Impact of tracheostomy on clinical outcomes in ventilated patients with severe pneumonia: a propensity-matched cohort study

Hayoung Seong; Hyojin Jang; Wanho Yoo; Saerom Kim; Soo Han Kim; Kwangha Lee

doi:10.3904/kjim.2024.248

Korean J Intern Med > Volume 40(2); 2025 > Article

Seong, Jang, Yoo, Kim, Kim, and Lee: Impact of tracheostomy on clinical outcomes in ventilated patients with severe pneumonia: a propensity-matched cohort study

Original Article

Pulmonology

The Korean Journal of Internal Medicine 2025;40(2):286-298.

Published online: March 1, 2025

DOI: https://doi.org/10.3904/kjim.2024.248

Impact of tracheostomy on clinical outcomes in ventilated patients with severe pneumonia: a propensity-matched cohort study

Hayoung Seong^1,^2,³, Hyojin Jang^1,^2,³, Wanho Yoo^1,^2,³, Saerom Kim^1,^2,³, Soo Han Kim^1,^2,³, Kwangha Lee^1,^2,³

¹Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Internal Medicine, Pusan National University Hospital, Busan, Korea

²Department of Internal Medicine, School of Medicine, Pusan National University, Busan, Korea

³Biomedical Research Institute, Pusan National University Hospital, Busan, Korea

Correspondence to: Kwangha Lee, M.D., Ph.D., Department of Internal Medicine, Pusan National University School of Medicine, 179 Gudeok-ro, Seo-gu, Busan 49241, Korea, Tel: +82-51-240-7743, Fax: +82-51-254-3127, E-mail: jubilate@pusan.ac.kr, https://orcid.org/0000-0001-9878-201X

Received July 16, 2024 Revised September 11, 2024 Accepted September 23, 2024

(open-access):

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background/Aims

Tracheostomy is a crucial intervention for severe pneumonia patients requiring prolonged mechanical ventilation (MV). However, debate persists regarding the influence of tracheostomy timing and performance on long-term survival outcomes. This study utilized propensity score matching to assess the impact of tracheostomy timing and performance on patient survival outcomes.

Methods

A retrospective observational study employing propensity score matching was conducted of respiratory intensive care unit (ICU) patients who underwent prolonged acute MV due to severe pneumonia from 2008 to 2023. The primary outcome was the 90-day cumulative mortality rate, with secondary outcomes including ICU medical resource utilization rates.

Results

Out of 1,078 patients, 545 underwent tracheostomy with a median timing of 7 days. The tracheostomy group exhibited lower 90-day cumulative mortality and a higher survival probability (hazard ratio [HR] 0.52, 95% confidence interval [CI] 0.43–0.63) than the no-tracheostomy group. The tracheostomy group had higher ICU medical resource utilization rates and medical expenditures. The early tracheostomy group (≤ 7 days) had lower ICU medical resource utilization rates and medical expenditures than the late tracheostomy group (> 7 days). However, there were no significant differences in the 90-day cumulative mortality rate and survival probability based on tracheostomy timing (HR 0.94, 95% CI 0.70–1.28).

Conclusions

Tracheostomy in patients with severe pneumonia requiring prolonged MV significantly reduced the 90-day mortality rate, and early tracheostomy may offer additional benefits for resource utilization efficiency. These findings underscore the importance of considering tracheostomy timing in optimizing patient outcomes and healthcare resource allocation.

Keywords: Tracheostomy; Pneumonia; Mechanical ventilation; Mortality; Propensity score

Graphical abstract

INTRODUCTION

The rising incidence of severe pneumonia requiring mechanical ventilation (MV) in intensive care units (ICUs) often requires long-term ventilator care and substantial utilization of medical resources [1–3]. However, sustaining long-term ventilator care through extended translaryngeal intubation poses risks of laryngeal and tracheal damage and elevates the likelihood of ventilator-associated pneumonia (VAP) [4]. Tracheostomy is a useful option for patients requiring prolonged MV, offering benefits such as decreased airway dead space and resistance, improved patient comfort, reduced sedative usage, and a lower incidence of nosocomial pneumonia [5,6].

Although numerous studies have demonstrated the favorable impact of tracheostomy on hospital outcomes for patients requiring prolonged MV, its effects on long-term outcomes remain controversial [7–9]. Additionally, the criteria for determining the timing of tracheostomy, whether early or late, and its subsequent influence on survival outcomes remain subjects of ongoing debate [5,10,11]. These controversies may be due to significant heterogeneity related to the specific diseases studied, the timing of tracheostomy, characteristics of the ICU setting, and the extent of medical resource utilization.

In Korea, severe pneumonia is one of the most common conditions among critically ill patients and imposes a significant burden on medical resources [12]. Despite this, research on the effect of tracheostomy on patient outcomes remains limited. We hypothesize that the timing and performance of tracheostomy in patients with severe pneumonia requiring prolonged MV will influence both patient outcomes and medical resource utilization. Therefore, this study aimed to investigate the impact of tracheostomy on survival outcomes and medical resource utilization in patients requiring MV due to severe pneumonia, using propensity score matching (PSM) to overcome the inherent limitations of retrospective studies.

