Reducing Time to Postintubation Sedation in a Pediatric Emergency Department Free

Monroe Carell Jr Children’s Hospital at Vanderbilt (Monroe Carell) is a 358-bed quaternary care pediatric referral center in Nashville, TN. Monroe Carell currently has more than 16 000 inpatient admissions and approximately 53 000 ED visits annually. The ED team includes PEM-trained physicians, general emergency medicine physicians, PEM fellows, emergency medicine and pediatric residents, general pediatricians, pediatric nurse practitioners, pediatric nurses, paramedics, and respiratory therapists. Critically ill children are admitted to the PICU, where care is directed by a pediatric intensivist. Providers in both inpatient and emergency settings use an electronic medical record system and computerized physician order entry (EPIC, Verona, WI).

We included all children aged younger than 18 years old presenting to Monroe Carell’s ED who underwent endotracheal intubation in the PED. Children who were intubated before PED arrival, suffered intracranial hemorrhage, had hemodynamic instability, or cardiac arrest in the PED were excluded.

An interprofessional team was formed in March 2020 to standardize PIS care. The team included a nurse educator, nurses, paramedics, PEM physicians, and PEM fellows. The team used the Model for Improvement as the foundation of the improvement work and created a key driver diagram to outline theories for improvement and potential interventions (Fig 1).¹¹  Where appropriate, Lean principles were incorporated to inform the team’s understanding of the system and eliminate waste. Some of the Lean principles used were GEMBA walks and visual reminders. Initial interventions comprised testing and modifying a PIS sedation protocol that was used in the PICU.

Over the subsequent 6 months, the protocol was reviewed by multiple oversight committees to ensure compliance with patient safety and nursing scope of practice, ultimately approved in September 2020. Nursing educational modules were created before formal rollout of the protocol as well as PEM physician educational sessions. After achieving our threshold of completed education by all PEM physicians and more than 75% of nursing staff, the protocol was initiated in December 2020. Additional interventions included monthly review of the data with staff, development of a PIS order set, and then the modification of an existing preintubation checklist to include PIS in April 2021.

The primary measure was time to administration of sedation medication after intubation in the PED, defined as the time from endotracheal tube placement to when the first non-RSI sedative medication was given. This metric was tracked via the resuscitation navigator (an electronic medical record tool within EPIC), which the nursing staff uses to document during all procedural intubations. The time of intubation and timing of medications administered were clearly documented in the ED tracking information. Any sedative medication used postintubation counted toward this metric. Sedative medications included were fentanyl, ketamine, propofol, lorazepam, and midazolam.

The secondary measure was unplanned endotracheal tube dislodgement, which was defined as an unexpected event of an endotracheal tube being removed during the patient’s ED encounter or during transfer to next location of care. The balancing measure was PIS-related hypotension, which was defined as a patient requiring pressors after the initiation of a sedation infusion. This information was also documented in the resuscitation navigator and identified if pressors were started after a sedation infusion was documented in the medication administration record.

Data were collected retrospectively for the baseline period of December 2017 to March 2020 via an electronic dashboard for all intubations encounters in the ED and prospectively thereafter through the end of our data collection. Intubations were grouped into consecutive groups of 3. Data were reviewed on a weekly basis. Statistical process control charts were used to study timeliness of PIS and to monitor any temporal relationships between interventions and performance using accepted rules for special cause variation.¹² 

This work was submitted to the institutional review board and approved as a quality improvement/nonresearch determination.

During the study period, there were a total of 291 intubation encounters within the PED that met study criteria for sedation. During the entire study period, 139 patients were excluded for cardiac arrest, hemodynamic instability, intubation before arrival to the PED, and penetrating head injuries. Baseline data included 176 intubations from December 2017 to March 2020, which established a mean time to PIS of 39 minutes. Special cause variation occurred and the mean time to PIS decreased to 21 minutes immediately following institutional protocol approval. Over the next 8 months, the team enhanced understanding through plan-do-study-act (PDSA) cycles, education to the department on implementation of protocol, development of a PIS order set, and updates to an existing preintubation checklist to include PIS plan. The culmination of these events was associated with special cause variation with a reduction in time to PIS to 13 minutes. An S-chart also revealed a reduction in variability with initiation of PIS (Fig 2).

There were 2 unplanned endotracheal tube dislodgements in the baseline period and none during the intervention period.

We observed no episodes of PIS-related hypotension during the intervention period.

At the beginning of the study, the interprofessional team collaborated to create a PIS protocol. At the time, the PICU had already successfully developed a protocol for PIS that used fentanyl for analgesia and sedation with intermittent doses of fentanyl and midazolam for uncontrolled sedation. If sedation remained uncontrolled, the infusion rate of fentanyl would increase until it reached a threshold. At that point, the addition of a midazolam infusion would be added. The level of sedation was measured using the Richmond Assessment Sedation Score (RASS).¹³ 

The first PDSA cycle involved applying this protocol to 1 patient in our PED. We quickly appreciated that the layout of this protocol did not translate well to our work environment. Our next PDSA cycle involved creating a decision tree to help simplify the directions for those using the protocol. It was helpful but determining which medication to give for uncontrolled sedation remained confusing. Given this feedback, we simplified the protocol by only using intermittent fentanyl injections while titrating the fentanyl infusion for uncontrolled sedation (Fig 3).

Our next PDSA cycle involved creating physician and nursing education. Nursing education was in the form of modules, with a required goal of module completion and a minimum 80% postmodule test score by more than 75% of all nursing staff. The nurses and paramedics also received in-person education from designated staff involved in the design of the protocol. Case studies were used to apply the new knowledge learned in the module, guide them through the PIS protocol, and practice RASS scoring. This extra education offered support to identify gaps in knowledge, the opportunity to ask questions, and gain confidence in using the PIS protocol.

