eFigure. Low-Risk Decannulation Pathway for Patients With Tracheotomy.
Smith KA, Matthews TW, Dubé M, Spence G, Dort JC. Changing Practice and Improving Care Using a Low-Risk Tracheotomy Clinical Pathway. JAMA Otolaryngol Head Neck Surg. 2014;140(7):630-634. doi:10.1001/jamaoto.2014.921
Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Tracheotomy is a common procedure. Postoperative care is usually managed by nonexpert clinicians. Prolonged decannulation is associated with a high incidence of complications. At present, no clinical protocol exists to guide clinicians through decannulation. To address this deficiency, we developed a low-risk tracheotomy clinical pathway.
To determine the effect of our low-risk tracheotomy clinical pathway on the time to decannulation and to determine its safety and sustainability by assessing the incidence of adverse events.
Design, Setting, and Participants
Our study combined retrospective and prospective cohorts from July 1, 2008, through January 31, 2012. Low-risk adult patients undergoing tracheotomy at a tertiary care hospital constituted the study population. A baseline cohort of 26 patients underwent retrospective assessment. After development of the pathway, a pilot group of 34 consecutive patients underwent evaluation; of these, 13 were ineligible because of high-risk factors, which included potential upper airway obstruction, unfavorable neck anatomy, or medical factors such as coagulopathy. To assess the sustainability of the pathway, a follow-up cohort underwent assessment. Of 107 consecutive patients, 39 met the low-risk criteria. Length of follow-up was 30 days after decannulation.
The low-risk tracheotomy clinical pathway, which provides a stepwise approach to decannulation.
Main Outcomes and Measures
Total time to decannulation (in days). We hypothesized that the pathway would reduce the total time to decannulation. The secondary outcome constituted adverse events. All hypotheses were formulated before data collection.
Mean (SD) total time to decannulation in the baseline cohort was 15.50 (12.08) days. After implementation of the pathway in the pilot cohort, mean (SD) total time to decannulation decreased to 5.74 (2.79) days (P < .001). In the follow-up cohort, mean (SD) total time to decannulation was 8.13 (7.09) days (P = .003). We found no association between adverse events and use of the pathway.
Conclusions and Relevance
Our low-risk tracheotomy clinical pathway is associated with a sustainable decrease in total time to decannulation without any associated increase in adverse events. We therefore believe that this pathway is a safe and effective tool to guide clinicians in the management of tracheotomy.
Tracheotomy is a very common procedure, and more than 500 are performed at our center annually. Most tracheotomies are performed for prolonged intubation, with a small proportion for emergency airway management and the treatment of head and neck cancer.1 A tracheotomy is usually a temporary measure with the ultimate long-term goal of decannulation. Tracheotomies are associated with a variety of complications that are generally classified as intraprocedural, early postprocedural, and late postprocedural. Subglottic stenosis, tracheal stenosis, tracheomalacia, and tracheoinnominate, tracheoesophageal, and tracheocutanous fistulas are examples of late complications that are more common with prolonged tracheotomy placement.2- 4 Because of the risks associated with tracheotomies, cannulation should be maintained for the shortest time appropriate. Another factor that contributes to the risk for complications is that postoperative tracheotomy care is often managed by nonexperts or health care professionals who are unfamiliar with such care, thereby creating the potential for suboptimal care.5,6
At present, no protocol exists to guide nonexperts through the process of decannulation in tracheotomy, and existing guidelines are based on expert opinion alone.7- 12 Other centers have identified variable and unorganized tracheotomy care as an ongoing issue.5,9,12,13 Although this concern is recognized, few attempts have been made to improve and standardize tracheotomy care.11 Decannulation remains an ill-defined and vague process, with few attempts to address this issue.
We surveyed our center’s experience with tracheotomy care and became concerned regarding the practice variations in the management of tracheotomy. We observed that the total time to decannulation varied significantly and that the process of decannulation lacked well-defined milestones and standardization. These observations confirmed the need for a tool to assist clinicians through the process of decannulation. We formed a team of expert physicians (otolaryngologists), respiratory therapists, nursing staff, and other health care workers whose goal was to create a clinical care pathway that would allow safe and efficient decannulation.