METHODS

Study design and patient selection

This study was conducted in the 12-bed adult respiratory ICU of a 1,200-bed university-affiliated tertiary care hospital in Busan, Korea. The ICU maintained a nurse-to-bed ratio of 1:3 and was equipped with a hemodynamic monitoring system for each patient. Standard care guidelines for critically ill patients were followed [13].

This retrospective observational cohort study screened all consecutive patients requiring MV treatment in the respiratory ICU from March 1, 2008 to May 31, 2023. Inclusion criteria were patients diagnosed with acute respiratory failure due to community-acquired or hospital-acquired pneumonia [14,15], who also received MV for more than 96 hours (defined as prolonged acute MV, PAMV) [16]. Exclusion criteria included patients requiring MV for conditions other than pneumonia, those who had previously undergone tracheostomies, individuals with pre-existing directives that limited life-sustaining treatments, and individuals under the age of 18. For patients with multiple ICU admissions, data from only their first ICU admission were collected.

The primary outcome was the 90-day cumulative mortality rate. This included all cases of death that occurred within 90 days from the start of MV. Secondary outcomes included discharge outcomes, total medical expenditures during the hospital stay, and medical resource utilization rates in the ICU. Discharge outcomes were assessed based on the discharge destination (e.g., death, home, or step-down hospital) and decannulation rates before hospital discharge. Total medical expenditures were calculated as the cost of all medical resources used during the patient’s hospital stay. Health ratio indicators such as the bed occupancy rate (BOR), bed turnover rate (BTR), average length of stay (LOS), and turnover interval (TI) were used to evaluate the efficiency of medical resource utilization [17]. These indicators have been evaluated as factors impacting medical expenditures, playing an important role in understanding how effective management of medical resources contributes to reducing medical costs [18,19].

Data collection

Demographic and clinical data, including age, sex, body mass index (BMI), duration of MV, LOS in the ICU and hospital, and mortality in the ICU and hospital, were collected retrospectively from patients’ electronic medical records. The severity of illness was measured using the Acute Physiology and Chronic Health Evaluation (APACHE) II score, and accompanying organ failure was assessed according to the Sequential Organ Failure Assessment (SOFA) score [20,21]. The APACHE II and SOFA scores were calculated from laboratory and clinical data obtained in the first 24 hours of intubation. Charlson’s comorbidity index (CCI) was evaluated to determine concurrent comorbidities before ICU admission [22].

All tracheostomies were surgical and were determined by the attending physician and then referred to a thoracic surgeon [23]. Time to tracheostomy was defined as the duration between initiating MV care and tracheostomy. According to tracheostomy guidelines from the Korean Bronchoesophagological Society, early tracheostomy (ET) was defined as a tracheostomy within 7 days after tracheal intubation, and late tracheostomy (LT) was defined as a tracheostomy after 7 days [24]. The 90-day cumulative mortality rate was calculated from the first day of MV. For surviving patients discharged from the hospital, the mortality rate was obtained from the National Health Insurance Service Database [25]. Total medical expenditures for all medical resources used during the hospitalization were retrieved with the permission of the Institutional Review Board (IRB). All costs are shown in United States Dollars (USD) at an exchange rate of 1 USD = 1,357.30 Korean won (exchange rate on November 1, 2023).

Statistical method

Continuous variables were presented as the mean ± standard deviation or median (interquartile range) and compared using the Student t-test or Mann–Whitney U test, as applicable. Categorical variables were presented as numbers (percentages) and compared using the chi-square or Fisher’s exact test, as applicable. Survival analysis was performed to compare the survival time between the two groups, and the difference in Kaplan–Meier curves between the groups was assessed using the log-rank test.

To minimize potential selection bias due to the absence of randomization between the tracheostomy and no-tracheostomy groups, as well as between the ET and LT groups, PSM was performed using 1-to-1 nearest-neighbor matching without replacement. Propensity scores were calculated using the following covariates; age, sex, BMI, comorbidities, APACHE II score, SOFA score, CCI, and use of adjunctive therapy (vasopressor, neuromuscular blocking agents, and renal replacement therapy). The propensity score was estimated via logistic regression. In this process, all patients were categorized based on the propensity scores, and each group was matched within 0.2 standard deviations of the logit propensity score [26]. Subgroup analysis after PSM was conducted based on age, sex, BMI, CCI, APACHE II score, SOFA score, and ICU admission year to identify factors affecting the occurrence of 90-day cumulative mortality.

