Association of Care Engagement and Psychosocial Factors With Tracheostomy Decannulation

ABSTRACT

Objective

To evaluate associations between care engagement and psychosocial factors and decannulation success and timing among adults with laryngotracheal stenosis and tracheostomy dependence.

Methods

This retrospective cohort study included 60 adults with tracheostomy dependence in the setting of laryngotracheal stenosis who underwent airway surgery between September 2021 and November 2025. Patient demographics, clinical history, psychiatric and substance use history, tracheostomy information, surgical information, and follow‐up care were extracted from electronic medical records. Fisher's exact tests, chi‐squared tests, and accelerated failure time models were used to assess associations with decannulation timing and success.

Results

Forty‐seven patients (78.3%) achieved decannulation. No significant differences in ability to decannulate were found based on age, sex, race, body mass index (BMI), comorbidities, reason for tracheostomy, or stenosis characteristics. Patients who were not decannulated had significantly higher rates of missed appointments (38.5% vs. 8.5% with ≥ 2 no‐shows, p = 0.025) and were more frequently lost to follow‐up (p < 0.001). Each missed appointment was associated with a 25% increase in time to decannulation (p = 0.004). Patients with psychiatric diagnoses had a longer time to decannulation, though this relationship was not statistically significant (p = 0.091). Substance use diagnosis, distance to hospital, social vulnerability index, emergency department visits, and complications were not significantly associated with outcomes.

Conclusion

Care engagement patterns, including better appointment adherence, were significantly associated with improved decannulation rates in patients with laryngotracheal stenosis. Psychiatric comorbidities may delay decannulation. These findings suggest that interventions targeting consistent follow‐up and mental health support may improve tracheostomy outcomes in this complex patient population.

Level of Evidence

4.

This retrospective cohort study evaluated associations between care engagement and psychosocial factors and decannulation success and timing among adults with laryngotracheal stenosis and tracheostomy dependence. We found that care engagement patterns, including better appointment adherence, were significantly associated with improved decannulation outcomes, and that psychiatric comorbidities may delay decannulation. These findings suggest interventions targeting consistent follow‐up and mental health support may improve tracheostomy outcomes in this complex patient population.

1. Introduction

Tracheotomy is a common intervention for patients requiring prolonged intubation in the setting of respiratory failure and may also be indicated for patients with benign or malignant airway obstruction [1, 2, 3]. For survivors of critical illness, decannulation represents a major milestone: restoring airway independence, improving communication, reducing infection risk, and enhancing long‐term quality of life [4, 5, 6]. Despite these benefits, achieving decannulation remains a well‐recognized challenge, with rates varying widely based on comorbidities, airway pathology, and institutional practices [5, 6, 7, 8, 9, 10]. Identifying modifiable predictors of successful decannulation is therefore essential to optimizing tracheostomy care [11, 12].

Existing research has largely focused on physiologic or anatomical determinants, such as pulmonary reserve, secretion burden, airway stenosis, and swallowing function as the drivers of decannulation [13, 14, 15, 16, 17, 18, 19, 20]. Less emphasis has been placed on social determinants of health (SDoH) that shape patients' ability to engage in tracheostomy care. These factors influence surgical recovery and care adherence across Otolaryngology and critical care settings, but their contribution to tracheostomy outcomes remains insufficiently examined [21, 22, 23]. Indeed, individuals living with tracheostomies may face psychosocial and behavioral challenges, which could impact their ability to follow‐up reliably and reach decannulation [12, 24, 25, 26, 27, 28, 29].

This study assesses how care engagement and psychosocial factors relate to tracheostomy outcomes, in particular decannulation, among adults treated at a tertiary academic center. We hypothesized that patients with greater psychosocial vulnerability and worse care engagement would have lower decannulation rates. By characterizing these associations, we aim to identify potential targets for intervention, including mental health services, substance use treatment, and enhanced care navigation, that may improve the likelihood of safe, timely decannulation.

2. Methods

2.1. Patient Cohort

This was a retrospective, single‐institution cohort study conducted at an academic tertiary care medical center. All patients who presented with tracheostomy dependence in the setting of laryngotracheal stenosis (leading to speaking valve or capping intolerance), desired decannulation, and underwent at least one operation to improve airway patency from September 2021 through November 2025 were included. Patients were excluded if they did not desire decannulation, decannulation was deemed unsafe or inadvisable, and/or if they were ventilator dependent. Patients were identified using a prospectively maintained departmental database of patients who underwent airway surgery. This study was approved by Northwestern's Institutional Review Board as an expedited review (STU00223584).

