ABSTRACT
Objective
To quantify 30‐day emergency department (ED) visits and hospital readmissions post‐tracheostomy—distinguishing between tracheostomy and non‐tracheostomy‐related causes—and to identify revisit risk factors.
Methods
Retrospective cohort study of adult patients undergoing all‐cause tracheostomy within the system between April 1, 2018, and April 1, 2025. Electronic Medical Record review captured demographic variables, outcomes, and mortality. Fisher's exact test and Fine and Gray's subdistribution hazard modeling assessed associations with in‐hospital death and risk of revisit.
Results
Among 618 patients, 119/618 (19.3%) died prior to discharge. Of the 499 patients discharged alive, 23 (4.6%) died within 30 days, for a cumulative 30‐day mortality of 142/618 (23.0%). Among those alive at discharge, 57 patients (11.4%) accounted for 68 ED visits and 95 (19.0%) for 101 readmissions. ED‐presenting patients were more likely to be admitted than discharged (36/57, 63.2%) and accounted for 36/95 (37.9%) of all readmissions—indicating most readmissions occurred as direct admissions from outside the ED. Most revisits were due to non‐tracheostomy‐related complications. Cardiovascular disease was a significant predictor of in‐hospital mortality, while gastrostomy dependence was protective (p = 0.012, p < 0.001). Emergent tracheostomy, revision‐tracheostomy, and discharge home increased ED utilization (p < 0.001, p = 0.042, p = 0.017). Gastrostomy dependence and presence of a ventriculoperitoneal shunt predicted readmission (p = 0.005, p < 0.001). Non‐English primary language increased ED revisits, while Hispanic/Latino ethnicity was protective against readmission (p = 0.051, p = 0.049). Longer initial hospitalization decreased likelihood of both ED visits and readmissions (p = 0.017, p = 0.025).
Conclusions
Identification of risk factors for postdischarge hospital utilization can inform trajectory, allowing for realistic discharge planning, improving outcomes while reducing healthcare costs.
Level of Evidence
3b
Keywords: emergency department visit, hospital readmission, in‐hospital mortality, outcomes, tracheostomy
1. Introduction
Tracheostomy is a frequently performed surgery associated with complications ranging from minor hemorrhages and mucous plugs to serious tracheoinnominate arterial or tracheoesophageal fistula and acute obstruction [1, 2]. Comorbidities such as cardiovascular disease (CVD), diabetes, and gastroesophageal reflux disease (GERD), among others, underscore the medical complexity of this patient population. This results in a group highly vulnerable to complications, high healthcare resource utilization, and notable inpatient and postdischarge mortality. Due to the high acuity of this patient population, it is important to examine survival and care trajectory after discharge.
The breadth of complications likely contributes to the high rate of 30‐day readmission, defined as readmission to any hospital within 30 days of discharge. The need to define factors that influence the frequency of readmissions and emergency department (ED) visits post‐tracheostomy across all age groups was recognized by the American Academy of Otolaryngology‐Head and Neck Surgery Foundation (AAO‐HNSF) [3]. Subsequently, various studies have characterized a range of patient‐specific factors, sociocultural determinants of health, and comorbidities as contributing to increased readmission [4, 5, 6]. However, most existing studies on complications and readmissions post‐tracheostomy focus on the pediatric population. Additionally, little has been described about ED visits after tracheostomy and whether those visits result in readmission. This leaves an opportunity for improvement, especially in the adult population.
Proper risk stratification must be done on readmissions to properly use 30‐day readmits as a quality metric, distinguishing between avoidable readmissions and those reflecting appropriate clinical need. Currently, 30‐day readmissions are tracked via the Hospital Readmissions Reduction Program (HRRP), established under the 2010 Affordable Care Act [7]. Specifically, hospital readmissions contribute to a payment adjustment factor, a weighted average performance metric applied to all‐cause discharges in the applicable fiscal year. This can lead to a percentage decrease in Medicare reimbursement for hospitals [8]. Timing of readmissions can also directly impact payment from a billing perspective as the Department of Health & Human Services (DHHS) Centers for Medicare & Medicaid Services (CMS) states that readmissions within 24 h may be billed under the initial admission, though some providers may bundle claims with justification up to 72 h [9, 10].
The Agency for Healthcare Research and Quality (AHRQ) reported respiratory system diseases were associated with a 17% 30‐day readmission rate, the fifth highest among all procedures performed [11]. The high frequency of readmissions post‐tracheostomy further increases the cost burden placed on the healthcare system, incentivizing investment in readmission prevention. As physician and hospital reimbursement become increasingly tied to this metric, understanding readmission rates becomes an important comparative tool. Preventing readmissions is important for reducing costs to the healthcare system and improving patient outcomes.
