Reevaluating a Standardized Sedation Weaning Protocol for Pediatric Laryngotracheal Reconstruction for Continuous Quality Improvement

Key Points

Question  How does a new electronic health record (EHR) effect a series of process, outcome, and balance measures for sedation weaning management in pediatric patients undergoing laryngotracheal reconstruction?

Findings  This quality improvement study included a focus group review of 13 patients in a postweaning group and 11 patients in a post-EHR group. Despite there being a drop in the presence of our standardized wean document (process measure) following implementation of a new EHR, length of weaning (outcome measure) remained stable at 9.45 days in the post-EHR group compared with 9.08 days in the postweaning group.

Meaning  By assessing a broad range of measures, we were able to understand the impact of the new EHR on our sedation weaning practice and target improvement opportunities, such as the development of a flexible sedation weaning template, more appropriately.

Importance  Health care organizations are complex and evolving systems. To date, longitudinal evaluation to ensure the sustainability of quality improvement (QI) initiatives has been missing from the otolaryngology literature. We sought to reassess perioperative management of laryngotracheal reconstruction, which requires adequate sedation.

Objective  Using principles of continuous QI, the objectives of this study were to (1) describe step-by-step methods to sustain QI efforts and (2) revisit a series of process, outcome, and balance measures for sedation weaning management following implementation of a new electronic health record (EHR).

Design, Setting, and Participants  A standardized sedation weaning protocol was previously developed and instituted in February 2013. To address healthcare system-wide changes, a 7-step, Institute for Healthcare Improvement methodology was used to reevaluate a series of measures comparing a previous postweaning group (2013-2014; 13 patients) and current post-EHR group (2016; 11 patients). We conducted a focus group review of these 24 patients.

Main Outcomes and Measures  The primary outcome measure was length of sedation weaning. Secondary outcome, process, and balance measures included total length of sedation, absence of standardized wean document, absence of specific recommendations on weaning regimen, length of stay, continued weaning at discharge, discharge location, absence of discharge instructions on weaning regimen or iatrogenic withdrawal syndrome (IWS), discharge within 72 hours of stopping weaning, and readmission.

Results  The postweaning and post-EHR groups were similar in age (20.5 months [95% CI, 11.92-29.15] vs 26.5 months [95% CI, 17.68-35.40]), as well as male sex (11 of 13 [85%] vs 10 of 11 [91%]), respectively. In the post-EHR group, the standardized sedation wean document was missing from 9 of 11 (82%) medical records. However, the primary outcome measure, length of sedation weaning, remained stable at 9.45 (95% CI, 7.62-11.29) days in the post-EHR group compared with 9.08 (95% CI, 7.00-11.18) days in the postweaning group. In addition, only 5 of 11 (46%) of discharges in the post-EHR group had specific guidance on weaning since the standardized template was no longer in use. As a result, in the post-EHR group, patients were 15.2 (95% CI, 0.46-242.34) times as likely to lack discharge instructions on weaning or IWS.

Conclusions and Relevance  Quality improvement is meant to be a continuous process in which reevaluation of care practices are regularly performed. System-wide redesign can be achieved using a formal methodological approach. Moving forward, notable QI opportunities for our institution included the development of a flexible sedation weaning template, as well as enhancements to discharge instructions to include IWS diagnosis and treatment.

In the pediatric population, laryngotracheal stenosis continues to remain a considerable health issue.^1-3 Laryngotracheal reconstruction (LTR) has evolved to include a variety of techniques for expanding the stenotic airway, including reconstruction with rib cartilage grafting.^4,5 With widespread use of open surgical techniques, decannulation rates surpass 90%.^3,6 To achieve a successful outcome, the perioperative treatment of these patients requires coordinated, multidisciplinary care.^7-11