2.2. Data Collection and Analysis

All data was collected from each patient's electronic medical record (EMR) by reviewing clinic notes, operative reports, and scheduling documentation. Data retrieved included demographic information, clinical history, psychiatric and substance use history, social vulnerability index (SVI) via geocoded addresses to census tracts, tracheostomy information, surgical information, and follow‐up care. Tracheostomy history and pertinent information were recorded at the patient's initial visit and throughout their clinical course. Follow‐up and scheduling information were determined from the EMR including cancellations and patient communications with the team regarding scheduling.

To determine the reason for tracheotomy, etiology of stenosis, and anatomic location of stenosis, each laryngologist that was responsible for the patient's care verified the relevant records. Multilevel stenosis was defined as stenosis at two unique, discontinuous locations within the airway.

Outcomes were assessed through careful review of internal and external records to identify complications across all health systems of each patient's' care. Records were queried to delineate comorbidities before tracheotomy as well as pre‐ and post‐decannulation complications. This approach ensured accurate reporting of complex variables and clear differentiation between pre‐existing comorbidities and those that developed after tracheotomy.

2.3. Clinic Follow‐Up Protocol

Routine clinic visits are essential to ensure patients progress toward their goals while reducing the risk of unforeseen complications or compromise to airway stability. Following endoscopic airway surgery, patients were routinely scheduled to return to clinic within 3–4 weeks for an initial postoperative visit. If they were doing well and progressing toward decannulation, we would schedule another follow‐up 4–6 weeks later, followed by a third visit about 4–6 weeks after that. At the third postoperative visit, which on average would land about 3 months after surgery, we would decannulate patients if they had been tolerating capping all day and night for multiple weeks without issue. After decannulation, patients were recommended to return 4–5 weeks later, followed by additional visits at three, six, and 12 months post‐decannulation to ensure airway caliber remained stable.

Appointments were scheduled for each patient by our clinical team before their visit ended, and confirmation messages via email or text (per patient preference) are routinely sent at 2 weeks, 1 week, and 48 h before the visit. Patients who did not attend a scheduled visit were contacted the same day by the attending physician or a member of the clinical staff to determine if they would be able to make their appointment later in the day, ascertain the reason for the missed appointment, and/or facilitate rescheduling.

Certainly, follow‐up visit timing was variable based on each patient's specific clinical picture, needs, and the response of their airway to initial endoscopic surgery. If patients required multiple endoscopic surgeries or open airway reconstruction to achieve decannulation, then their follow‐up care pathway would have differed than that outlined above.

2.4. Statistical Analysis

Statistical analysis was performed using RStudio version 4.5.1. Fisher's exact tests were used for contingency tables with expected cell frequencies less than 5. Chi‐squared tests with Yates continuity correction were used for 2 × 2 contingency tables when all expected cell frequencies were 5 or greater. Pearson's chi‐squared tests were utilized for contingency tables 2 × 3 or larger when all expected cell frequencies were 5 or greater. These tests were used to compare demographic characteristics, comorbidities, tracheostomy and stenosis characteristics, psychiatric and substance use comorbidities, follow‐up care, and complications between patients who were and were not decannulated.

To assess factors associated with time to decannulation, we employed a time‐to‐event framework to better account for variable follow‐up durations and patients who did not achieve decannulation. A series of separately adjusted accelerated failure time (AFT) models with a log‐logistic distribution were fit, each evaluating the association between time to decannulation and an individual covariate of interest, with reason for tracheostomy and stenosis characteristics (etiology and location of stenosis) included as adjustment variables. Patients who did not achieve decannulation were treated as censored observations. Two‐sided p‐values were used for all analyses with the α‐criterion at 0.05.

3. Results

3.1. Patient Demographics and Airway Characteristics

Sixty patients were included and 47 patients (78.3%) achieved decannulation. There were no significant differences in decannulation status based on age, sex, race, body mass index (BMI), or comorbidities (Table 1). Of those who achieved decannulation, the duration of tracheostomy tube dependence ranged from 76 days to 14.5 years (Figure 1A). However, after establishing care with our team, time to decannulation ranged from 7 days to 19 months (Figure 1B).