Objectives of this study included (1) quantifying rates of ED visits and hospital readmissions following discharge post‐tracheostomy, (2) distinguishing between tracheostomy‐related, respiratory‐related, and non‐tracheostomy‐related complications, and (3) identifying risk factors and potentially preventable contributors to return visits. These insights may inform improvements in patient education, discharge counseling, and resource allocation.
2. Methods
2.1. Eligibility Criteria
This study was approved by the University of Nebraska Medical Center Institutional Review Board (IRB # 0503‐25‐EP). All adult patients ≥ 19 years old who underwent a tracheostomy, regardless of indication or surgical service, within the University of Nebraska system between April 1, 2018, and April 1, 2025, were included. Patients discharged after revision tracheostomy were also included. Patients were excluded if they were < 19 years of age at the date of tracheostomy, if they had a tracheostomy performed at another institution, or had a total laryngectomy. This resulted in 618 patients who fit our criteria. This approach for inclusion and exclusion criteria is similar to prior studies on readmission rates [3].
2.2. Study Variables
The primary outcome was nonelective readmission within 30 days of tracheostomy. Secondary outcomes included ED visits within 30 days of discharge. A 30‐day readmission was defined as any subsequent inpatient stay occurring within 30 days of initial discharge, regardless of primary visit diagnosis. This definition has been utilized in prior studies on readmission rates [3]. The primary visit or readmission diagnosis was identified by the first three digits of the International Classification of Diseases, Tenth Revision (ICD‐10) code [12]. This database‐derived variable identifies the admission diagnosis responsible for most of the resource utilization during length of stay. Further characterization, utilizing ICD‐10 codes, grouped visit diagnoses into tracheostomy‐related, non‐tracheostomy‐related, or respiratory‐related complications.
In the case of multiple identical outcomes (i.e., multiple readmissions within the 30‐day period), only the first index was used in further analysis. Multiple outcomes were considered only when plotting time course of events. Covariates collected a priori included demographic characteristics, comorbidities, and preoperative risk factors based on clinical relevance and prior literature, primary insurance payer, and discharge disposition. Public insurance included programs such as Medicaid, Medicare, and Tricare; private insurance referred to commercial plans; and self‐pay indicated uninsured patients. Discharge disposition was grouped into facility‐based care (long‐term acute care hospitals [LTACs], skilled nursing facilities [SNFs], and extended care facilities [ECFs]); transitional or step‐down care (temporary or intermediate facilities bridging acute hospitalization to long‐term placement or community return); and home‐ or community‐based care (discharge to home with or without skilled services, or to community settings such as homeless shelters). Comorbidities and technology dependencies were defined by the first three digits of the International Classification of Diseases, Tenth Revision (ICD‐10) code.
2.3. Statistical Analysis
Standard descriptive statistics were used to summarize data for all patients. Fisher's exact test was used to look at the association of patient characteristics with in‐hospital mortality. For analysis of return ED visits or hospital readmission, only patients who survived the initial hospitalization were included in the analysis (n = 499). Time to revisit was defined as days from hospital discharge to ED visit or readmission. Observations were right censored on May 20, 2025, as the last patient—discharged on May 18, 2025, with no ED visit or readmission—completed follow‐up on that date. Death from any cause prior to the outcome of interest was treated as a competing event. Gray's test compared cumulative incidence functions between groups, and Fine and Gray's proportional subdistribution hazards model estimated covariate effects on ED revisits and readmissions [13]. Results are shown as hazard ratios and 95% confidence intervals. All analyses were done using SAS, version 9.4 and a p < 0.05 was considered statistically significant.
3. Results
3.1. Patient Demographics and Characteristics
A total of 618 adults underwent tracheostomy from April 1, 2018, to April 1, 2025, within the University of Nebraska system. Table 1 summarizes demographic and preoperative risk factors of the patient population. Most patients were in the 40 to 65 years old age group (295/618, 47.7%), male (375/618, 60.7%), English‐speaking (559/618, 90.5%), and Caucasian (436/618, 70.6%). The primary insurance payer was primarily public (351/618, 56.8%). The average patient had 2.2 markers of clinical complexity, most common being obesity (129/618, 20.9%), CVD (132/618, 21.4%), GERD (144/618, 23.3%), anxiety (154/618, 24.9%), diabetes (195/618, 31.6%), and gastrostomy dependence (213/618, 34.5%). Regarding tracheostomy considerations, most tracheostomies were initial surgeries (n = 601, 97.2%) and performed for elective indications (557/618, 90.1%). Most patients had an average hospital stay of 31–90 days (304/618, 49.2%) and were discharged to facility‐based care (268/618, 43.4%).
TABLE 1.
