al.2 Both of these trials used similar methods to identify patients for inclusion. Both compared similar approaches to weaning: T-piece, pressure support (PSV), and synchronized intermittent mandatory ventilation (SIMV). However, the Brochard trial found that weaning was most rapid with PSV, while the Esteban trial indicated that use of the T-piece was the best approach. Sorting out why the results of these trials were so different is difficult to do, but there are differences in the approaches used in each experimental arm in these two studies that may help to explain the results.
Table 1
Results of recent protocolized weaning trials
Ely3 Kollef4
Protocol Control Protocol Control
Number of Pts 149 151 179 178
Weaning Time* 1 (0-2) 3 (2-7) 35 (15-114) 44 (21-209)
Ventilator Days* 4.5 (2-9) 6 (3-11) 69 ± 123 102 ± 169
Hospital Days 14 (9-26) 15.5 (6-30) 12.7 ± 9.4 14.2 ± 11.7
Reintubation Rate 4% 10% 12.8% 10.1%
Complication Rate 41% 20%
Mortality 40% 40% 40% 42%
*Ely expressed time in days, Kollef expressed time in hours.
In the Brochard study, patients were required to tolerate a PSV level of 8 cm H2O for 24 hours with a respiratory rate (RR) less than 30/min before being considered weaned, while Esteban et al required tolerance of a PSV level of 5 cm H2O for two hours with a RR less than 25/min. In the SIMV arm, Brochard required tolerance of a rate of four mandatory breaths/min for 24 hours (with RR < 30/min), while Esteban required tolerance of a rate of 5/min for two hours (RR < 30/min) before considering the patient weaned. In the T-piece arms, both studies required successful completion of a two-hour trial with the RR less than 30/min. However, in the Brochard study, two patients were required to successfully complete two T-piece trials and one patient was required to successfully complete three T-piece trials before they were considered weaned.
The one overriding conclusion of both groups was that SIMV as applied in these studies was the least successful method of weaning. The more demanding criteria of the Brochard trial across all arms may have artificially lengthened weaning time. When days of weaning expressed in mean ± SD are compared across both trials, the shortest weaning time was within the T-piece group in the Esteban trial (3.4 vs 3.8 ± 0.5 days; Tobin MJ, personal communication). The PSV group of the Brochard trial required 5.7 ± 0.7 days. My personal interpretation is that for the vast majority of patients who require mechanical ventilation, weaning is not the issue but, rather, identifying readiness for spontaneous breathing. The most rapid method of identifying readiness for spontaneous breathing is a T-piece trial, which is what both Ely and associates3 and Kollef et al4 tested in their comparisons of protocolized weaning to standard physician orders.
Table 2
Criteria used to screen patients on a daily basis for readiness for a weaning trial
• Intact airway reflexes
• PaO2/FIO2 > 200 mmHg
• PEEP £ 5 cm H2O
• RR/VT £ 105
• No need for continuous infusion of vasopressors or sedatives
Patients were required to pass all of the other criteria before movement to
Protocolized Weaning Trials
Recently, two groups in single-center studies have randomly compared the use of therapist/nurse-driven protocols to individual physician ordering practices with respect to the rapidity of weaning from ventilatory support (See Table 1).3,4 In the protocol arms of these studies, respiratory therapists screened patients each morning using the criteria outlined in Table 2. It was assumed that the process requiring ventilatory support had been reversed prior to screening. Those patients passing the screen received up to a two-hour T-piece trial after consultation with the medical staff. Patients successfully completing the trial were considered weaned and were extubated unless there was an overriding medical concern (See Table 3). In the control arm, normal physician ordering practices were followed.
As noted in Table 1, approximately 150 patients were randomized to each arm of both studies. In both studies, weaning time was shorter, as were total ventilator and ICU days, in the protocol group. In the Ely study, reintubation rates were also lower in the protocol group, as were overall complication rates. In the Kollef study, complications were similar in both groups. The results of both of these studies favored the use of therapist/nurse-directed protocols over individual physician ordering practices.