TABLE 1.

Demographic characteristics and comorbidities of patients by decannulation status.

Characteristic	Decannulated, N = 47 (78.3%)	Not decannulated, N = 13 (21.7%)	p
Age	μ = 55.0 ± 15.0	μ = 53.5 ± 19.8	0.257
20–44 years	14 (29.8%)	4 (30.8%)
45–64 years	21 (44.7%)	3 (23.1%)
65+ years	12 (25.5%)	6 (46.2%)
Sex			0.722
Male	24 (51.1%)	8 (61.5%)
Female	23 (48.9%)	5 (38.5%)
Race			0.125
White	16 (34.0%)	8 (61.5%)
Black	18 (38.3%)	2 (15.4%)
Hispanic	11 (23.4%)	2 (15.4%)
Asian or Pacific Islander	2 (4.3%)	0 (0%)
Other	0 (0%)	1 (7.7%)
Body mass index	μ = 30.5 ± 9.1	μ = 29.6 ± 7.7	0.786
Underweight	0 (0%)	0 (0%)
Normal	10 (21.3%)	4 (30.8%)
Overweight	17 (36.2%)	4 (30.8%)
Obese	20 (42.6%)	5 (38.5%)
Comorbidities
Asthma	14 (29.8%)	1 (7.7%)	0.153
Hypertension	34 (72.3%)	9 (69.2%)	1.000
Diabetes	21 (44.7%)	4 (30.8%)	0.560
Cancer	9 (19.1%)	2 (15.4%)	1.000
Endocrine disorder	12 (25.5%)	3 (23.1%)	1.000
Obstructive sleep apnea	11 (23.4%)	2 (15.4%)	0.713
Gastroesophageal reflux disease	15 (31.9%)	2 (15.4%)	0.314

FIGURE 1.

Duration of tracheostomy dependence. Histograms showing total time with tracheostomy tube before decannulation (A) and time with tracheostomy before decannulation once established care with our team at Northwestern Memorial Hospital (NMH) (B). X‐axis scale is not continuous in (A).

Each patient's clinical history and reason for tracheotomy was case dependent but generally fit into three categories: prolonged intubation for respiratory failure (n = 44, 73.3%), malignancy‐related airway obstruction (n = 10, 16.7%), and benign airway obstruction/stenosis (n = 6, 10%). The reason for tracheotomy was not significantly different between patients based on decannulation status (p = 0.679, Table S1). Table S1 further outlines the etiology and location of stenosis for each patient, which did not show any statistically significant differences between patients who were versus were not decannulated.

All patients underwent endoscopic airway surgery initially to try and improve their airway caliber, which generally included microdirect laryngoscopy with CO₂ laser excision of stenosis, steroid injection, balloon dilation, and mitomycin C application. Some patients required multiple endoscopic surgeries, and four needed open airway surgery.

3.2. Care Engagement and Complications

The number of clinic visits prior to decannulation was not significantly different between groups, although patients that did not achieve decannulation required more appointments (p = 0.053, Table 2). Patients that remained tracheostomy dependent had a statistically higher no‐show/late cancellation rate, with 38.5% of this cohort missing two or more scheduled visits, while only 8.5% of patients that were decannulated missed 2+ appointments (p = 0.025). Moreover, 53.8% (n = 7) of patients that did not reach decannulation were lost to follow‐up (p < 0.001).

TABLE 2.

Clinical course before decannulation: Clinic visits and no shows/late cancellations, emergency department visits, and complications by decannulation status.

