Vol. 11 •Issue 9 • Page 16
Ventilation Today
Protocol-directed Weaning Improves Outcomes and Lessens Costs in Pediatrics
An administrative review of adult and pediatric mechanically ventilated patients at the Medical University of South Carolina Hospital revealed that this high risk, high cost patient group in one year had accounted for $80 million in charge activity and utilized 23,000 hospital days. Something had to be done.
The pediatric intensive care unit formed a multidisciplinary process improvement team with the goal of identifying and implementing the best practices in weaning pediatric patients and reducing the mechanical ventilation hours per patient. “Trench workers” on this committee provided bedside observations and reality testing of any new plans for change. They included the head respiratory therapist of the PICU, two experienced staff nurses, the unit’s doctor of pharmacy, and the .critical care nursing education coordinator.
Clearly economic issues initially determined the team’s focus. Hypothetically, more expeditious weaning would decrease the length of stay and treatment costs. Published data have suggested caring for a ventilated patient in the ICU is two-to-four times more costly than a nonventilated ICU patient and eight-to-10 times the cost of a general ward patient.1
However, in addition to improving the institution’s fiscal outcome, they quickly recognized substantial clinical and functional outcomes would be realized with more efficacious care of these complex patients. In fact, the team became zealously convinced that safely liberating the patient from mechanical ventilation is an important goal for no other reason than moving the patient out of potential harm’s way.
Reduced mechanical ventilator time could decrease co-morbidities, such as ventilator-associated pneumonia, atelectasis, airway complications, barotrauma, or at times life-threatening endotracheal tube obstruction or accidental extubation. Patient functional and satisfaction goals also would be achieved because less patient sedation, long-term morbidity and discomfort would occur.
ANALYZING THE PROCESS
The first major step after identifying the goals was to analyze the facility’s current approach to mechanical ventilation weaning and formulate a plan to reach better outcomes. The team used personal experience, relevant medical literature and baseline institutional data to critically look at the system in place. They ultimately made a number of conclusions:
• Most physicians don’t wean mechanical ventilation expeditiously. The personal observations of the respiratory and nursing staff were poignant. For example, one of the three intensivists assigned to the 10-bed PICU was known as the “clean-up attending.” This intensivist had a reputation for aggressively weaning and extubating children — and decreasing the acuity and census — every time he came on service. On the other extreme, resident physicians were observed to be very reticent in making any significant changes in the ventilator settings, especially during night call.
The medical literature supports the observation that physicians don’t keep pace with patients’ improvement. A randomized, controlled study demonstrated this point by allowing a group of mechanically ventilated adult patients who met certain respiratory criteria to undergo a two-hour trial of spontaneous breathing (T-bar or a “flow-triggered” circuit with continuous positive airway pressure).2 The attending physicians were then informed of the patients who had successfully tolerated the trial. This group of patients had a significantly shorter mechanical ventilator time than a control group who weren’t subjected to the trial but managed by the same group of attending physicians.
• Most physicians and bedside ICU nurses don’t effectively use the immediate biomechanical feedback available at the bedside. The new breed of sophisticated ventilators that were developed in the 1990s .gave accurate and valuable expiratory tidal volumes and minute ventilation of small pediatric patients that might allow safe weaning without a multitude of blood gases. Pulse oximetry allows safe and rapid weaning of FIO2, which is a potential signal of reduced ventilator-perfusion mismatch. Observing indirect measures of dynamic compliance (peak inspiratory pressure) and wheezing (auto-peep) could signal a patient’s readiness for aggressive weaning.
• The medical literature showed the efficiency of weaning protocols, although there were no studies in the pediatric literature at that time.3 The adult literature suggested that a timely and systematic approach to weaning patients may be more important than the actual technical method utilized and that RT/nurse driven protocols could reduce average mechanical ventilation hours by approximately one-third.4
• Patient characteristics have a major effect on the duration of mechanical ventilation. Clearly an acute respiratory distress syndrome patient’s ventilatory needs are much different than a child with croup. It was too complex to devise a different care plan for each type of mechanically ventilated patient. (Our attendings already had adopted permissive hypercapneic strategies for our ARDS patients in the early 1990s, and thus volutrauma issues weren’t considered major factors.)