Health ratio indicators were collected using the following variables: (1) number of active beds, which refers to the number of functional beds for each hospital year; (2) active bed days, which refers to the number of functional beds in the hospital for a given period, usually 1 year, calculated as the number of active beds multiplied by 365 days; (3) number of discharges in a given year; and (4) occupied bed days, which refers to the sum of the total number of days all admitted patients spent in the hospital for a given year. Using the above variables, four ratios or indices were computed as follows:

BOR=Occupied bed days (total inpatient days)Active bed days×100%Average LOS=Occupied bed days (total inpatient days)Number of dischargesBTR=Number of discharges in 1 yearActive bedsTI=365BTR-Average LOS

All tests were two-tailed and p values of < 0.05 were considered statistically significant. Statistical analyses were performed using the R statistical language (version 4.3.2; R Core Team, 2023) and additional packages (MatchIt).

Ethics statement

Approval of the IRB of Pusan National University Hospital was obtained, and the need for patient consent was waived because of the retrospective and observational nature of the study (IRB No. 2403-009-137). Study procedures followed the Declaration of Helsinki of 1975 and the Strengthening the Reporting of Observational Studies in Epidemiology guidelines [27,28].

RESULTS

During the study period, 3,296 patients underwent MV in the respiratory ICU, with 1,675 (50.8%) requiring PAMV. In total, 277 patients who received MV for reasons other than pneumonia, 241 patients who had limited life-sustaining treatments, and 79 who had a tracheostomy before ICU admission were excluded. The final study cohort comprised 1,078 patients (Fig. 1). In the entire population, the ICU, hospital, and 90-day cumulative mortality rates were 30.1% (n = 325), 36.1% (n = 389), and 47.6% (n = 513), respectively. In total, 545 patients (50.6%) underwent tracheostomy during the ICU stay, and the median time to tracheostomy was 7 days (range, 0–44 days) (Fig. 2).

Table 1 summarizes the baseline characteristics of the study cohort before and after PSM. At ICU admission, the tracheostomy group was older and had a higher CCI than the no-tracheostomy group. In PSM analysis, 914 patients with and without tracheostomy were matched (n = 457 per group). The selected covariates did not significantly differ between the matched groups (all standardized mean differences < 0.2), with histograms of the logit of propensity scores demonstrating the high quality of the matching process (Supplementary Fig. 1). The tracheostomy group experienced significantly longer durations of MV as well as ICU and hospital LOS than the no-tracheostomy group. Furthermore, the tracheostomy group had significantly lower ICU, hospital, and 90-day cumulative mortality rates. However, the tracheostomy group had a significantly higher transfer rate to step-down hospitals compared to the no-tracheostomy group (55.1% vs. 28.0%, p < 0.001). Additionally, the tracheostomy group had higher medical expenditures, BOR, and average LOS than the no-tracheostomy group (Table 2).

Survival analysis indicated that over 90 days, the tracheostomy group had a higher survival probability than the no-tracheostomy group (hazard ratio [HR] 0.52, 95% confidence interval [CI] 0.43–0.63) (Fig. 3A). Subgroup analysis showed that in all cases, the tracheostomy group had a lower 90-day cumulative mortality rate than the no-tracheostomy group (Fig. 4A).

In the tracheostomy group, 275 patients were classified into the ET group, while 270 patients were classified into the LT group. After PSM, 205 patients in each of the ET and LT groups were matched. Table 3 summarizes the baseline characteristics of patients classified according to the timing of tracheostomy. The selected covariates did not significantly differ between the matched ET and LT groups (all standardized mean differences < 0.2), with histograms of the logit of propensity scores demonstrating the high quality of the matching process (Supplementary Fig. 2).

The ET group had significantly shorter durations of MV, ICU LOS, and hospital LOS and incurred lower medical expenditures than the LT group (Table 4). However, there were no significant differences between the two groups in terms of discharge outcomes, decannulation rates, 90-day cumulative mortality rate, or survival (HR 0.94, 95% CI 0.70–1.28) (Fig. 3B). Additionally, the mortality rate did not significantly differ between the two groups in all subgroup analyses (Fig. 4B).

DISCUSSION

This study is the first investigation to assess the influence of tracheostomy on 90-day outcomes among PAMV patients with severe pneumonia, while also exploring the distinctions between ET and LT following Korean guidelines, using PSM analysis. Our findings provide critical care physicians with valuable insights into the role of tracheostomy in planning long-term strategies for patients with severe pneumonia who require prolonged MV.

A key aspect of our study was the examination of positive outcomes associated with tracheostomy using PSM analysis. Our findings revealed that patients who underwent tracheostomy had a lower 90-day mortality rate. This survival benefit was consistent across all subgroups, irrespective of factors such as age, sex, severity, comorbidities, and presence of organ failure at ICU admission. The improved outcomes may be attributable to various advantages of tracheostomy, including improvements in lung mechanics, easier oral hygiene management, reduced sedative usage promoting easier expectoration by patients, and a subsequent decrease in the risk of VAP [5,29,30]. However, patients who underwent tracheostomy have longer LOS, higher medical expenditures, and often require long-term care facilities for additional respiratory and swallowing rehabilitation [31,32]. Tracheostomy should be considered for patients who have difficulty being liberated from MV, as indicated by various clinical indicators [33,34]. Our findings also suggest that tracheostomy can play a crucial role in managing critically ill patients with acute respiratory failure due to severe pneumonia and highlight the need for appropriate allocation of medical resources during tracheostomy.