Characteristic	Decannulated N = 47 (78.3%)	Not decannulated N = 13 (21.7%)	p
Number of clinic visits			0.053
≤ 3 visits	21 (44.7%)	7 (53.8%)
4–7 visits	19 (40.4%)	1 (7.7%)
8–11 visits	5 (10.6%)	3 (23.1%)
≥ 12 visits	2 (4.3%)	2 (15.4%)
No show/late cancel (< 24 h)			0.025
None	32 (68.1%)	7 (53.8%)
1	11 (23.4%)	1 (7.7%)
≥ 2	4 (8.5%)	5 (38.5%)
Lost to follow‐up			< 0.001
No	47 (100%)	6 (46.2%)
Yes	0 (0%)	7 (53.8%)
Emergency department visits			0.326
None	17 (36.2%)	4 (30.8%)
1	13 (27.7%)	3 (23.1%)
2	10 (21.3%)	1 (7.7%)
3+	7 (14.9%)	5 (38.5%)
Tracheostomy complications
Granulation	42 (89.4%)	12 (92.3%)	1.000
Skin issues	10 (21.3%)	3 (23.1%)	1.000
Bleeding	10 (21.3%)	4 (30.8%)	0.478
Tube obstruction	12 (25.5%)	3 (23.1%)	1.000
Trach dislodgement	18 (38.3%)	9 (69.2%)	0.095

Note: Bolded values denote statistical significance (p < 0.05).

The number of emergency department (ED) visits, though not statistically different between the groups (p = 0.326), showed that 38.5% of those who were not decannulated had three or more ED visits compared with 14.9% of those that achieved decannulation (Table 2). Additionally, routine tracheostomy complications including granulation, skin issues, bleeding, and tube obstruction were not statistically different between groups (all p > 0.478). Tracheostomy tube dislodgement, while not statistically significant, occurred more often in patients who were not decannulated compared with those who were decannulated (69.2% vs. 38.3%, respectively; p = 0.095).

3.3. Psychosocial Factors and Substance Use

Distance to the hospital showed no significant difference between groups, with mean distance of 92.0 miles for those decannulated versus 116.1 miles for those not decannulated (p = 0.619, Table 3). SVI did not differ based on decannulation status (p = 0.948), and the majority of both groups scored in the highest vulnerability quartile of 0.75–1.00 (66.0% decannulated, 69.2% not decannulated).

TABLE 3.

Psychosocial factors categorized by decannulation status.

Characteristic	Decannulated N = 47 (78.3%)	Not decannulated N = 13 (21.7%)	p
Distance to hospital	μ = 92.0 ± 305.6	μ = 116.1 ± 219.4	0.619
Within 10 miles	6 (12.8%)	2 (15.4%)
10–25 miles	17 (36.2%)	2 (15.4%)
25–50 miles	12 (25.5%)	4 (30.8%)
50–100 miles	6 (12.8%)	3 (23.1%)
100+ miles	6 (12.8%)	2 (15.4%)
Social vulnerability index	μ = 0.759 ± 0.246	μ = 0.753 ± 0.197	0.948
0–0.25	3 (6.4%)	0 (0%)
0.25–0.5	5 (10.6%)	2 (15.4%)
0.5–0.75	8 (17.0%)	2 (15.4%)
0.75–1.00	31 (66.0%)	9 (69.2%)
Prior psychiatric diagnosis			0.204
Depression alone	2 (4.3%)	0 (0%)
Anxiety alone	5 (10.6%)	2 (15.4%)
Multiple diagnoses	7 (14.9%)	5 (38.5%)
None	33 (70.2%)	6 (46.2%)
Substance use diagnosis			0.892
Tobacco/nicotine dependence	8 (17.0%)	1 (7.7%)
Alcohol	4 (8.5%)	2 (15.4%)
Marijuana	2 (4.3%)	0 (0%)
Polysubstance abuse	4 (8.5%)	1 (7.7%)
No diagnosis	29 (61.7%)	9 (69.2%)

Prior psychiatric diagnoses were present in 29.8% of those who achieved decannulation compared with 53.8% of those who were not decannulated, and patients who were not decannulated had a higher proportion with multiple psychiatric diagnoses (38.5% vs. 14.9%, respectively). However, these differences did not reach statistical significance (p = 0.204, Table 3). Finally, substance use did not differ significantly between groups (p = 0.892), and over 60% of patients in each group reported no history of substance use.

3.4. Time to Decannulation

On AFT analysis, each no show/late cancellation was associated with a 25% longer time to decannulation (TR: 1.25, 95% CI: 1.07–1.45, p = 0.004, Table 4). Although not statistically significant, patients with prior psychiatric diagnoses had a longer time to tracheostomy removal, with these patients taking 72% longer to decannulate (TR: 1.72, 95% CI: 0.92–3.21, p = 0.091). Distance to hospital, substance use diagnosis, SVI, ED visits, and complications did not demonstrate significant associations with the duration of tracheostomy dependence (all p > 0.127).

