Abstract
Background
Controversy exists on the optimal timing for performing a tracheostomy in critically ill patients, particularly in terms of clinical outcomes such as weaning failures, mechanical ventilation duration, ventilator-associated pneumonia, and mortality. The present study aimed to determine whether prolonged mechanically ventilated patients had a higher success rate in ventilator weaning after an “early” tracheostomy”, defined as ≤ 14 days following endotracheal intubation.
Methods
An observational, retrospective single-center study of 738 prolonged ventilated, tracheotomized patients treated at a national weaning center over 12 years. Propensity score matching and binary logistic regression analysis were used to evaluate whether an early tracheostomy independently predicted prolonged weaning failures, defined as the transition to home mechanical ventilation.
Results
The entire cohort comprised 507 early procedures (69%), and propensity score matching yielded 220 patients in each group undergoing either an early or a late tracheostomy. Prolonged weaning failure rates (34% vs. 33%, P = 0.762) and other secondary outcomes – decannulation failures, frequencies of long-term oxygen therapy at hospital discharge, mortality rates – were not different between these groups, and an early tracheostomy was not independently associated with failure to wean in logistic regression analysis. However, the groups differed significantly in the total duration of mechanical ventilation (40 days [IQR 32–56] vs. 51 days [52–70], P < 0.01), primarily due to the additional days spent on ventilators before admission to the weaning center in cases of late tracheostomy.
Conclusions
No significant difference in weaning failure rates or other secondary outcomes was observed among prolonged mechanically ventilated, tracheotomized patients treated at a specialized facility, regardless of early (≤ 14 days) or late tracheostomy.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12890-026-04252-9.
Keywords: Mechanical ventilation, Tracheostomy timing, Prolonged weaning, Spontaneous breathing trial
Background
The tracheostomy procedure is commonly performed on patients in the intensive care unit suffering from prolonged mechanical ventilation, loss of airway-protecting reflexes (e.g., in neurological diseases), or obstruction of the upper airways [1]. The use of a tracheostomy is reputed to offer several advantages over endotracheal tubes, relating to a reduction in sedative requirements, such as improved communication and swallowing, lower risk of ventilator-associated pneumonia, a shorter duration of mechanical ventilation, and decreased mortality rates, with clinical trial data, however, providing inconsistent evidence to support each of the outlined assertions [2]. Moreover, it is essential to weigh these benefits against the potential adverse effects of the procedure, such as major bleeding, wound infection (surgical tracheostomy dominates), or tracheal stenosis [3]. Most research on tracheostomy in intensive care focuses on the optimal technique (dilatational versus surgical) [4] and the ideal timing of the procedure [3]. Theoretically, “early tracheostomies” could benefit patient outcomes (e.g., weaning outcomes, total ventilation duration, mortality rates), given their potential advantages over endotracheal tubes. However, studies have reported conflicting results on whether early tracheostomies are more beneficial than those performed later, likely due to patient heterogeneity and the wide range of definitions of an “early tracheostomy” [5–9].
To the best of our knowledge, no research has been conducted to date that exclusively evaluates tracheotomized subjects treated in a specialized weaning facility, focusing on specific outcomes such as prolonged weaning failures, transfer rates to home mechanical ventilation, decannulation failures, or frequencies of long-term oxygen therapy (LTOT) at hospital discharge. Our study hypothesized that prolonged mechanically ventilated patients treated at a national weaning center were more likely to be successfully weaned off the ventilator after an early tracheostomy, defined as ≤ 14 days following endotracheal intubation.
Methods
This retrospective study evaluated prolonged mechanically ventilated, tracheotomized patients consecutively admitted at a specialized national weaning facility. Institutional review board approval was obtained for this project (Ethics Committee of the State Chamber of Physicians of Baden-Wuerttemberg, Germany, file number F-2025–050), allowing written informed consent to be waived.
Patient selection
Subjects included in the study had been referred from ICUs across Germany for prolonged weaning from mechanical ventilation, meeting both the ICC (International Consensus Conference) and WIND criteria, which relate to failure of at least three weaning attempts or requirement of more than seven days of mechanical ventilation after the first separation attempt [10, 11] (Fig. 1). Additional preconditions for admittance were mechanical ventilation via tracheostomy tube with positive end-expiratory pressure (PEEP) less than 10 cmH2O and a fraction of inspired oxygen (FiO2) less than 0.6, hemodynamic stability without the need for vasopressors or inotropes, and the absence of deep sedation. No restrictions were imposed on transferring patients to the weaning center based on the primary reason for mechanical ventilation or concomitant chronic disorders.
Fig. 1.