Instead, the team decided to focus on a weaning protocol that would be directed by the patients’ physiologic readiness to tolerate reduction .in ventilatory support. However, the group recommended an aggressive enteral feeding protocol and a neuromuscular/sedation protocol should be devised for the longer term, more complex mechanical ventilation patients.
MAKING IMPROVEMENTS
The team developed a protocol that reduces the unnecessary variation and delay in the mechanically ventilation weaning process. The protocol is RT/nurse driven with attending input at key stages. The momentum for weaning shifts to previously established patient respiratory milestones — not a resident or attending’s bedside assessment or repeated blood gases.
The entry criteria for the protocol (See Table 1.) were purposely conservative. FIO2 is rapidly weaned based on pulse oximetry to a lower limit of 0.30. The rational for this lower limit is that ventilation-perfusion matching is never normal on positive pressure ventilation, and tolerance of a FIO2 of 0.30 has been associated with low risk of extubation failure.5 A prospective study of 208 mechanically ventilated pediatric patients concluded that several parameters predicted a low risk (< 10 percent chance) of extubation failure. Some of these criteria included spontaneous tidal volume of ³ 6.5 ml/kg and PIP ² 25 cm H2O.
By protocol, the synchronized intermittent mandatory ventilation is weaned to maintain minute ventilation within 25 percent of the initial satisfactory minute ventilation established prior to protocol initiation. The rate is weaned to a minimum of 8 for patients younger than 5 years old and rate of 4 for patients older than 5 years old.
Mechanical ventilation providing < 20 percent of the total minute ventilation is consistent with a low risk of extubation failure.5 The team surmised that if the children appeared comfortable on these minimum rates (which represent approximately 20 percent to 30 percent of their minute ventilation requirements), they would have a good likelihood of extubation success. Additionally, the group empirically believed that low positive pressure ventilatory rates helped prevent atelectesis in young children.
The protocol then advances to bedside weaning parameters in cooperative patients or in younger patients with at least a spontaneous tidal volume, minute ventilation and respiratory rate. The majority of patients undergo a CPAP/pressure support trial, or in some older patients, a T-bar trial for 30 minutes to 60 minutes with a blood gas and reassessment of the bedside parameters. CPAP/pressure support is generally reduced to a minimum 2 cm H2O and 5 cm H2O, respectively, for the potential advantage of reducing the work of breathing associated with endotracheal tube/ventilator circuit resistance.
The attending is notified at that point in time for approval of extubation. In occasional patients with risk of fatigue (patients with neuromuscular disease or a prolonged ICU course), a preliminary blood gas is obtained prior to any trials, which are then individualized (most often in the form of repetitive trials for longer lengths of time) depending on the patient’s tolerance.
RESULTS AND DISCUSSION
The facility’s outcomes manager checked the effectiveness of these interventions by analyzing retrospective clinical and financial databases. Total mechanical ventilation hours, mechanical ventilation hours per patient and per PICU day, PICU and hospital length of stay, and multiple other parameters, including number of blood gases, cost of sedation, patient demographics and mortality, were tabulated for periods before and after protocol initiation.
Notably, the average mechanical ventilator hours decreased by nearly 50 percent in years 1997 and 1998 after the protocol had gone into effect. The standard deviation decreased concomitantly, documenting less variation in mechanical ventilation hours and suggesting less variation in the ventilator usage. (See Table 2) The number of blood gases decreased by 48 percent during the 1996 and 1997 time frame.
In an attempt to better define the protocol’s impact, a subset of patients who underwent mechanical ventilation in the PICU several months before the formation on the team was compared with a subset of patients who were ventilated several months after the initiation of the protocol. Importantly, in these patient subset groups, any patient who had expired or was determined to be a home ventilator patient was excluded. (This isn’t true of the data in Table 2.)