TABLE 4.

Separately adjusted accelerated failure time models of associations between time to decannulation and care engagement or psychosocial factors.

Characteristic	TR	95% CI	p
No show/late cancel (< 24 h)	1.25	1.07, 1.45	0.004
Emergency department visits	1.06	0.89, 1.27	0.530
Distance to hospital	1.00	1.00, 1.00	0.127
Prior psychiatric diagnosis—yes	1.72	0.92, 3.21	0.091
Substance use diagnosis—yes	1.46	0.82, 2.59	0.199
Social vulnerability index	0.66	0.22, 1.91	0.438
Complications—yes	1.21	0.33, 4.38	0.777
Granulation	0.98	0.34, 2.83	0.975
Skin issues	0.88	0.41, 1.9	0.742
Bleeding	1.04	0.49, 2.21	0.917
Tube obstruction	0.89	0.45, 1.78	0.749
Trach dislodgement	1.12	0.62, 2.05	0.706

Note: Analyses conducted while adjusting for reason for tracheostomy, etiology of stenosis, and location of stenosis. Bolded values denote statistical significance (p < 0.05).

Abbreviations: 95% CI, 95% confidence interval; TR, time ratio.

4. Discussion

This work is one of the first to examine how psychosocial and behavioral factors relate to tracheostomy outcomes in adults with laryngotracheal stenosis and tracheostomy dependence. Specifically, we evaluated associations between care engagement, psychiatric comorbidities, and substance use disorders on decannulation success and timing. Since all patients in this study required surgery to improve airway patency, they also needed longitudinal follow‐up before decannulation. Importantly, the data presented here show patients with inconsistent appointment attendance or those lost to follow‐up were less likely to achieve decannulation or experienced a longer duration of tracheostomy tube dependence. Indeed, this patient population may be vulnerable to disruptions in care, amplifying the influence of psychosocial factors on airway outcomes.

Appointment adherence emerged as one of the most clinically relevant findings. Very few patients who were decannulated missed more than one scheduled visit. On the other hand, patients who were not decannulated had a significantly higher proportion of two or more no‐shows/late cancellations. Notably, each no‐show or late cancellation was associated with a 25% longer time to decannulation. In this manner, missed visits were associated with both reduced likelihood of decannulation and prolonged tracheostomy dependence, highlighting consistent follow‐up as a valuable marker of decannulation success. Indeed, while appointment adherence may be most consequential to achieving safe decannulation, missed visits after tracheostomy tube removal remain valuable to ensure recurrent dyspnea or restenosis does not occur.

Missed visits may serve as an independent predictor, due to a behavioral pattern of disengagement from care, or as a marker of underlying social and clinical complexity, including transportation barriers, competing life demands, mental health burden, and/or substance use, rather than patient motivation alone. Distinguishing between these possibilities has important implications for intervention: patients who are disengaged may benefit from targeted outreach, communication, and care navigation, while those facing structural barriers may require more upstream support such as transportation assistance, social work involvement, or flexible scheduling. In either case, identifying patients at risk for missed visits early in their clinical course and connecting them with appropriate resources may be among the most impactful strategies to improve decannulation outcomes in this population.

Moreover, while the total number of clinic visits did not differ significantly between groups, patients who were not decannulated had more appointments, reflecting prolonged clinical courses. Interestingly, there was also a group of over 50% of patients who were not decannulated that had fewer than three visits. We suspect these patients with a lower number of appointments could be due to being lost to follow‐up, which arose in 53.8% of patients that did not reach decannulation.

ED utilization and tracheostomy‐related complications were not significantly associated with decannulation success or duration of tracheostomy dependence. One interpretation is that short‐term complications may not impede eventual decannulation, and in some cases, complications could prompt closer follow‐up or earlier consideration of decannulation. Prior literature remains mixed, with some studies reporting lower decannulation rates in patients with complications while others, such as Zaga et al., demonstrate higher adverse event rates among patients following decannulation [6, 10, 17]. Our results align with the heterogeneity of previous literature by suggesting that ED visits and complications alone may be insufficient predictors of decannulation outcomes.