Patient characteristics (n = 618).
| Characteristics | Number (%) |
|---|---|
| Age | |
| 19–39 a | 130 (21.0%) |
| 40–65 | 295 (47.7%) |
| > 65 | 193 (31.2%) |
| Ethnicity | |
| Not Hispanic or Latino a | 556 (90.0%) |
| Hispanic or Latino | 50 (8.1%) |
| Other | 12 (1.9%) |
| Race | |
| White or Caucasian a | 436 (70.6%) |
| Black or African American | 87 (14.1%) |
| Other/unknown/undisclosed | 95 (15.4%) |
| Sex assigned at birth | |
| Male a | 375 (60.7%) |
| Female | 243 (39.3%) |
| Primary language | |
| English a | 559 (90.5%) |
| Spanish/other | 59 (9.5%) |
| Primary insurance payer | |
| Public a | 351 (56.8%) |
| Private | 195 (31.6%) |
| Self‐pay/uninsured | 72 (11.7%) |
| Disposition b | |
| Facility‐based care a | 268 (43.4%) |
| Transitional/step‐down care | 127 (20.6%) |
| Home‐ and community‐based care | 102 (16.5%) |
| Indication for tracheostomy | |
| Elective a | 557 (90.1%) |
| Emergent | 61 (9.9%) |
| Procedure type | |
| Tracheostomy | 601 (97.2%) |
| Revision‐tracheostomy | 17 (2.8%) |
| Length of stay | |
| < 10 days a | 19 (3.1%) |
| 10–30 days | 233 (37.7%) |
| 31–90 days | 304 (49.2%) |
| > 90 days | 62 (10.0%) |
| Comorbidities c | |
| GERD | 144 (23.3%) |
| Asthma | 49 (7.9%) |
| Movement disorder | 27 (4.4%) |
| Anxiety | 154 (24.9%) |
| Depression | 38 (6.1%) |
| CVD | 132 (21.4%) |
| Stroke | 98 (15.9%) |
| Diabetes | 195 (31.6%) |
| End‐stage renal failure | 15 (2.4%) |
| Obesity | 129 (20.9%) |
| Technology dependence c | |
| Technology dependence | 115 (18.6%) |
| Wheelchair | 8 (1.3%) |
| Gastrostomy dependence | 213 (34.5%) |
| Ventriculoperitoneal shunt | 15 (2.4%) |
| Implanted port | 7 (1.1%) |
| Other medical devices | 21 (3.4%) |
Abbreviations: CVD, cardiovascular disease; GERD, gastroesophageal reflux disease.
Reference category used in comparative analyses.
Facility‐based care (LTAC, SNF, ECF); transitional/step‐down care (temporary/intermediate facilities); home‐ & community‐based care (home ± services, hospice, shelter).
Comorbidities and technologies dependencies were defined by the first three digits of the International Classification of Diseases, Tenth Revision (ICD‐10) code.
3.2. Quantification of Mortality and Revisit
Of the 618 patients who received tracheostomies during our study, 119/618 (19.3%) died prior to initial discharge. Among the 499 patients alive at discharge, 23/499 (4.6%) died within 30 days, yielding a cumulative 30‐day mortality of 142/618 (23.0%).
Of those discharged alive, 95/499 (19.0%) were readmitted within 30 days and 57/499 (11.4%) presented to the ED within this period (Table 2). The 57 patients who presented to the ED accounted for 68 total visits (48 with one visit, eight with two visits, and one with four visits). The 95 patients who were readmitted totaled 101 readmissions (89 with one readmission and six with two readmissions).
TABLE 2.
Care trajectory within 30‐days post‐tracheostomy (discharged alive, n = 499).
| Care pathway | TR (n, %) | RESP (n, %) | NTR (n, %) | Total (n, % of discharged) |
|---|---|---|---|---|
| No ED visit or readmission | — | — | — | 383 (76.8%) |
| ED visit only (no readmission) | 8 (38.1%) | 1 (4.8%) | 12 (57.1%) | 21 (4.2%) |
| ED visit → readmission | 4 (11.1%) | 8 (22.2%) | 24 (66.7%) | 36 (7.2%) |
| Direct readmission (no ED visit) | 8 (13.6%) | 10 (17.0%) | 41 (69.5%) | 59 (11.8%) |
| Total ED visits | 12 (21.1%) | 9 (15.8%) | 36 (63.2%) | 57 (11.4%) |
| Total readmissions | 12 (12.6%) | 18 (18.9%) | 65 (68.4%) | 95 (19%) |
Note: Only the initial event per patient was counted. Patients with multiple ED visits or readmissions were categorized according to their first occurrence; ED visits directly resulted in readmission were classified under “ED visit → readmission.”