Clinical care pathways have been gaining popularity for a variety of reasons. For example, they improve patient outcomes and decrease hospital costs.14,15 However, the sustainability of these pathways is not as certain. As with many new interventions, the initial success of clinical care pathways is partially a result of early enthusiasm and commitment. Over time as that initiative dies, too often the pathway and its success fade with it. This phenomenon is well-documented and presents a challenge to those attempting to improve patient care with these tools. Clinical care pathways are rarely examined for long-term efficacy after their initial success.
Postoperative tracheotomy care is a well-recognized and persistent issue, with potentially grave consequences. Our center confirmed the need for a tool to address this issue and created a clinical care pathway to define the process of decannulation. Our objective was to assess the impact and efficacy of this pathway and to examine the sustainability of the pathway in clinical practice.
This longitudinal study involved 3 consecutive cohorts. The first baseline cohort consisted of patients who underwent tracheotomy from July 1 through November 30, 2008. Data from the baseline cohort were collected by retrospective medical record review. In a second pilot cohort, the low-risk tracheotomy decannulation pathway was developed and prospectively evaluated from February 1, 2009, through February 28, 2010. The pilot implementation was performed in multiple nursing units at a single hospital site. A third follow-up cohort underwent prospective evaluation from March 1, 2010, through January 31, 2012, to assess the continued performance of the low-risk tracheotomy decannulation pathway. The follow-up cohort included patients in multiple nursing units at the participating hospitals. Potentially eligible patients underwent initial screening by their attending physician and/or a respiratory therapist. The inclusion and exclusion criteria (see below) were applied to determine whether the patient was eligible for inclusion in the low-risk pathway. The protocol was reviewed by the Conjoint Health Research Ethics Board and approved as a quality improvement initiative. Patients did not provide informed consent as this is not deemed necessary for quality improvement initiatives. All patients admitted to the hospital formally consent to treatment and use of their clinical data for quality improvement.
Patients were deemed eligible for the low-risk pathway if they were transferred to a regular-acuity nursing unit with a new tracheotomy and if they were older than 18 years. Patients with percutaneous and open tracheotomies were included. The presence of high-risk criteria meant that the patient was not treated under the low-risk protocol. Patients were deemed at high risk for decannulation and explicitly excluded from the pathway for a variety of reasons owing to the potential for airway obstruction (Box 1). The pathway was applied to inpatients only because outpatient decannulation occurs in a population of patients with compromised airways that may not be addressed appropriately by our pathway.
Documented traumatic intubation
Major oral/facial swelling
Major head and neck surgery
Upper airway surgery
Severe vocal cord damage/dysfunction
Anticipated difficulty with tracheotomy tube change or reinsertion due to
(1) Unfavorable neck anatomy (eg, severe obesity)
(2) Cervical spine limitations
(3) History of a coagulopathy
(4) Initial placement of a specialty tracheotomy tube (eg, Shiley XLT [Covidien])
All eligible patients were entered into the low-risk pathway (eFigure in the Supplement). The low-risk tracheotomy decannulation pathway provides an in-depth stepwise approach to initial cuff deflation, tracheotomy tube change, initiation of cork occlusion trials, and decannulation based on the patient meeting a series of defined milestones. Scheduled clinical assessments and physiological measures of these milestones determined the individual patient’s progress through the pathway. Patients failing to achieve a given pathway milestone after 3 attempts received a mandatory otolaryngology consultation to determine the cause and appropriate management.
The primary outcome measure was time to decannulation, which was defined from the time of arrival on the nursing unit. We also reviewed adverse events to assess whether the pathway was causally related to complications. Adverse events included death, loss of airway, code 66 (unstable patient), or code blue (cardiopulmonary arrest) (Box 2) and reinsertion of the tracheotomy tube or endotracheal intubation after decannulation. Adverse events were identified from the intensive care unit database and medical record review. Events that occurred before the patients’ tracheotomy or more than 30 days after decannulation were excluded.