Table 3
Criteria for failure of a spontaneous breathing trial
• Pulse > 140/min or sustained 20% increase
• SaO2 < 90%
• RR > 35/min
• Systolic BP > 180 mmHg
• Diastolic BP < 90 mmHg
• Increased anxiety
• Increased diaphoresis
Failure of any of the above dictated failure of the spontaneous breathing
Why Do Protocols Work?
Protocols resulted in more rapid weaning not because respiratory therapists and nurses were more capable of weaning patients than physicians but because the approach was well defined and based on established data, application was consistent, and movement through the protocol was directed and constantly assessed by practitioners at the bedside. There was no room for variation in the application of the protocol based on individual therapist/nurse bias, and there was no delay in decision making because therapists and nurses were constantly at the bedside.
In the individual physician ordering group, numerous approaches to weaning were possible and, potentially the approach used could change from day to day depending on the bias of the physician managing the patient on a given day. Delays in weaning are common when individual physician orders are required to progress from one phase of weaning to the next. If the ordering physician only sees the patient in the morning and then again in the afternoon, the weaning process can be markedly delayed by the lack of orders to proceed.
In order for protocolized approaches to weaning to be successful, there must be physician champions directing the overall process. Physician direction is essential in the development of the protocol, for the buy-in of the protocol by other physicians, and to assure that the protocol is followed by staff at the bedside. This process, although operationalized at the bedside by therapists and nurses, requires physician direction and intervention at defined stages if it is to be employed successfully.
Performance of a Spontaneous Breathing Trial
Where controversy still exists is how best to perform a spontaneous breathing trial. Do we use a T-piece? Do we leave the patient attached to the ventilator? Do we apply continuous positive airway pressure (CPAP), or do we apply low levels of pressure support? A recent randomized, multicenter study by Esteban5 indicated that the use of 7 cm H2O PSV during the two-hour spontaneous breathing trial identifies a greater percentage of patients ready for extubation than a simple T-piece trial. In this study, patients were randomized into T-piece and PSV groups after meeting criteria for weaning (See Table 2). In the T-piece group, 22% failed the trial, while only 14% failed the PSV trial. The reintubation rates for both groups were the same.
It is unclear how best to integrate this single study into a weaning protocol. My bias for spontaneous breathing trials is to leave the patient attached to the ventilator provided the ventilator is of the most recent generation and flow triggering is available.6 This insures that all of the monitors and alarms on the ventilator will be active during the trial. Of concern is the use of older ventilators, with which imposed work of breathing during spontaneous breathing can be excessive. The addition of 5 cm H2O CPAP during the trial is indicated in patient with COPD and auto-PEEP,7 whereas the addition of PSV (7 cm H2O) during the trial may logically be considered for patients with small endotracheal tubes or overall increased airways resistance.8 Of critical importance when designing a spontaneous breathing trial protocol is to insure that objective guidelines are established for each aspect of the protocol so that similar judgments will be made by the different practitioners in carrying out the protocol.
References
1. Brochard L, et al. Am J Respir Crit Care Med 1994; 150:896-903.
2. Esteban A, et al. N Engl J Med 1995;332:345-350.
3. Ely EW, et al. N Engl J Med 1996;335:1864-1869.
4. Kollef MH, et al. Crit Care Med 1997;25:567-574.
5. Esteban A, et al. Am J Respir Crit Care Med 1997;156: 459-465.
6. Hirsch C, et al. Resp Care 1991;36:815-828.
7. Petrof BJ, et al. Am Rev Respir Dis 1990;141:281-289.
8. Brochard L, et al. Anesthesiology 1991;75:739-745.
Criteria for termination of a spontaneous breathing trial inc1ude:
a. pulse rate exceeding 140 beats/min.
b. systolic blood pressure exceeding 180 mmHg.
c. respiratory rate exceeding 35 breaths/min.
d. arterial saturation by pulse oximetry less than 90%.
e. all of the above.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.