The literature on socioeconomic status and tracheostomy outcomes remains variable [30, 31, 32, 33]. While greater travel distance was hypothesized to impede decannulation, we found no significant findings in our cohort. This may reflect referral patterns to a tertiary care center or the increasing use of telehealth, although broader investigations across practice settings are needed. Furthermore, SVI scores were uniformly high across all patients in this work, so there were not any statistically significant findings between SVI and decannulation success. Although SVI score alone was not significantly associated with decannulation, we believe these patients may benefit from enhanced social support and care coordination considering the high baseline SVI for this entire patient cohort.

Prior studies have investigated the influence of neurological issues on tracheostomy outcomes, but none have assessed psychiatric comorbidities [9, 10, 13, 34]. Psychiatric complexity may influence tracheostomy outcomes through effects on care engagement, appointment adherence, and/or treatment tolerance. Over half of patients who were not decannulated had a documented psychiatric diagnosis, and a higher proportion had multiple psychiatric comorbidities compared with those who were decannulated (38.5% vs. 14.9%, respectively). Moreover, patients with a prior psychiatric diagnosis took 72% longer to decannulate, though this finding was not statistically significant. Finally, patients with substance use disorders, often comorbid with psychiatric conditions, demonstrated no statistical association with prolonged tracheostomy dependence. Since neither psychiatric comorbidities nor substance use disorders reached statistical significance in chi‐square or time‐to‐event analyses, it may suggest that this study was underpowered due to the small sample size rather than a lack of a clinical relationship. Future work would benefit from further investigation with larger cohorts.

Several limitations of this study should be acknowledged. First, its retrospective methodology limits the conclusions that can be drawn with respect to causation versus association. This also prevents prospective recording of these variables, so greater understanding and nuance may be lost with respect to some of the more descriptive outcomes. Second, the small sample size and imbalance between decannulated and non‐decannulated patients limit statistical power, particularly for analyses involving psychiatric and substance use subgroups. Due to this constraint, separately adjusted models were employed for each covariate rather than a single unified multivariable model, which limits control for confounding and means findings should be interpreted with appropriate caution pending validation in larger cohorts. Additionally, this single‐institution cohort represents a narrowly defined population of patients with tracheostomy dependence due to laryngotracheal stenosis, which may limit generalizability. Referral patterns to a tertiary academic center may also introduce selection bias, as patients seen at such institutions may differ from those managed in community settings with respect to disease severity, access to care, and psychosocial complexity. However, this focus also represents a strength, as this population requires close longitudinal follow‐up, which could be affected by behavioral and psychosocial factors.

Future studies should prioritize prospective, multicenter validation of these findings to determine whether they translate consistently across different populations, clinical settings, and geographic regions. Comparative studies examining decannulation outcomes across different tracheostomy dependent populations would help clarify if certain psychosocial factors are most impactful in specific clinical contexts. Importantly, interventions targeting appointment adherence and close clinical follow‐up merit attention, potentially utilizing automated tools within the EMR as well as identifying patients who may require greater care coordination. Finally, further research should explore whether integrated mental health and substance use resources can improve decannulation rates and timing for select patients.

5. Conclusion

This study represents one of the first investigations into associations between behavioral and psychosocial factors and decannulation outcomes in adults with tracheostomy dependence and laryngotracheal stenosis. Care engagement patterns, particularly appointment adherence, were significantly associated with both decannulation success and timing. Psychiatric comorbidities were more prevalent in patients with prolonged tracheostomy dependence, but the association with an increased time to decannulation was not statistically significant. These results highlight the importance of addressing psychosocial vulnerabilities alongside anatomic and physiologic issues in tracheostomy care. By identifying modifiable behavioral and social factors, this work paves the way for enhanced care coordination, mental health support, and substance use treatment, which may improve decannulation rates for this complex patient population.

Funding

The authors have nothing to report.

Disclosure

No AI tools were used in the preparation of this manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Reason for tracheostomy and etiology & location of stenosis by decannulation status.

Abrahamson C. W., Landini A. L., Frost M., Gutierrez J. A. III, Burns J. A., and Stein A. P., “Association of Care Engagement and Psychosocial Factors With Tracheostomy Decannulation,” Laryngoscope Investigative Otolaryngology 11, no. 4 (2026): e70543, 10.1002/lio2.70543.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.