The 30‐day rate of all‐cause ED visits was 57/499 (11.4%), totaling 68 revisits. Among ED revisits, other non‐tracheostomy related complications accounted for 39/68 (57.4%), followed by tracheostomy‐related complications (17/68, 25.0%) and respiratory‐related complications (12/68, 17.7%).
The 30‐day rate of all‐cause readmissions was 95/499 (19.0%), totaling 101 readmissions. Among total readmissions, other non‐tracheostomy‐related complications accounted for 69/101 (68.3%), followed by respiratory‐related complications (20/101, 19.8%) and tracheostomy‐related complications (12/101, 11.9%).
When restricted to initial revisits only, patterns were similar: non‐tracheostomy‐related complications remained the most common cause of both ED visits (36/57, 63.2%) and readmissions (65/95, 68.4%). Respiratory‐related complications were less frequent causes of ED revisits (9/57, 15.8%), while tracheostomy‐related complications were the least common cause of readmission (12/95, 12.6%).
3.3. Care Trajectory
Among patients presenting to the ED, 36/57 (63.2%) were subsequently readmitted (Table 2), representing 36/95 (37.9%) of all readmissions. Most readmissions occurred as direct admissions rather than through the ED (59/95, 62.1%), indicating the majority originated elsewhere.
3.4. Timing of Events
A total of 169 health care touch point events (101 readmissions and 68 ED visits) occurred within 30 days. Patients experiencing multiple revisits were accounted for, with each subsequent visit counted as a separate event in analysis. The median time to revisit after discharge was 9.0 days overall (8.0 days for ED visits and 9.0 days for readmissions; Figure 1). Among ED presentations, tracheostomy‐related complications occurred earliest (median 5.0 days), whereas respiratory‐related and non‐tracheostomy‐related complications occurred later (median 9.5 and 10.0 days, respectively; Figure 2a). A similar pattern was observed for readmissions: tracheostomy‐related complications occurred earlier (median 5.5 days), compared with respiratory‐related and non‐tracheostomy‐related causes (median 10.0 days; Figure 2b).
FIGURE 1.
Distribution of all‐cause revisits 30‐days postdischarge from tracheostomy. Median postdischarge day for all readmissions was 9.0 (tracheostomy‐related complications: 5.5 days; airway‐related complications: 10.0 days; non‐tracheostomy complications: 10.0 days). Median time to ED visit was 8.0 days (tracheostomy‐related: 5.0 days; airway‐related: 9.5 days; non‐tracheostomy: 10.0 days). Patients with multiple visits may be represented more than once. Day 0 corresponds to the day of discharge.
FIGURE 2.
Distribution of readmissions (a) and emergency department visits (b) by day and association with tracheostomy following hospital discharge. Each visit is classified into one of three mutually exclusive categories: Tracheostomy‐related (TR), other airway‐related (RESP), or non‐tracheostomy‐related (NTR) complications. Day 0 corresponds to the date of discharge.
3.5. Risk Factors
CVD was a predictor of in‐hospital mortality; whereas gastrostomy dependence was protective (p = 0.012, p < 0.001; Table 3).
TABLE 3.
In‐hospital mortality: Cardiovascular disease as a risk factor and gastrostomy dependence as a protective factor.
| Total (n = 618) | In‐hospital death (n = 119) | Survived to discharge (n = 499) | P | |
|---|---|---|---|---|
| Comorbidities a | ||||
| CVD | 132 (21.4%) | 36 (30.3%) | 96 (19.2%) | 0.0121* , b |
| Technology dependence a | ||||
| Gastrostomy dependence | 213 (34.5%) | 20 (16.8%) | 193 (38.7%) | < 0.001* , b |
Note: Compares survivors of the index hospitalization to those who died prior to discharge. Only variables with significant associations are shown.
Abbreviation: CVD, cardiovascular disease.
Comorbidities and technologies dependencies were defined by the first three digits of the International Classification of Diseases, Tenth Revision (ICD‐10) code.
Indicates Fisher's exact p‐value.
p < 0.05 denotes statistical significance.
Revision tracheostomy (SHR 2.93, 95% CI: 1.04–8.23, p = 0.042), emergent indication (SHR 2.81, 95% CI: 1.53–5.19, p < 0.001), and discharge to home‐ or community‐based care (SHR 2.05, 95% CI: 1.14–3.67, p = 0.017) were associated with increased risk of ED revisit (Table 4a). Speaking a primary language other than English showed a marginal association with increased ED revisit risk (SHR 1.96, 95% CI: 1.0–3.83, p = 0.051). Longer initial hospitalization was protective, as length of stay > 30 days was associated with lower risk of ED revisit (31–90 days: SHR 0.33, 95% CI: 0.13–0.82, p = 0.017; > 90 days: SHR 0.26, 95% CI: 0.07–0.94, p = 0.039). No comorbidities or technology dependencies were significantly associated with ED revisit.