Previous studies produced varied conclusions regarding the survival benefits of tracheostomy in patients requiring prolonged MV [8,9,30,35]. By contrast, our study indicates that tracheostomy may offer notable benefits for patients with severe pneumonia requiring prolonged MV. These differences might be attributable to the heterogeneity in the enrolled subjects whose data were reported [5,36], which often included subgroups with diverse etiological conditions requiring MV and different ICU environments. Our study focused exclusively on patients with severe pneumonia in a respiratory ICU. By focusing on a more homogeneous patient population, the present study provides clearer insights into the benefits of tracheostomy for severe pneumonia patients receiving prolonged MV.

Although the timing of tracheostomy did not demonstrate a survival benefit according to our national guidelines [24], ET positively influenced ICU capacity by alleviating strain. This led to lower medical expenditures. Therefore, ET could help to optimize the clinical course and manage critical care resources [37]. Given the current state of critical care medicine in Korea, which is characterized by a chronic shortage of ICU beds relative to hospital size and a lack of nursing staff, full-time intensivists, and well-designed MV weaning protocols [38,39], the decision by attending physicians to perform ET may be advantageous for patients anticipated to require long-term ventilator care upon ICU admission. Despite potential adverse effects of tracheostomy, such as procedure-related complications and later cosmetic concerns [40,41], ET could offer significant benefits in terms of resource management and long-term patient care. However, our findings are based on a single institution; therefore, larger multi-center studies are needed to confirm the effectiveness of ET in terms of patient outcomes. Future research should also consider the cost-effectiveness and long-term benefits of ET across various patient populations and healthcare settings to provide a more comprehensive understanding of its impact.

The optimal timing of tracheostomy remains controversial, with criteria for ET and LT varying between studies [5,10]. Currently, there is no international guideline that clearly defines the distinction between ET and LT, which leads to inconsistencies in clinical practice and research. Although Korea has established criteria for ET and LT, it is uncertain if these criteria are applicable to ventilated patients with severe pneumonia. Therefore, extensive large-scale, multicenter studies are necessary to evaluate the performance and optimal timing of tracheostomy. Essential considerations in these studies should include defining patient populations likely to require prolonged MV, establishing clear definitions of ET and LT, and assessing adverse events associated with tracheostomy. These studies will help determine whether the performance and timing of tracheostomy can be universally applied across different patient populations and healthcare settings, as well as their impact on outcomes such as survival, MV duration, and complication rates. This understanding is critical for improving clinical decision-making and patient care.

According to historical studies, tracheostomy was performed in 9–13% of patients receiving MV [2,8]. One study reported that up to 34% of patients requiring MV for more than 48 hours underwent tracheostomy [42]. In our study, tracheostomy was performed in 16.5% of patients receiving MV and in 50.6% of those requiring MV for more than 96 hours, which is a higher rate than previously reported. This higher rate in our study may be explained by the fact that we focused on patients with severe respiratory failure due to severe pneumonia, and our patient population was older than in previous studies.

This study has several limitations. First, it is difficult to predict which patients will require prolonged MV or tracheostomy, and the subjects may be heterogeneous due to the diversity of underlying diseases and severity. We implemented PSM analysis to minimize potential selection bias related to the subjects and timing of tracheostomy. Second, this study was limited to data from a single institution, making it difficult to generalize the findings to the overall situation in South Korea. Third, we hypothesized that tracheostomy is associated with a decreased incidence of VAP, however, we could not definitively confirm the association between tracheostomy and VAP due to the retrospective design. Lastly, the clinical differences between community-acquired pneumonia and hospital-acquired pneumonia, as well as complications related to tracheostomy (such as bleeding, laryngeal damage, etc.), are important considerations in patient outcomes. However, the retrospective design of this study limited our ability to analyze these factors due to insufficient data accurately. Despite these limitations, our PSM analysis provides evidence that tracheostomy, when performed timely, can significantly reduce mortality and improve the efficiency of medical resource utilization, highlighting its value as a critical treatment strategy.

In conclusion, analysis of the effect of tracheostomy on outcomes of pneumonia patients showed that 90-day cumulative mortality was reduced. Additionally, the ET group had shorter durations of MV and LOS as well as lower medical expenditures than the LT group. To accurately assess the effect of tracheostomy on patient outcomes, large-scale, multi-center research tailored to the specific conditions of the healthcare environment is essential.

KEY MESSAGE

1. Tracheostomy reduces the 90-day cumulative mortality rate in severe pneumonia patients requiring prolonged MV.

2. ET has a positive impact on alleviating ICU capacity and reducing healthcare expenditures.

3. Considering tracheostomy may help manage resources and improve long-term patient outcomes, suggesting that it may be an important treatment strategy for severe pneumonia patients requiring prolonged MV.