We acknowledged during the protocol development that protocol-driven titration needed to be based on an objective score. Historically, nurses needed to seek out a physician for approval for each medication administration or modification. Given this need, the protocol underwent rigorous evaluation to ensure the language and instruction maintained the appropriate scope of practice and to identify necessary moments when physicians should be notified of a change.

In late November 2020, small trials were performed to ensure the protocol and the accompanying order set were functional. Ultimately, multiple small tests occurred without any interference prompting the protocol and order set to go live in December 2020. Printed protocols were available in our resuscitation bays as were visual reminders regarding the protocol.

During the small trials, we appreciated that children with delayed initiation of sedation were typically secondary to late entry of the order set by the physician. The order set was modified to provide automatic ordering of the sedation fentanyl infusion rate and initial bolus doses based on the patient’s weight, which also occurred in December 2020 (Fig 4). The input required from providers included mean arterial pressure goals and heart rate goals. This workflow modification allowed for ease of ordering in high-stress situations.

Our last PDSA cycle sought to improve team awareness by including PIS plans in our preintubation checklist in April 2021. This checklist is part of the PED standard workflow and is performed before every intubation. This checklist is printed and hangs on the wall of our resuscitation bays, where most patients are intubated in the PED. Using Lean principles, our team initiated a visual reminder through updating this checklist. On April 1, 2021, “Postintubation Sedation Plan” was added to the checklist (Fig 5). Now, during every preparation for an intubation procedure, the PIS plan is discussed. The physician verbalizes if the patient meets criteria for the PIS Protocol and who will order the medications. As the providers started implementing this part of the checklist, some started having the initial bolus of fentanyl drawn up with RSI medications and verbalized when safe to give after the intubation procedure was complete. The mean time to PIS was sustained at 13 minutes from June 2021 until the end of the study period in February 2022.

The use of quality improvement methodology led to a sustained reduction in time to initiation of PIS in a PED from a baseline mean of 39 minutes to 13 minutes without an occurrence of PIS-related hypotension.

Multiple PDSA cycles contributed to our understanding of processes and informed interventions that were associated with a sustained reduction in PIS. In the initial PDSA cycle, applying the protocol to 1 patient allowed us to clearly identify where the protocol needed modifications and what type of education was needed for effective protocol utilization by team members. The nursing leadership involved in the project recognized this quickly and modules were developed to educate the team. Without the baseline education and this iterative process, it would not have been possible to achieve our future goals.

During this process the importance of empowering and using all team members at their maximum scope of practice was demonstrated. Our team learned that nurses in other departments in our hospital were titrating medications for sedation based on RASS scoring, and after collaborating with the PED nursing leadership, we believed that this was within the scope of practice for our nursing staff. Several studies have shown that nursing-driven protocols can help with analgesia in adult and pediatric settings.^14 –16 

As we have continued the protocol and process, we have seen nursing take a leadership role in ensuring patients get adequate sedation after intubation in the department. Several of our bedside nurse leaders were in the initial protocol development and discussions. This is a possible reason we saw a decrease in the time to PIS immediately following protocol approval in September 2020 but before full education of the ED team and implementation of the protocol in December 2020. Nursing and physicians in the initial discussions knew the importance of adequate PIS and began to advocate for our patients within our department. Overall, this allowed for a shared responsibility, which was 1 of our key drivers (Fig 1).

The PIS Order Set and PIS Checklist are examples of higher-level reliability interventions. After establishing the PIS Protocol and once the team felt comfortable using it, we were able to make these additional interventions. Once we implemented these higher-level reliability interventions, we did see a second demonstration of special cause variation and a mean-line reduction in time to PIS from 21 minutes to 13 minutes. Also, as the system continued, we have noted a decrease in variability, demonstrated on the concurrent S-chart (Fig 2).

Our improvement project was developed using Model for Improvement methodology but there are elements of Lean principals that made this project successful. There is crossover between these 2 quality improvement practices and using these 2 methods together can be very beneficial.¹⁷  Our first few PDSA cycles included Gemba Walks. Without doing this practice, we would not have identified the barriers within the protocol or identified the higher-level reliability changes that needed to occur. Visual Reminders is another Lean concept that was implemented. Our PED has several different types of providers and team members that rotate throughout our department, so adding a visual reminder of the PIS Plan to the intubation checklist that hangs in the resuscitation bay was an effective way to remind all team members that sedation is an important part of the intubation process. This project is another quality improvement example where both the Model for Improvement and Lean can be used together for the betterment of patient care.

Our improvement project was performed at a single institution and therefore may not be generalizable to all institutions, which is a limitation of this project. Also, patients who were intubated before arrival to the PED were not included in this project, so timing to sedation for these patients was not accounted for in these data. Patients with intracranial injury, hemodynamic instability, or cardiac arrest were not included in this project, which could result in some limitations to application of the data to these patient populations.

The use of the Model for Improvement and Lean methodology led to a sustained reduction in time to initiation of sedation postintubation in a PED without the occurrence of PIS-related hypotension. The application of PDSA cycles with sequential learning, maximizing all team members scope of practice, and implementation of higher-level reliability interventions allowed for this sustained reduction. Overall, we were able to make an impact on patient safety for some of our most critically ill patients.

ED

emergency department

PSDA

Plan Study Do Act

PED

pediatric emergency department

PEM

pediatric emergency medicine

PIS

postintubation sedation

RASS

Richard Assessment Sedation Score

RSI

rapid sequence intubation