Following LTR, the patient is often nasotracheally intubated, requiring analgesia and sedation, with or without neuromuscular blockade. This level of pharmacologic restraint helps minimize excessive neck and endotracheal tube movement, which could disrupt suture lines and cartilage grafts, cause mucosal trauma, or result in accidental extubation.⁷ The length of intubation, and subsequently sedation and analgesic needs, varies from 3 to 9 days, depending on the type of airway reconstruction.⁵ Unfortunately, tolerance and physical dependency to these medications develops in up to 50% to 60% of patients.¹² As a result, medication tapering is necessary to decrease the risk of withdrawal, which could complicate the process of tracheal extubation and recovery. Iatrogenic withdrawal syndrome (IWS) is characterized by autonomic dysfunction, gastrointestinal disturbances, and neurologic and motor abnormalities, which can manifest after abrupt cessation or rapid weaning in physically dependent patients. Recommendations for optimal sedation and analgesia in critically ill children, as well as assessment methods for IWS, continue to evolve as new research emerges.^13-17

Although weaning protocols are associated with decreased rates of withdrawal, ventilator duration, intensive care unit and hospital length of stay, and cost, lack of adherence is common.¹⁸ One reason for poor adherence involves the complexity of care processes.¹⁸ At our institution, we previously used quality improvement methodology to improve adherence to a sedation weaning protocol and found a significant decrease in the length of sedation weaning following LTR, a critical aspect of postoperative care.¹⁹ Although achieving improvement is ideal, leading healthcare organizations also recognize the importance of sustaining improvement efforts. Quality improvement is meant to be a continuous process in which reevaluation of care practices are regularly performed. Thus, using principles of continuous quality improvement, our objectives were to (1) describe step-by-step methods to sustain improvement efforts and (2) revisit a series of process, outcome, and balance measures for sedation weaning management following implementation of a new electronic health record (EHR).

The study took place at the Massachusetts Eye and Ear Infirmary (MEEI) and Massachusetts General Hospital for Children (MGHfC) Boston. The Pediatric Airway, Swallowing, and Voice Center is a unique collaboration between the MEEI and MGHfC. Airway reconstructive surgeries are performed within the MEEI operating rooms and patients are subsequently transferred for postoperative care to the MGHfC PICU. Once patients no longer require intensive care, they are transferred to the floor at MGHfC or MEEI, depending on individual patient complexity and needs. The institutional review board of the Massachusetts Eye and Ear Infirmary deemed the study exempt from review, because it was part of a quality improvement initiative.

The Institute for Healthcare Improvement (IHI) is a recognized health care quality improvement organization that provides resources for implementing system-wide change. In particular, the IHI considers the execution of effective communication and coordination at the interfaces between care settings to be a high-leverage initiative leading to a significant improvement in outcomes. As a result, they have developed multiple guides that provide a step-wise approach to improve transitions in care across a variety of settings (Figure 1).^20,21

For this project, the "forming a team" step involved 2 key groups of stakeholders. First, the previous study’s focus group included attending pediatric otolaryngologists, pediatric intensivists, pediatric hospitalists, fellow physicians, resident physicians, nurses, pharmacists, and social workers.¹⁹ All individuals who contributed to the original work and were still functioning in a similar role were invited to join the focus group. In addition, any new clinicians were also invited to participate. The impetus to reconvene the focus group was centered on a postoperative readmission of a patient who underwent LTR to the hospital for sedation withdrawal. As a result, the group reviewed the experience of patients who underwent LTR from 2015 to 2016. One key issue stood out during this time, which was the implementation of a new EHR in April 2016. In the prior electronic health record, the standardized sedation weaning protocol was uploaded as a stand-alone document at the time of patient transfer, which was readily accessible to clinicians in all locations. This method of communication was disrupted with the change in EHR systems. As a result, the revived focus group recommended reevaluation of all patients who underwent LTR following implementation of the new EHR in April 2016.