Supplementary Information

kjim-2024-248-Supplementary-Fig-1.pdf

kjim-2024-248-Supplementary-Fig-2.pdf

Notes

CRedit authorship contributions

Hayoung Seong: methodology, investigation, data curation, formal analysis, software, writing - original draft, writing - review & editing, visualization; Hyojin Jang: investigation, validation; Wanho Yoo: investigation, validation; Saerom Kim: investigation, validation; Soo Han Kim: investigation, validation; Kwangha Lee: conceptualization, methodology, resources, investigation, software, writing - review & editing, supervision

Conflicts of interest

The authors disclose no conflicts.

Funding

None

Figure 1

Patient selection flow chart. MV, mechanical ventilation; ICU, intensive care unit. ^a)Included leukemia, cardiomyopathy, pulmonary edema, arrhythmia, diabetic ketoacidosis, and connective tissue disease.

Figure 2

Distribution of time to tracheostomy (from mechanical ventilator care to tracheostomy) (n = 545).

Figure 3

Kaplan–Meier survival curves of 90-day survival in (A) the matched tracheostomy and no-tracheostomy groups and (B) the matched early tracheostomy and late tracheostomy groups. HR, hazard ratio; CI, confidence interval.

Figure 4

Subgroup analysis of 90-day mortality in (A) the matched tracheostomy and no-tracheostomy groups and (B) the matched early tracheostomy and late tracheostomy groups. HR, hazard ratio; CI, confidence interval; BMI, body mass index; CCI, Charlson’s comorbidity index; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; NA, not available.

Table 1

Comparison of baseline characteristics of patients in the tracheostomy and no-tracheostomy groups before and after propensity score matching

Characteristic	Before matching			After matching

	Tracheostomy (n = 545)	No-tracheostomy (n = 533)	SMD	Tracheostomy (n = 457)	No-tracheostomy (n = 457)	SMD
Age (yr)	69.4 ± 12.8	66.7 ± 13.8	0.211	68.9 ± 12.6	68.3 ± 12.8	0.043

Sex (male)	389 (71.4)	377 (70.7)	0.006	327 (71.6)	326 (71.3)	0.002

BMI (kg/m²)	21.6 ± 3.7	22.1 ± 4.0	0.112	21.7 ± 3.6	22.0 ± 3.7	0.064

CCI	5.7 ± 3.0	5.1 ± 2.9	0.188	5.5 ± 2.9	5.4 ± 2.9	0.022

APACHE II score^a)	21.9 ± 6.8	21.3 ± 7.6	0.097	21.5 ± 6.7	21.7 ± 7.7	0.030

SOFA score^a)	7.9 ± 3.2	7.9 ± 3.7	0.004	7.9 ± 3.2	8.0 ± 3.7	0.033

Comorbidity

Pulmonologic	303 (55.6)	315 (59.1)	0.035	274 (60.0)	261 (57.1)	0.028

Hemato-oncologic	81 (14.9)	94 (17.6)	0.028	71 (15.5)	77 (16.8)	0.013

Cardiologic	87 (16.0)	57 (10.7)	0.053	54 (11.8)	54 (11.8)	0.000

Neurologic^b)	87 (16.0)	49 (9.2)	0.068	47 (10.3)	47 (10.3)	0.000

Nephrologic	67 (12.3)	76 (14.3)	0.001	26 (5.7)	27 (5.9)	0.002

Hepatologic	10 (1.8)	13 (2.4)	0.006	10 (2.2)	9 (2.0)	0.002

Adjunctive therapy

Vasopressor^a)	305 (56.0)	328 (61.5)	0.056	264 (57.8)	265 (58.0)	0.002

NMBAs^a)	291 (53.4)	310 (58.2)	0.048	250 (54.7)	255 (55.8)	0.011

RRT^a)	67 (12.3)	76 (14.6)	0.020	59 (12.9)	62 (13.6)	0.007

Values are presented as mean ± standard deviation or number (%).

SMD, standardized mean difference; BMI, body mass index; CCI, Charlson’s comorbidity index; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; NMBA, neuromuscular blocking agent; RRT, renal replacement therapy.

^a) All clinical data were calculated or obtained from medical records within the first 24 hours of mechanical ventilation.

^b) Included cerebrovascular accidents, intracerebral hemorrhage, subdural hemorrhage, subarachnoid hemorrhage, and Alzheimer’s dementia.

Table 2

Comparison of clinical outcomes of propensity-matched patients in the tracheostomy and no-tracheostomy groups

Clinical outcomes	Crude			Adjusted

	Tracheostomy (n = 545)	No-tracheostomy (n = 533)	p value	Tracheostomy (n = 457)	No-tracheostomy (n = 457)	p value
Days of MV (d)	27.1 ± 25.9	10.5 ± 17.2	< 0.001	27.9 ± 26.8	10.6 ± 18.3	< 0.001

ICU LOS (d)	26.6 ± 20.9	13.3 ± 19.7	< 0.001	26.4 ± 19.4	13.4 ± 9.9	< 0.001

Hospital LOS (d)	57.4 ± 72.2	29.0 ± 28.7	< 0.001	57.5 ± 75.4	29.1 ± 28.9	< 0.001

Mortality

ICU mortality	117 (21.5)	208 (39.0)	< 0.001	95 (20.8)	180 (39.4)	< 0.001

90-day cumulative mortality	220 (40.4)	293 (55.0)	< 0.001	182 (39.8)	256 (56.0)	< 0.001