TABLE 4a.
Significant demographic and preoperative risk factors of emergency department visit following adult tracheostomy (n = 618).
| Index admission number (%) | ED visit number (% readmission within category) | No ED visit number (% readmission within category) | Hazards ratio (95% CI) | P | |
|---|---|---|---|---|---|
| Primary language | |||||
| English a | 559 (90.5%) | 47 (8.4%) | 512 (91.6%) | — | |
| Spanish/other | 59 (9.5%) | 10 (16.9%) | 49 (83.1%) | 1.96 (1.0–3.83) | 0.051 c , * |
| Disposition b | 0.035 c , * | ||||
| Facility‐based care a | 268 (43.4%) | 26 (9.7%) | 242 (90.3%) | — | |
| Transitional/step‐down care | 127 (20.6%) | 12 (9.4%) | 115 (90.6%) | 0.99 (0.50–1.96) | 0.971c |
| Home‐ and community‐based care | 102 (16.5%) | 19 (18.6%) | 83 (81.4%) | 2.05 (1.14–3.67) | 0.017 c , * |
| Indication for tracheostomy | |||||
| Elective a | 557 (90.1%) | 44 (7.9%) | 513 (92.1%) | — | |
| Emergent | 61 (9.9%) | 13 (21.3%) | 48 (78.7%) | 2.81 (1.53–5.19) | < 0.001 c , * |
| Procedure type | 0.0414 c , * | ||||
| Tracheostomy a | 601 (97.2%) | 51 (8.5%) | 550 (91.5%) | — | |
| Revision‐tracheostomy | 17 (2.8%) | 6 (35.3%) | 11 (64.7%) | 2.93 (1.04–8.23) | |
| Length of stay | 0.090c | ||||
| < 10 days a | 19 (3.1%) | 5 (26.3%) | 14 (73.7%) | — | |
| 10–30 days | 233 (37.7%) | 23 (9.9%) | 210 (90.1%) | 0.41 (0.17–1.03) | 0.059c |
| 31–90 days | 304 (49.2%) | 25 (8.2%) | 279 (91.8%) | 0.33 (0.13–0.82) | 0.017 c , * |
| > 90 days | 62 (10.0%) | 4 (6.5%) | 58 (93.5%) | 0.26 (0.07–0.94) | 0.039 c , * |
Note: Univariate subdistribution hazard ratios (HR) with 95% confidence intervals (CI) assess associations between patient characteristics and ED visits within 30 days of discharge. ED visits and readmissions are mutually exclusive. Only variables with significant associations are shown.
Reference group.
Facility‐based care (LTAC, SNF, ECF); transitional/step‐down care (temporary/intermediate facilities); home‐ & community‐based care (home ± services, hospice, shelter).
Fisher's exact p‐value.
p < 0.05 denotes statistical significance.
Ventriculoperitoneal shunt presence (SHR 3.76, 95% CI: 1.91–7.38, p < 0.001) and gastrostomy dependence (SHR 1.78, 95% CI: 1.19–2.67, p = 0.005) were associated with increased risk of readmission, whereas Hispanic or Latino ethnicity (SHR 0.32, 95% CI: 0.10–0.99, p = 0.049) and initial hospitalization > 90 days (SHR 0.27, 95% CI: 0.09–0.85, p = 0.025) were associated with decreased risk (Table 4b).
TABLE 4b.
Significant demographic and preoperative risk factors of readmission following adult tracheostomy (n = 618).
| Index admission number (%) | Readmission number (% readmission within category) | No readmission number (% readmission within category) | Hazards ratio (95% CI) | P | |
|---|---|---|---|---|---|
| Ethnicity | 0.134c | ||||
| Not Hispanic or Latino a | 556 (90.0%) | 91 (16.4%) | 465 (83.6%) | — | |
| Hispanic or Latino | 50 (8.1%) | 3 (6.0%) | 47 (94.0%) | 0.32 (0.10–0.99) | 0.049 c , * |
| Other | 12 (1.9%) | 1 (8.3%) | 11 (91.7%) | 0.68 (0.10–4.65) | 0.693c |
| Length of stay | 0.082c | ||||
| < 10 days a | 19 (3.1%) | 6 (31.6%) | 13 (68.4%) | — | |
| 10–30 days | 233 (37.7%) | 30 (12.9%) | 203 (87.1%) | 0.46 (0.21–1.03) | 0.060c |
| 31–90 days | 304 (49.2%) | 54 (17.8%) | 250 (82.2%) | 0.63 (0.29–1.34) | 0.226c |
| > 90 days | 62 (10.0%) | 5 (8.1%) | 57 (92.0%) | 0.27 (0.09–0.85) | 0.025 c , * |
| Technology dependence b | |||||
| Gastrostomy dependence | 213 (34.5%) | 48 (22.5%) | 165 (77.5%) | 1.78 (1.19–2.67) | 0.005 c , * |
| Ventriculoperitoneal shunt | 15 (2.4%) | 8 (53.3%) | 7 (46.7%) | 3.76 (1.91–7.38) | < 0.001 c , * |
Note: Univariate subdistribution hazard ratios (HR) with 95% confidence intervals (CI) assess associations between patient characteristics and readmissions within 30 days of discharge. ED visits and readmissions are mutually exclusive. Only variables with significant associations are shown.