With the focus on continuous quality improvement, we kept consistency with measurement between the original and current study.¹⁹ The 2 groups being compared were those patients after the implementation of the original standardized sedation weaning document (2013-2014; 13 patients; data available from the prior study¹⁹) and the new set of patients following implementation of the new EHR (2016; 11 patients). The primary outcome measure that was assessed was the length of sedation weaning. To simplify the presentation of data and discussion, these will be refered to as the “postweaning group” and “post-EHR group,” respectively. Secondary measures included the total length of sedation (outcome), absence of standardized sedation wean document at time of transfer (process), absence of specific recommendations on weaning regimen (process), length of hospital stay (balance), length of intensive care unit (ICU) stay (balance), length of floor stay (balance), continued weaning at time of discharge (balance), and discharge location (balance). Since reevaluation of the LTR sedation weaning protocol was prompted by a readmission, 3 additional measures were examined, including absence of discharge instructions on weaning regimen or information on IWS (process), discharge within 72 hours of stopping weaning (balance), and readmission (balance).

Descriptive statistics are provided as frequency count and percentage for categorical data and mean and 95% CI for continuous data. Baseline demographic and clinical information was compared using the Fisher exact test for categorical data and t test for continuous data. Results were considered statistically significant at P < .05. When comparing the process, outcome, and balance measures, effect size was determined using relative risk for categorical data and Cohen d for continuous data. In addition, a statistical process control chart of sedation weaning length was constructed with an individual moving range (Xmr) chart using the 3-σ rule for defining upper and lower control limits. Statistical analysis was performed in JMP statistical software (version 13.1, SAS Institute).

Baseline demographic and clinical information of the postweaning and post-EHR patients were reasonably similar. There were no statistical differences between age 20.5 (95% CI, 11.92-29.15) months vs 26.5 (95% CI, 17.68-35.40) months; P = .40), male sex (11 of 13 [85%] vs 10 of 11 [91%]; P > .99), patients with rib cartilage graft (9 of 13 [69%] vs 8 of 11 [73%]; P = .70), and continuous sedation infusion duration (9.15 days; 95% CI, 7.51-10.78 vs 8.64 days; 95% CI, 8.10-9.18; P = .60). A shorter length of mechanical ventilation was identified in the post-EHR group compared with the postweaning group (10.46 days; 95% CI, 8.63-12.29 vs 8.09 days; 95% CI, 7.48-8.71, respectively; P = .03).

In the previous EHR system, a sedation weaning template, “MGH/MEEI Sedation Wean Plan” (Figure 2), was loaded into the computer and easily accessible. In the new EHR, the document was no longer uploaded with the same consistency, such that it was 10.6 (95% CI, 1.59-71.37) times as likely to be missing from the record compared to the previous EHR. Although there was a notable decrease in the use of the specific template, the medical record was also assessed for any other representative recommendations regarding sedation weaning. Every PICU-to-floor transfer note mentioned the presence of ongoing sedation weaning and gave instructions, although the consistency and clarity of the medication plan varied substantially. Furthermore, in 7 of 11 (60%) cases, tapering of sedation medications varied from the standardized protocol either by dosing strength, dosing frequency, or both.

Table 1 and Table 2 summarize additional measures between the postweaning group and the post-EHR group. For the primary outcome measure, the length of sedation wean remained stable at 9.45 days (95% CI, 7.62-11.29) in the post-EHR group compared with 9.08 days (95% CI, 7.00-11.18) in the postweaning group, suggesting a trivial difference (Cohen d, 0.12; 95% CI, −0.68 to 0.92) between the 2 groups (Table 2) (Figure 3).²² In respect to length of sedation, there was a moderate difference (Cohen d, −0.44; 95% CI, −1.25 to 0.34), with the post-EHR group having an overall shorter course. When examining the balance measures, a trivial difference (Cohen d, −0.17; 95% CI, −0.97 to 0.63), was noted in the total length of stay (Table 2). There was, however, a large difference noted (Cohen d, −1.21; 95% CI, −2.09 to 0.35), with a shorter PICU length of stay and moderate difference noted (Cohen d, 0.79; 95% CI, −0.30 to 1.63), with a longer floor length of stay in the post-EHR group (Table 2).