Code 66 (unstable patient requiring urgent assessment for resuscitation and transfer to the intensive care unit) for
(1) Airway compromise
(2) Respiratory rate <8 or >30 breaths/min
(3) Acute change in oxygen saturation level to <90% despite oxygen administration of >5 L/min
(4) Heart rate <40 or >140 beats/min
(5) Systolic blood pressure <90 mm Hg
(6) A sudden decrease in level of consciousness or in Glasgow Coma Scale score of >2 points
(7) Prolonged or repeated seizures
(8) Acute change in urinary output to <50 mL in 4 h
(9) Any other medical concerns requiring intensive care unit services
Data were collected by registered respiratory therapists and entered into a database. Supervisors and educators were actively involved in monitoring data collection for accuracy and reliability. We compared the pilot and follow-up cohort data with the baseline cohort data for the analysis. We compared categorical outcomes using a χ2 or a Fisher exact test as appropriate and continuous outcomes using a nonparametric (Wilcoxon rank sum) test. A prospective power calculation was not performed for this study; however, a post hoc power calculation for the pilot cohort revealed that a sample size of 17 participants per group would detect a difference in the primary outcome as great as (or greater than) that found with a power of 0.90 and an α value of .05. A similar calculation for the follow-up cohort revealed that it was relatively underpowered, primarily because of the high SD in the nonpathway group. P < .05 was considered significant for all tests. Statistical analysis was performed using commercially available software (STATA, version 12; StataCorp).
Twenty-six consecutive patients were included in the baseline cohort, and their primary data were used for the subsequent comparisons. In the pilot cohort, 34 patients in multiple inpatient units underwent evaluation for inclusion; 13 were excluded. One hundred seven patients underwent evaluation for eligibility in the follow-up cohort in multiple inpatient units; 68 were excluded. The decannulation procedure of 60 patients in all was managed according to the low-risk tracheotomy decannulation pathway (pilot and follow-up cohorts).
We found no significant difference in age or sex among the 3 groups. The indication for tracheotomy in the baseline cohort was airway management in a greater number of patients than in the pilot and follow-up cohorts, in which the indication was prolonged intubation in most of the patients (P < .01) (Table 1).
In the baseline cohort, the mean (SD) total time to decannulation was 15.50 (12.08) days. Mean total time to decannulation decreased by 9.76 days between the baseline and pilot cohorts to 5.74 (2.79) (P < .001). In the follow-up cohort, the mean total time to decannulation was reduced to 8.13 (7.09) days compared with the baseline cohort (P = .003).
In the baseline cohort, 7 adverse events occurred: code 66 in 5 patients and code blue in 2 patients. No episodes of airway reinsertion occurred. In the pilot cohort, we identified 2 adverse events, both consisting of code 66; both patients required reinsertion of their tracheotomy tube. Three adverse events occurred in the follow-up cohort. Two patients required reintubation, and one of these eventually required a repeated tracheotomy (Table 2).
This prospective longitudinal study is the first to examine the effect of a care pathway on time to decannulation in low-risk patients. The need for a clinical care tool to define and guide clinical practice through the process of decannulation has been previously identified, although no clear, sustainable solution has been found. After clarifying the issue and identifying areas for improvement, our center created a low-risk tracheotomy decannulation pathway. Examination of this clinical care pathway displayed a consistent and sustainable decrease in the total time to decannulation without any increase in complications.
The need for a tool to guide clinicians became clear after a survey of tracheotomy care at our center. We found that the times to first tracheotomy tube change, to initiation of cork occlusion trials, and from successful cork occlusion trial to decannulation varied from 1 to 19, 0 to 19, and 0 to 50 days, respectively. The care being provided to tracheotomy patients varied widely for no identifiable medical reason. Through the formation of the Tracheotomy Management Initiative, consisting of otolaryngologists and various other health care workers, the pathway was created. After the creation of clear milestones based on expert opinion to address this variability, the pathway underwent multiple revisions during team discussions before its clinical implementation. Since it has been introduced to clinical practice, no changes to the pathway have been necessary.
The total time to decannulation increased slightly in the follow-up cohort compared with the pilot cohort. One of the phenomena associated with the implementation of clinical care pathways is initial success with subsequent decreased efficacy due to fading enthusiasm, care, and commitment. Although we noted a 2.39-day increase in the total time to decannulation, this change still translated to a significant overall reduction in total time to decannulation from the baseline cohort. This change suggests that during the 3 years that the decannulation pathway has been in use and during the 2 years after the pilot study, the pathway continued to have a significant clinical impact. The total time to decannulation in the follow-up cohort may actually provide a more accurate assessment of the pathway’s impact, because during this period the pathway was used by health care practitioners who were not dedicated to its success. We believe this assessment reflects the sustainability of the pathway in general hospital practice over time. At the same time, our assessment also highlights the need for ongoing education and communication of pathway outcomes to health care workers. These measures are effective in keeping clinical staff engaged and focused on continued improvement of quality of care.