Discharge outcomes

Death	155 (28.4)	234 (43.9)	< 0.001	128 (28.0)	204 (44.6)	< 0.001

Home	89 (16.3)	146 (27.4)	< 0.001	77 (16.8)	125 (27.4)	< 0.001

Step-down hospital	301 (55.2)	153 (28.7)	< 0.001	252 (55.1)	128 (28.0)	< 0.001

Cost (USD)

Total medical expenditures	38,785.1 ± 73,050.9	22,383.6 ± 32,514.8	< 0.001	39,022.4 ± 78,785.9	21,762.7 ± 28,857.4	< 0.001

Out-of-pocket medical expenditures	8,269.4 ± 16,093.2	4,562.6 ± 4815.9	< 0.001	8,243.1 ± 17,324.2	4,549.1 ± 4,251.1	< 0.001

Resource utilization rate of ICU

BOR (%)	19.3 (11.6–27.4)	10.1 (8.5–11.9)	0.002	13.3 (10.2–22.7)	8.0 (6.8–10.3)	0.003

Average LOS (in days)	29.2 (24.8–32.6)	13.8 (11.6–15.6)	< 0.001	29.7 (22.5–31.4)	13.3 (11.7–14.6)	< 0.001

BTR (inpatient per bed per year)	2.0 (1.3–5.3)	2.6 (1.6–4.1)	0.971	1.5 (1.0–4.3)	2.3 (1.4–3.4)	0.843

TI (in days)	104.2 (51.1–256.3)	126.6 (80.5–143.8)	0.709	159.8 (68.8–311.8)	142.0 (93.5–214.9)	0.958

Values are presented as mean ± standard deviation, number (%), or median (interquartile range).

MV, mechanical ventilation; ICU, intensive care unit; LOS, length of stay; USD, United State Dollars; BOR, bed occupancy rate; BTR, bed turnover rate; TI, turnover interval.

Table 3

Comparison of baseline characteristics of patients in the ET and LT groups before and after propensity score matching

Characteristic	Before matching			After matching

	ET (n = 275)	LT (n = 270)	SMD	ET (n = 205)	LT (n = 205)	SMD
Age (years)	71.8 ± 10.8	66.9 ± 14.1	0.451	71.1 ± 11.6	70.3 ± 10.8	0.077

Sex (male)	200 (72.7)	189 (70.0)	0.027	143 (69.8)	145 (70.7)	0.010

BMI (kg/m²)	21.7 ± 3.8	21.6 ± 3.5	0.001	21.6 ± 4.1	21.8 ± 3.6	0.050

CCI	5.9 ± 3.2	5.4 ± 2.9	0.177	5.9 ± 3.2	5.7 ± 2.7	0.060

APACHE II score^a)	21.1 ± 6.5	22.8 ± 7.0	0.270	21.8 ± 6.6	21.7 ± 6.2	0.008

SOFA score^a)	7.7 ± 3.0	8.2 ± 3.4	0.150	7.8 ± 3.0	7.8 ± 3.2	0.016

Comorbidity

Pulmonologic	149 (54.2)	154 (57.0)	0.029	117 (57.1)	118 (57.6)	0.005

Hemato-oncologic	44 (16.0)	37 (13.7)	0.023	29 (14.1)	27 (13.2)	0.010

Cardiologic	42 (15.3)	45 (16.7)	0.014	34 (16.6)	31 (15.1)	0.015

Neurologic^b)	44 (16.0)	43 (15.9)	0.001	32 (15.6)	32 (15.6)	0.000

Nephrologic	14 (5.1)	18 (6.7)	0.016	10 (4.9)	10 (4.9)	0.005

Hepatologic	4 (1.5)	6 (2.2)	0.008	4 (2.0)	4 (2.0)	0.000

Adjunctive therapy

Vasopressor^a)	163 (59.3)	142 (52.6)	0.067	119 (58.0)	117 (57.1)	0.010

NMBAs^a)	155 (56.4)	136 (50.4)	0.060	113 (55.1)	104 (50.7)	0.010

RRT^a)	30 (10.9)	37 (13.7)	0.028	25 (12.2)	23 (11.2)	0.044

Values are presented as mean ± standard deviation or number (%).

ET, early tracheostomy; LT, late tracheostomy; SMD, standardized mean difference; BMI, body mass index; CCI, Charlson’s comorbidity index; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; NMBA, neuromuscular blocking agent; RRT, renal replacement therapy.

^a) All clinical data were calculated or obtained from medical records within the first 24 hours of mechanical ventilation.

^b) Included cerebrovascular accidents, intracerebral hemorrhage, subdural hemorrhage, subarachnoid hemorrhage, and Alzheimer’s dementia.