The 1 readmission in the post-EHR group represented 9.1% of patients, such that patients in this group were 3.5 (95% CI, 0.15-78.19) times as likely to be readmitted (Table 1). Patients in the post-EHR group were also 1.18 (95% CI, 0.60-2.32) times as likely to be discharged within 72 hours of stopping their sedation weaning (Table 1). In addition, patients in the post-EHR group were 1.18 (95% CI, 0.46-3.04) times as likely to complete a sedation weaning plan at home (Table 1). Discharge documents were assessed for instructions on weaning regimen or information on IWS. In the postweaning group, every discharge summary included information on the sedation weaning protocol as the MGH/MEEI Sedation Wean Plan was included in the home-going materials. In contrast, only 5 of 11 [45.46%] discharges in the post-EHR group had specific guidance on weaning because the template was no longer in use. As a result, in the post-EHR group, patients were 15.17 (95% CI, 0.46-242.34) times as likely to lack discharge instructions on weaning or IWS.

All too often, successful improvement efforts are lost as attention shifts to other priorities while systems and staff revert to the “old way” of doing things. Previously at our institution, there was substantial practice variability in terms of sedation management following LTR procedures. A key focus point for improvement centered on the development of a standardized sedation weaning document. Institute for Healthcare Improvement methodology was used to implement system-wide practice changes, resulting in a significant decrease in the length of sedation weaning postoperatively.^19-21 When engaging in quality improvement, it is crucial to ensure that processes continue to perform at the newly established levels of quality and safety. Thus, the readmission of one of our LTR patients for sedation withdrawal renewed our focus on the sedation weaning process. And, as noted, 1 key issue stood out, which was the implementation of a new EHR in April 2016. As a result, based on the principles of continuous quality improvement, we reevaluated a series of process, outcome, and balance measures for our sedation weaning practices. This longitudinal perspective is necessary to ensure sustainability of improvement efforts and has been missing from the otolaryngology literature.²³

The 2 groups examined, postweaning and post-EHR, were well matched in respect to age, sex, and need for a rib graft, which can serve as a proxy for extent of surgery, postoperative pain, and analgesic needs. It is also important to assess each group for differences in length of mechanical ventilation and continuous sedation. In particular, higher doses and longer duration of sedation has a substantial impact on the risk for IWS and thus consideration for a slower taper.¹⁵ In the post-EHR group, both a shorter length of mechanical ventilation and length of time for infusion of sedation medications was identified. Thus, the 2 cohorts had relatively similar baseline demographic and clinical characteristics. Furthermore, the distinction regarding mechanical ventilation and length of sedation would have favored a reduced risk of IWS in the post-EHR cohort.

Overall, the primary outcome measure, the length of sedation weaning, showed stability over time. In contrast, the standardized sedation wean document was missing from 9 of 11 (82%) medical records in the post-EHR group. This finding illustrates the importance of assessing multiple measures when performing quality improvement work. Although the process measure, the sedation wean document, appeared to fail in the new EHR system, the outcome measure remained stable. Then, when the medical record was assessed in further detail, sedation weaning schedules were identified, but just in different formats when compared with the original form. Currently, the optimal pharmacologic treatment in the postoperative setting following complex airway surgery remains challenging. The original sedation wean document was developed by the MGHfC PICU Withdrawal Committee based on published literature.^12,13 There are, however, multiple other care organizations that have developed similar evidence-based protocols, but with slightly different recommendations that appear equally effective and safe in practice.^15,24,25 In 7 of 11 (63.6%) of our post-EHR records, tapering of sedation medications varied from the previously developed, standardized protocol either by dosing strength, dosing frequency, or both. These subtle differences, however, did not appear to affect the primary outcome measure of sedation weaning length. Thus, it appears that although the wean document itself was not present, there was retention of a more formalized hand-off process describing the weaning plan, which is perhaps more important than rigid compliance with any specific weaning protocol.