Reference group.
Technology dependencies were defined by the first three digits of the International Classification of Diseases, Tenth Revision (ICD‐10) code.
Indicates Fisher's exact p‐value.
p < 0.05 denotes statistical significance.
4. Discussion
The primary objectives of this study were to determine the rates and primary causes of return visits in adults post‐tracheostomy and to identify specific patient characteristics and risk factors contributing to in‐hospital mortality and healthcare utilization postdischarge. These findings can play an important role in counseling for patients, families, and healthcare providers.
4.1. Quantification of Mortality and Revisit
High in‐hospital mortality underscores the medical complexity of this population, with 19.3% of patients dying prior to discharge. Although patients exhibited a mean of two comorbid conditions or technology dependencies, mortality was not associated with a greater number of these factors, suggesting outcomes may be more strongly driven by illness severity than cumulative comorbidity burden.
Considering mortality up to 30 days of discharge post‐tracheostomy, the cumulative mortality of our cohort was 23.0%. Heightened mortality is characteristic of the adult tracheostomy population. In recent agreement with our results, Mehta et al. showed 30‐day mortality after tracheostomy as 22.1% [14].
Among the 499 patients discharged alive, 11.4% of patients presented to the ED within 30 days and 19.0% were readmitted within this time. Notably, a tracheostomy‐specific rounding team was implemented at our institution during the duration of this study; however, no changes in outcome measures were observed. Tracheostomy patients continue to exhibit a high frequency of postdischarge healthcare utilization.
Non‐tracheostomy‐related complications accounted for 66.5% of revisits within 30 days. Tracheostomy‐related complications were less frequent—accounting for 21.1% of initial ED revisits and 12.6% of initial readmissions. However, because ED visits occurred less often overall, the patient‐level risk of presenting to the ED or experiencing readmission for a tracheostomy‐related complication was identical at 2.4%.
4.2. Care Trajectory
Fewer patients presented to the ED than were readmitted. This paradox may be attributed to the large area served by the University of Nebraska system, in which the closest ED may infrequently be within the system. As this study is among the first to characterize ED revisits post‐tracheostomy, it is important to acknowledge any factors that may contribute to trends shown in our cohort. Readmissions that came directly through the ED only accounted for 37.9% of total readmissions, indicating in most cases that patients originated outside of the hospital system. This suggests fragmentation in the postdischarge care plan as many patients may re‐present to external hospitals. Standardized education and counseling across all discharge destinations can offset the cost and utilization burden placed on the healthcare system.
Notably, 63.2% of patients who presented to the ED within 30 days of discharge required care escalation necessitating readmission. In agreement with our results, Shay et al. found for pediatric patients, 59.2% of ED visits post‐tracheostomy resulted in readmission [15]. This suggests ED presentation often reflects significant care needs rather than minor concerns, underscoring the value of early identification of high‐risk patients.
4.3. Timing of Events
The median postdischarge date for all health care touch point events was 9.0 days; 8.0 days for ED visits, and 9.0 days for readmission, respectively. Tracheostomy‐related complications presented earlier, with a median revisit timing of 5.0 days for ED visits and 5.5 days for readmission. The earlier revisit rate for tracheostomy‐related complications is justified by the acuity of post‐tracheostomy care.
4.4. Risk Factors
CVD was a significant predictor of in‐hospital mortality, while gastrostomy tube dependence was protective. The leading cause of death nationally, CVD has been shown to be a pertinent risk factor for mortality in tracheostomy patients in prior literature as well [16, 17]. The protective mechanism of gastrostomy tube presence may be attributed to enhanced nutritional support via enteral feeding, which can mitigate the chronic malnutrition observed in tracheostomy patients, potentially reducing mortality [18].
Primary language other than English, discharge to home‐ or community‐based care, emergent indication, and revision‐tracheostomy increased risk of ED revisit. Initial length of stay > 30 days decreased risk of ED revisit. Language‐associated risk factors may uniquely manifest in the post‐tracheostomy setting, given the high acuity of this patient population and the necessity of proper education and communication throughout discharge. Home healthcare can reduce overall healthcare expenditures, but for medically complex patients, such as post‐tracheostomy, reliance on patient, and caregiver education may increase postdischarge hospital utilization [19]. Tracheostomies performed under emergent conditions have been shown to have a threefold increase in the likelihood of postoperative complications, which aligns with our revisit results [20]. Patients undergoing emergent tracheostomies and revision‐\ tracheostomies likely represent a higher acuity group.