Table 4

Comparison of clinical outcomes of propensity-matched patients in the ET and LT groups

Clinical outcomes	Crude			Adjusted

	ET (n = 275)	LT (n = 270)	p value	ET (n = 205)	LT (n = 205)	p value
Time to tracheostomy (d)	4.4 ± 2.0	13.4 ± 5.7	< 0.001	4.5 ± 2.0	13.5 ± 6.0	< 0.001

Days of MV (d)	22.2 ± 23.3	32.0 ± 27.4	< 0.001	23.5 ± 23.1	32.2 ± 28.7	< 0.001

ICU LOS (d)	21.2 ± 19.7	32.2 ± 20.6	< 0.001	21.1 ± 16.1	32.0 ± 20.6	< 0.001

Hospital LOS (d)	46.2 ± 39.9	68.9 ± 93.0	< 0.001	46.1 ± 38.4	67.0 ± 71.0	< 0.001

Mortality

ICU mortality	50 (18.2)	67 (24.8)	0.060	41 (19.8)	52 (25.1)	0.100

90-day cumulative mortality	105 (38.2)	115 (42.6)	0.295	75 (36.2)	87 (42.0)	0.484

Discharge outcomes

Death	66 (24.0)	89 (33.3)	0.019	54 (26.1)	65 (31.4)	0.191

Home	44 (16.0)	45 (16.7)	0.833	32 (15.6)	36 (17.6)	0.595

Step-down hospital	165 (60.0)	136 (50.4)	0.024	119 (58.0)	104 (50.7)	0.137

Decannulation rate	47 (17.1)	40 (14.8)	0.468	35 (17.1)	28 (13.7)	0.338

Cost (USD)

Total medical expenditures	29,506.1 ± 25,755.0	48,236.0 ± 99,692.5	0.003	30,502.8 ± 27,247.9	44,324.2 ± 57,167.9	0.002

Out-of-pocket medical expenditures	6,586.9 ± 6,089.9	9,983.1 ± 21,912.1	0.014	6,840.3 ± 6,193.5	9,037.7 ± 9,775.9	0.007

Resource utilization rate of ICU

BOR (%)	6.5 (2.9–14.6)	12.4 (8.9–15.0)	0.131	5.7 (2.5–12.5)	9.7 (7.8–12.5)	0.161

Average LOS (in days)	23.0 (17.5–29.4)	31.6 (26.6–37.2)	0.015	23.1 (17.3–28.1)	30.8 (26.4–36.6)	0.017

BTR (inpatient per bed per year)	1.1 (0.4–3.3)	1.5 (0.8–2.0)	0.677	0.8 (0.3–2.5)	1.0 (0.7–1.8)	0.519

TI (in days)	224.6 (86.6–1,010.9)	165.1 (143.4–369.7)	0.697	297.7 (123.5–1,425.7)	295.0 (171.7–449.8)	0.822

Values are presented as mean ± standard deviation, number (%), or median (interquartile range).

ET, early tracheostomy; LT, late tracheostomy; MV, mechanical ventilation; ICU, intensive care unit; LOS, length of stay; USD, United State Dollars; BOR, bed occupancy rate; BTR, bed turnover rate; TI, turnover interval.

REFERENCES

1. Law AC, Tian W, Song Y, Stevens JP, Walkey AJ. Decline in prolonged acute mechanical ventilation, 2011–2019. Am J Respir Crit Care Med 2022;206:640–644.

2. Mehta AB, Syeda SN, Wiener RS, Walkey AJ. Epidemiological trends in invasive mechanical ventilation in the United States: a population-based study. J Crit Care 2015;30:1217–1221.

3. Fagon JY, Chastre J, Vuagnat A, Trouillet JL, Novara A, Gibert C. Nosocomial pneumonia and mortality among patients in intensive care units. JAMA 1996;275:866–869.

4. Holzapfel L, Chevret S, Madinier G, et al. Influence of long-term oro- or nasotracheal intubation on nosocomial maxillary sinusitis and pneumonia: results of a prospective, randomized, clinical trial. Crit Care Med 1993;21:1132–1138.

5. Freeman BD. Tracheostomy update: when and how. Crit Care Clin 2017;33:311–322.

6. Whitmore KA, Townsend SC, Laupland KB. Management of tracheostomies in the intensive care unit: a scoping review. BMJ Open Respir Res 2020;7:e000651.

7. Quinn L, Veenith T, Bion J, Hemming K, Whitehouse T, Lilford R. Bayesian analysis of a systematic review of early versus late tracheostomy in ICU patients. Br J Anaesth 2022;129:693–702.

8. Abe T, Madotto F, Pham T, et al. Epidemiology and patterns of tracheostomy practice in patients with acute respiratory distress syndrome in ICUs across 50 countries. Crit Care 2018;22:195.

9. Clec’h C, Alberti C, Vincent F, et al. Tracheostomy does not improve the outcome of patients requiring prolonged mechanical ventilation: a propensity analysis. Crit Care Med 2007;35:132–138.