The results of this patient selection manipulation yielded more modest results; still, there was a significant decrease in ICU length of stay per patient and a notable drop in the average mechanical ventilation hours. (See Table 3.)
Very little benchmark data in the medical literature exists regarding mechanical ventilation length of pediatric patient groups. A 1997 study provided descriptive data on a random sample of 198 mechanically ventilated pediatric patients in a tertiary PICU and noted an average mechanical ventilation time of 8.3 days (199 hours).6 The mean mechanical ventilation values in our patient population, both before (148 hours) and especially after the protocol (89 hours) are less, but there’s insufficient data to allow statistical comparison between the two institutions.
Comparing the median or the 50th percentile mechanical ventilator hours per patient helps to offset the effect of “outliners” — the handful of patients who are on the ventilator for literally thousands of hours before extubation or weaning; these patients can significantly affect average values. A Kaplan Meier plot of the patient data from Table 3 demonstrates a progressive difference in the two groups after approximately 40 percent of the patients had been weaned off the ventilator. (See Figure 1.)
It’s tempting to speculate that the weaning protocol, and perhaps aggressive enteral nutrition and neuromuscular protocols, played a part in the post-protocol group that required prolonged ventilation. Lastly, a control chart demonstrates that over the past couple of years we have maintained a statistically significant trend of decreased mechanical ventilation hours per ventilated patient days. Further statistical analysis of some of the outcome data can be found elsewhere.7
FURTHER GAINS
It will be important for the team to reassess the protocol to maintain outcomes and to achieve further gains. Several current practices in the PICU need to be incorporated into the protocol. For example, the respiratory rate entry criteria needs to be less conservative as the upper limits aren’t strictly followed if the patient has little in the way of retractions or work of breathing.
Another suggested arm of the protocol may address the abrupt spontaneous breathing trial that has been shown to be an effective tool in adults screened for basic pulmonary criteria.8 Already some of the attendings routinely perform CPAP/pressure support trials during morning rounds in patients with mild pulmonary dysfunction to better assess their pulmonary status. Similarly, it’s not an uncommon practice in the unit to quickly perform such trials on awakening patients who were ventilated for status epilepticus but have little in the way of pulmonary disease. End-tidal CO2 monitoring has yet to be incorporated into the protocols, but this technology may offer further bedside support for rapid weaning.
The process improvement efforts and protocol led to a demonstrable effect on patient outcomes. This project demonstrated that RTs and nurses could successfully wean pediatric patients more expeditiously with a protocol than was accomplished in preprotocol periods when physicians were exclusively responsible for mechanical ventilation weaning.
Another benefit of developing a protocol-based therapy has been an enhancement of the multidisciplinary clinical effort of the PICU. Initially, some physicians expressed concern that there would be loss of physician autonomy or an effect on medical education with a shift to a clinical pathway.
The opposite occurred. The bedside staff and resident physicians have a better understanding of weaning parameters and ventilator function by understanding of the rationale and utilizing the protocol. Lastly, it should be emphasized that process improvement efforts must emanate from the personnel and units who are going to perform the process and will require administrative support and commitment to achieve.
All of the authors work at the Medical University of South Carolina Children’s Hospital. Dr. Tecklenburg serves as the director of the division of pediatric emergency/critical care. Dr. Habib is the medical director of the PICU and pediatric respiratory care. Hughes serves as project manager in the department of outcomes management research and development. Dr. Smith is an associate professor of surgery and pediatrics.
Visit “Online Extras” at www.Respiratory-care-sleep-medicine.advanceweb.com to see the Medical University of South Carolina’s PICU Ventilation Weaning protocol.
For a list of references, please call Mike Bederka at (610) 278-1400, ext. 1128, or visit www.Respiratory-care-sleep-medicine.advanceweb.com/mrreflist.html.