In this study, we also focused more specifically on the discharge process. Evidence suggests that several specific interventions can reduce the rate of avoidable rehospitalizations, including medication reconciliation and customized patient education.²¹ In the post-EHR group, patients who were being discharged on a sedation weaning protocol to be completed at home were provided with specific guidance on administration of medications. Notably, however, 6 of 11 (54%) of patients in the post-EHR group did not receive any information on weaning or IWS. As a result, patients in the post-EHR group were 15.17 (95% CI, 0.46-242.34) times as likely to lack discharge instructions on weaning or IWS.This is particularly pertinent because withdrawal symptoms usually appear within 12 hours of stopping sedatives, but may not reach their peak until 72 hours.^15-17 Perhaps more concerning is that the number of patients discharged within 72 hours of stopping their sedation weaning was high in both groups (7 of 13 [54%] postweaning and 6 of 11 [64%] post-EHR). Thus, it seems that many patients are being discharged within the time window for IWS without provision of adequate education on the condition or recommended actions if concern arises.

Several potential limitations exist in our study. Because of the unique patient population, the sample size was very small, leading to effect sizes with broad 95% CIs, which must be interpreted cautiously. In addition, this is a single-institution evaluation and, owing to the substantial variability in practice environments, our results are not directly generalizable to other institutions. However, the processes with which we assessed our clinical environment using IHI methodology can serve as an example to other organizations. Furthermore, we did not examine the clinical efficacy of our medication and dosage regimen. Thus, we cannot provide any specific endorsements or support for our sedation weaning protocol.

Perioperative treatment for laryngotracheal reconstruction requires adequate sedation and analgesia with or without neuromuscular blockade to achieve success. A standardized sedation weaning protocol was previously developed and instituted in February of 2013.¹⁹ Because of the implementation of a new EHR system, the previous stand-alone weaning document was no longer accessible, resulting in roadblocks to communication and medication ordering. In addition, there was a readmission of a postoperative patient owing to IWS. Using IHI methodology, we reevaluated a series of process, outcome, and balance measures for sedation weaning management, providing an iterative view of quality improvement efforts in a clinical setting, displaying measures over time. Despite a change in the sedation wean document process measure, the length of sedation weaning remained the same. Therefore, it appears that a formalized hand-off process describing the weaning plan is more important than rigid compliance with any specific weaning protocol. In addition, it appears that many patients are being discharged within the time window for IWS without provision of adequate education on the condition or recommended actions if concern arises. As a result, notable improvement opportunities for our institution included the development of a flexible sedation weaning template in the new EHR system, as well as enhancements to the discharge instructions to include IWS diagnosis and treatment. Finally, quality improvement is meant to be a continuous process in which reevaluation of care practices are regularly performed. System-wide redesign can be achieved using a formal methodological approach.

Corresponding Author: Christopher J. Hartnick, MD, MS Epi, Department of Otolaryngology, Massachusetts Eye & Ear Infirmary, 243 Charles St, Boston, MA 02114 (christopher_hartnick@meei.harvard.edu).

Accepted for Publication: December 11, 2018.

Published Online: February 14, 2019. doi:10.1001/jamaoto.2018.4348

Author Contributions: Drs Bowe, Cummings, and Hartnick had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Bowe, Colaianni, Hartnick.

Acquisition, analysis, or interpretation of data: Bowe, Yamasaki, Cummings.

Drafting of the Manuscript: Bowe.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Bowe, Cummings.

Administrative, technical, or material support: Bowe, Yamasaki, Cummings, Hartnick.

Study supervision: Hartnick.

Conflict of Interest Disclosures: None reported.

Disclaimer: The views expressed herein are those of the author and do not reflect the official policy or position of the US Air Force Medical Service, the US Air Force Office of the Surgeon General, the Department of the Air Force, the Department of Defense, or the US Government.

Meeting Presentation: This study was an oral presentation at the Triological Society 2017 Combined Sections Meeting; January 20, 2017; New Orleans, Louisiana.

Additional Contributions: The authors thank James K. Aden, PhD, San Antonio Uniformed Services Health Education Consortium, JBSA-Ft Sam Houston, San Antonio, Texas, for his assistance with statistical analysis. He was not compensated.