10. Aquino Esperanza J, Pelosi P, Blanch L. What’s new in intensive care: tracheostomy-what is known and what remains to be determined. Intensive Care Med 2019;45:1619–1621.

11. Rose L, Messer B. Prolonged mechanical ventilation, weaning, and the role of tracheostomy. Crit Care Clin 2024;40:409–427.

12. Oh TK, Kim HG, Song IA. Epidemiologic study of intensive care unit admission in South Korea: a nationwide population-based cohort study from 2010 to 2019. Int J Environ Res Public Health 2022;20:81.

13. Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301–1308.

14. File TM Jr, Ramirez JA. Community-acquired pneumonia. N Engl J Med 2023;389:632–641.

15. Lanks CW, Musani AI, Hsia DW. Community-acquired pneumonia and hospital-acquired pneumonia. Med Clin North Am 2019;103:487–501.

16. Zilberberg MD, Nathanson BH, Ways J, Shorr AF. Characteristics, hospital course, and outcomes of patients requiring prolonged acute versus short-term mechanical ventilation in the United States, 2014–2018. Crit Care Med 2020;48:1587–1594.

17. Mehrtak M, Yusefzadeh H, Jaafaripooyan E. Pabon Lasso and Data Envelopment Analysis: a complementary approach to hospital performance measurement. Glob J Health Sci 2014;6:107–116.

18. Halpern NA, Pastores SM. Critical care medicine beds, use, occupancy, and costs in the United States: a methodological review. Crit Care Med 2015;43:2452–2459.

19. Halpern NA, Pastores SM. Critical care medicine in the United States 2000–2005: an analysis of bed numbers, occupancy rates, payer mix, and costs. Crit Care Med 2010;38:65–71.

20. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818–829.

21. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/ failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996;22:707–710.

22. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–383.

23. Landa K, Pappas TN. Zollinger’s atlas of surgical operations, 11th edition. Ann Surg 2023;277:e488.

24. Korean Bronchoesophagological Society Guideline Task Force; Nam IC, et al.; Shin YS. Guidelines for tracheostomy from the Korean Bronchoesophagological Society. Clin Exp Otorhinolaryngol 2020;13:361–375.

25. Seong SC, Kim YY, Khang YH, et al. Data resource profile: the National Health Information Database of the National Health Insurance Service in South Korea. Int J Epidemiol 2017;46:799–800.

26. Austin PC. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat 2011;10:150–161.

27. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 2013;310:2191–2194.

28. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453–1457.

29. Freeman BD, Morris PE. Tracheostomy practice in adults with acute respiratory failure. Crit Care Med 2012;40:2890–2896.

30. Combes A, Luyt CE, Nieszkowska A, Trouillet JL, Gibert C, Chastre J. Is tracheostomy associated with better outcomes for patients requiring long-term mechanical ventilation? Crit Care Med 2007;35:802–807.

31. Ishizaki M, Toyama M, Imura H, Takahashi Y, Nakayama T. Tracheostomy decannulation rates in Japan: a retrospective cohort study using a claims database. Sci Rep 2022;12:19801.

32. Cox CE, Carson SS, Lindquist JH, Olsen MK, Govert JA, Chelluri L. Differences in one-year health outcomes and resource utilization by definition of prolonged mechanical ventilation: a prospective cohort study. Crit Care 2007;11:R9.

33. MacIntyre NR. The ventilator discontinuation process: an expanding evidence base. Respir Care 2013;58:1074–1086.

34. Jeong ES, Lee K. Clinical application of modified Burns Wean Assessment Program scores at first spontaneous breathing trial in weaning patients from mechanical ventilation. Acute Crit Care 2018;33:260–268.

35. Cinotti R, Voicu S, Jaber S, et al. Tracheostomy and long-term mortality in ICU patients undergoing prolonged mechanical ventilation. PLoS One 2019;14:e0220399.

36. Raimondi N, Vial MR, Calleja J, et al. Evidence-based guidelines for the use of tracheostomy in critically ill patients. J Crit Care 2017;38:304–318.

37. Hernandez G, Ramos FJ, Añon JM, et al. Early tracheostomy for managing ICU capacity during the COVID-19 outbreak: a propensity-matched cohort study. Chest 2022;161:121–129.

38. Lim CM, Kwak SH, Suh GY, Koh Y. Critical care in Korea: present and future. J Korean Med Sci 2015;30:1540–1544.

39. Kwak SH, Jeong CW, Lee SH, Lee HJ, Koh Y. Current status of intensive care units registered as critical care subspecialty training hospitals in Korea. J Korean Med Sci 2014;29:431–437.

40. Epstein SK. Late complications of tracheostomy. Respir Care 2005;50:542–549.

41. Rana S, Pendem S, Pogodzinski MS, Hubmayr RD, Gajic O. Tracheostomy in critically ill patients. Mayo Clin Proc 2005;80:1632–1638.

42. Quality of Life After Mechanized Ventilation in the Elderly Study Investigators. 2-month mortality and functional status of critically ill adult patients receiving prolonged mechanical ventilation. Chest 2002;121:549–558.