The review of adverse events suggests that complications were not increased in patients undergoing pathway protocol decannulation. The primary goal of the review of adverse events was to determine whether a causal relationship existed between the pathway and complications. As such, we will describe only the pathway-treated complications in detail. A closer examination of the patients in the pathway groups revealed that 1 patient in the pilot cohort who experienced an adverse event had undergone decannulation against protocol (failure to perform the cork occlusion trial for 24 hours) owing to physician error. The second patient in the pilot cohort with an adverse event experienced a complication of percutaneous endoscopic gastrostomy feeding, which resulted in aspiration pneumonia 2 days after decannulation. In the follow-up cohort, 1 patient experienced pulsatile bleeding after the first tracheotomy tube change on postoperative day 5. The source was a bleeding superficial blood vessel that was controlled with a single suture. The second patient underwent reintubation 5 days after decannulation, during a recurrence of his original presenting symptoms (code blue after chest pain, hypertensive crisis, and seizure). After intubation, he eventually received a second tracheotomy for prolonged intubation. The patient underwent eventual decannulation with no complication, and the cause of his symptoms was never identified. The third patient underwent assessment on postdecannulation day 1 for a code 66 for acute agitation in the setting of a history of severe dementia and ongoing delirium. He underwent reintubation after medical management of his agitation and extubation with no complication or respiratory symptoms the following day. We have difficulty concluding that the pathway itself was responsible for these complications, but the adverse event review suggests that complications do not increase in patients undergoing decannulation according to the pathway.
The indication for tracheotomy in the baseline cohort was largely airway management, whereas the indication in the pilot and follow-up cohorts was primarily prolonged intubation. Airway management includes patients who received a tracheotomy for a variety of reasons other than prolonged intubation, including pulmonary toilet or upper airway swelling. Some of these indications might have placed these patients in the high-risk category, and we believe this categorization explains the difference in indication between the groups. However, we believe that the baseline cohort represents an appropriate comparison group because most of these patients would have been eligible for management according to the pathway.
This study provides insight into the effectiveness of a decannulation pathway but has some limitations. The baseline cohort ideally would have undergone prospective assessment. However, it represents the best available comparison group, and we believe the cohort appropriately represents the original starting point for tracheotomy care at our center. The sample sizes of the initial baseline and pilot cohorts are relatively small. The pathway requires careful patient selection: it is by definition a low-risk pathway, and we stress that it should only be applied to those patients. This pathway is not intended for patients with complex upper airway concerns, although perhaps it could be used as a framework for the development of a high-risk pathway. Despite these limitations, the findings of this study are strengthened by the reliable prospective data collection, the uniqueness of the design and pathway, and monitoring for adverse events associated with the care pathway.
This prospective longitudinal study evaluated the impact, sustainability, and safety of a low-risk tracheotomy decannulation pathway. We found that the pathway consistently decreased the time to decannulation and has had a sustainable impact in clinical practice over time. In addition, its use was not associated with any increase in adverse events, illustrating the safety of the pathway. This pathway has changed how we manage tracheotomy at our center and continues to be used at multiple sites in our health region. It provides a safe, systematic, and sustainable approach to tracheotomy decannulation.
We plan to improve the understanding of the roles and goals of the pathway in patients undergoing tracheotomy to increase its clinical use. We also plan to maintain the pathway over time with frequent retraining and educational sessions and continued data collection for ongoing analysis. This process will allow us to address any new issues that arise and to fine-tune the pathway as required.
Submitted for Publication: January 8, 2014; final revision received March 22, 2014; accepted April 25, 2014.
Corresponding Author: Joseph C. Dort, MD, Ohlson Research Initiative, Section of Otolaryngology–Head and Neck Surgery, Department of Surgery, University of Calgary, HRIC Room 2A02, 3280 Hospital Dr NW, Calgary, AB T2N 4Z6, Canada (firstname.lastname@example.org).
Published Online: June 12, 2014. doi:10.1001/jamaoto.2014.921.
Author Contributions: Dr Dort had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Smith, Matthews, Dubé, Dort.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Dort.
Administrative, technical, or material support: Smith, Dubé, Spence, Dort.
Study supervision: Matthews, Dubé, Dort.
Conflict of Interest Disclosures: None reported.
Additional Contributions: We thank the many respiratory therapists, nurses, physicians, and other health care providers whose collaboration and support made the development and implementation of this pathway possible. The Ohlson Research Initiative, in particular Steve Nakoneshny, BSc(Hons), supported the data management required for this project, for which he received compensation. Alberta Health Services provides continued support for this project.