Gastrostomy dependence and VP shunt presence increased risk of readmission. Hispanic or Latino ethnicity and initial hospitalization > 90 days decreased risk of readmission. While not previously identified as risk factors, gastrostomy tube dependence and VP shunt presence likely indicate heightened medical complexity, emphasizing comprehensive consideration of patient factors during discharge planning. The protective effect of Hispanic or Latino ethnicity aligns with the ‘Hispanic Paradox’—a phenomenon in which these populations exhibit better‐than‐expected health outcomes despite social risk factors and higher rates of chronic conditions [15]. Familism and community support may contribute to improved adherence and early recognition of complications.
The protective effect of longer initial hospital stay against ED revisits and readmission is paradoxical relative to prior literature [4, 9, 15]. To our knowledge, this is the first study to report such inverse relationship. The difficulty of postdischarge placement for tracheostomy patients in our region may contribute to prolonged hospitalization, in turn allowing for increased time allowance for education, management of comorbidities, and optimization of tracheostomy care.
4.5. Strengths
This study has several notable strengths. To our knowledge, it is the first to distinctly stratify risk factors for both ED revisits and readmissions in adult tracheostomy patients. Use of Fine & Gray's proportional hazards modeling, rather than standard descriptive statistics, allowed us to account for the high mortality in this cohort. Additionally, the inclusion of a broad range of patient characteristics, demographics, comorbidities, and medical complexities enabled cross‐variable analysis to assess their cumulative impact on patient outcomes.
4.6. Limitations
This study has several important limitations. Follow‐up data were constrained to a single tertiary care hospital, likely underreporting revisit rates, biasing associations toward the null—underscoring the mortality of this patient population. The medical complexity and comorbidities of our cohort may not be representative of patients at other institutions. Additionally, use of ICD‐10 billing codes to identify tracheostomy, patient demographics, and comorbidities is subject to misclassification and does not capture illness severity; however, these codes are commonly used throughout the literature.
5. Conclusion
This study retrospectively observed an in‐hospital post‐tracheostomy mortality rate of 19.3%, cumulative 30‐day postdischarge mortality of 23.0%, a 30‐day all‐cause ED revisit rate of 11.4%, and readmission rate of 19.0%. In‐hospital mortality was predicted by CVD and protected by gastrostomy tube presence. ED revisit was predicted by primary language other than English, discharge to home‐ or community‐based care, emergent indication, and revision‐tracheostomy, and was protected by length of stay > 30 days. Readmission was predicted by gastrostomy dependence and VP shunt presence and was protected by Hispanic or Latino ethnicity and length of stay > 90 days. This information provides insight into preoperative risk factors for mortality and postdischarge hospital utilization, which can better inform pre‐tracheostomy counseling and care trajectory. Incorporating these findings into realistic discharge planning and education may improve outcomes while reducing healthcare costs.
Funding
The authors report no funding for the conduct of this study. M.G. received a stipend through the University of Nebraska Medical Center College of Medicine Medical Student Summer Research Program (MSSRP).
Conflicts of Interest
The authors declare no conflicts of interest.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
References
- 1. Abril M. K., Berkowitz D. M., Chen Y., Waller L. A., Martin G. S., and Kempker J. A., “The Epidemiology of Adult Tracheostomy in the United States 2002–2017: A Serial Cross‐Sectional Study,” Critical Care Explorations 3, no. 9 (2021): e0523, 10.1097/CCE.0000000000000523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Das P., Zhu H., Shah R. K., Roberson D. W., Berry J., and Skinner M. L., “Tracheotomy‐Related Catastrophic Events: Results of a National Survey,” Laryngoscope 122, no. 1 (2012): 30–37, 10.1002/lary.22453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Mitchell R. B., Hussey H. M., Setzen G., et al., “Clinical Consensus Statement: Tracheostomy Care,” Otolaryngology‐Head and Neck Surgery 148, no. 1 (2013): 6–20, 10.1177/0194599812460376. [DOI] [PubMed] [Google Scholar]
- 4. Spataro E., Durakovic N., Kallogjeri D., and Nussenbaum B., “Complications and 30‐Day Hospital Readmission Rates of Patients Undergoing Tracheostomy: A Prospective Analysis,” Laryngoscope 127, no. 12 (2017): 2746–2753, 10.1002/lary.26668. [DOI] [PubMed] [Google Scholar]
- 5. Chew L., Su‐Velez B. M., Miller J. E., and West A. N., “30‐Day Readmission Rates, Diagnoses, and Risk Factors Following Pediatric Airway Surgery,” International Journal of Pediatric Otorhinolaryngology 136 (2020): 110141, 10.1016/j.ijporl.2020.110141. [DOI] [PubMed] [Google Scholar]
- 6. Levy L., Smiley A., and Latifi R., “Mortality Risk Factors in Patients Admitted With the Primary Diagnosis of Tracheostomy Complications: An Analysis of 8026 Patients,” International Journal of Environmental Research and Public Health 19, no. 15 (2022): 9031, 10.3390/ijerph19159031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Office of the Legislative Counsel , The Patient Protection and Affordable Care Act of 2010 (42 USC 18001) (US Government Publishing Office, 2010), https://www.cms.gov/regulations‐and‐guidance/legislation/legislativeupdate/downloads/ppaca.pdf?utm_source=chatgpt.com. [Google Scholar]
- 8. CMS , “Hospital Readmissions Reduction Program,” accessed December 4, 2025, https://www.cms.gov/medicare/payment/prospective‐payment‐systems/acute‐inpatient‐pps/hospital‐readmissions‐reduction‐program‐hrrp.
- 9. Centers for Medicare & Medicaid Services (CMS) andUS Department of Health & Human Services (DHHS) , “Revision of Common Working File (CWF) Editing for Same‐Day, Same‐Provider Acute Care Readmissions—Transmittal 266,” 2004, https://www.cms.gov/regulations‐and‐guidance/guidance/transmittals/downloads/r266cp.pdf.
- 10. Regence , “Inpatient Hospital Readmissions,” accessed December 4, 2025, https://www.regence.com/provider/library/policies‐guidelines/reimbursement‐policy/inpatient‐hospital‐readmissions?utm_source=chatgpt.com.
- 11. Bailey M. K., Weiss A. J., Barrett M. L., and Jiang H. J., “Characteristics of 30‐Day All‐Cause Hospital Readmissions, 2010–2016,” in Healthcare Cost and Utilization Project (HCUP) Statistical Briefs (Agency for Healthcare Research and Quality (US), 2006). [PubMed] [Google Scholar]
- 12. Silver S. A., Harel Z., McArthur E., et al., “30‐Day Readmissions After an Acute Kidney Injury Hospitalization,” American Journal of Medicine 130, no. 2 (2017): 163–172.e4, 10.1016/j.amjmed.2016.09.016. [DOI] [PubMed] [Google Scholar]
- 13. Fine J. P. and Gray R. J., “A Proportional Hazards Model for the Subdistribution of a Competing Risk,” Journal of the American Statistical Association 94, no. 446 (1999): 496–509, 10.2307/2670170. [DOI] [Google Scholar]
- 14. Mehta A. B., Walkey A. J., Curran‐Everett D., and Douglas I. S., “One‐Year Outcomes Following Tracheostomy for Acute Respiratory Failure,” Critical Care Medicine 47, no. 11 (2019): 1572–1581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Shay S., Shapiro N. L., and Bhattacharyya N., “Revisits After Pediatric Tracheotomy: Airway Concerns Result in Returns,” International Journal of Pediatric Otorhinolaryngology 104 (2018): 5–9, 10.1016/j.ijporl.2017.10.021. [DOI] [PubMed] [Google Scholar]
- 16. NCHS , “Leading Causes of Death,” accessed December 4, 2025, https://www.cdc.gov/nchs/fastats/leading‐causes‐of‐death.htm.
- 17. Kashlan K. N., Williams A. M., Chang S. S., Yaremchuk K. L., and Mayerhoff R., “Analysis of Patient Factors Associated With 30‐Day Mortality After Tracheostomy,” Laryngoscope 129, no. 4 (2019): 847–851, 10.1002/lary.27345. [DOI] [PubMed] [Google Scholar]
- 18. Henningfeld J., Lang C., Erato G., Silverman A. H., and Goday P. S., “Feeding Disorders in Children With Tracheostomy Tubes,” Nutrition in Clinical Practice 36, no. 3 (2021): 689–695, 10.1002/ncp.10551. [DOI] [PubMed] [Google Scholar]
- 19. Beams D. R., Chorney S. R., Kou Y. F., Teplitzky T. B., Wynings E. M., and Johnson R. F., “Emergency Department Visits and Hospitalizations After Pediatric Tracheostomy,” Laryngoscope 133, no. 8 (2023): 2018–2024, 10.1002/lary.30416. [DOI] [PubMed] [Google Scholar]
- 20. Natasha N., Khurshied S., Zahid M. A., et al., “Comparison Between Complications of Elective and Emergency Tracheostomies,” Cureus 15, no. 11 (2023): e48612, 10.7759/cureus.48612. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.