Patient flow diagram. *: Refers to the ICC or WIND criteria for prolonged weaning [10, 11]. Abbreviations: SBT, spontaneous breathing trial; ECLS, extracorporeal lung support; CCI, Charlson comorbidity index
Data collection
Patient records were assessed for demographics (e.g., age, gender, body mass index), clinical features at the time of admission to the weaning center (e.g., APACHE-II score, type of tracheostomy, the primary reason for mechanical ventilation, extracorporeal lung support during the ICU stay), and comorbidities (including the Charlson index). We summarized weaning outcomes, including prolonged weaning failure rates, weaning duration, total mechanical ventilation duration, ventilator-free days within 90 days of intubation (VFD-90), decannulation failures, LTOT frequency at hospital discharge, and hospital mortality rates. Moreover, we analyzed ventilatory variables and indexes immediately before the first protocolized spontaneous breathing trial (SBT) following admission to the center [12, 13] to evaluate gas exchange, lung dimensions, respiratory mechanics, and ventilatory efficiency in each case.
Classification of tracheostomy timing
Although tracheostomy is a frequently performed procedure on mechanically ventilated patients in the ICU, the optimal timing for tracheostomy placement remains a topic of debate and ongoing investigation. Considering most studies focus on the period between 10 and 14 days after endotracheal intubation [8, 9], we defined early and late tracheostomies based on the 14-day threshold, using the 10-day threshold for sensitivity analysis.
Prolonged weaning failure and success
The outcome of prolonged weaning at the end of the ventilator liberation process was categorized as either failure or success, based solely on patients’ spontaneous breathing abilities [14]. Prolonged weaning failure is defined as “long-term ventilator dependence” due to persistent ventilatory failure with transitioning to domiciliary non-invasive (NIV, by face mask) or invasive ventilation (IMV, by tracheostomy tube). Ventilatory failure describes recurrent hypercapnia [PaCO2 > 45 mmHg] during daily weaning trials, preventing the extension of spontaneous unassisted breathing or hypercapnia occurring within seven days after weaning completion, determined by the last day on which the patient was ventilated. These patients usually remain ventilator-attached at discharge. Conversely, the definition of prolonged weaning success is sustained “autonomic breathing” (≥ 7 days) without concomitant signs of ventilatory failure (hypercapnia) after weaning completion. These patients remain ventilator-detached at hospital discharge [14].
Respiratory system evaluation
After admission to the weaning center, pressure-controlled, assist-control (A/C) mechanical ventilation was used on all subjects to unload the respiratory pump effectively. A standardized ventilator liberation method, starting with a standardized first 30-minute SBT, was applied once weaning readiness criteria were met, including FiO2 ≤ 0.4, PEEP ≤ 8 cmH2O, stable hemodynamics without vasopressors or inotropic agents, and normocapnia on mechanical ventilation [12, 13]. Along with recording ventilatory variables, arterial blood gases (ABGs) were collected during A/C ventilation in the semi-recumbent position immediately before the onset of the weaning trial, which used a T-piece (with oxygen admixture at the same level as during mechanical ventilation) or CPAP (adjusted to PEEP levels). Each weaning trial ended with another ABG.
We collected the following ventilatory variables immediately before the onset of the first SBT: FiO2, respiratory rate (RR), tidal volume (VT), peak inspiratory airway pressure (Pmax), PEEP, and dynamic driving pressure (∆Paw = Pmax–PEEP). Next, we calculated the following parameters to evaluate gas exchange variables, lung dimensions, respiratory mechanics, and ventilatory efficiency within each case: P/F ratio, predicted body weight normalized tidal volume (VT/PBW; representing lung dimension in the specific setting of prolonged weaning) [14], ventilatory ratio (VR; a composite measure of ventilation efficiency and shunt, correlating with the pulmonary dead-space fraction) [15], dynamic respiratory system compliance (Cdyn = VT/∆Paw; reflecting respiratory mechanics) [13], and mechanical power (MP), with the latter further normalized to Cdyn (referred to as power density) [13, 14] (Additional file 1).
Study endpoints
We hypothesized that early tracheostomy (≤ 14 days following endotracheal intubation) would benefit weaning outcomes in patients on prolonged ventilation. Therefore, the primary objective was to compare prolonged weaning outcomes (failure versus success as defined above) between patients with early and late tracheostomy. Secondary outcomes included weaning duration from the first SBT upon admission to the weaning center, total mechanical ventilation duration, VFD-90, decannulation failures, LTOT frequency at hospital discharge, and mortality rates. Moreover, we compared ventilatory indexes reflecting gas exchange, lung dimensions, respiratory mechanics, and ventilatory efficiency among these groups at the time of the first SBT following admission to the center.
Statistical analysis
Categorical variables were compared using the Chi-square or Fisher’s exact test. Depending on the continuous variables` homogeneity of variance, determined by the Kolmogorov-Smirnov normality test, differences between groups were analyzed through Student’s t-test or Mann-Whitney U-test.
Propensity score-matched cohorts were used to directly compare outcomes between patients with early (≤ 14 days) versus late tracheostomies. Each patient’s propensity score was calculated employing a logistic regression model, derived from observed demographics, baseline characteristics, and comorbidities, to estimate the predicted probability of an early tracheostomy after intubation (Additional file 1). Subjects were matched 1:1 without replacement using the nearest-neighbor method, with a caliper of 0.1 SD of the propensity score logit. We performed additional binary logistic regression analysis on the entire study population to determine if tracheostomy timing was independently associated with prolonged weaning failure, introducing this parameter either as a categorical variable (≤ 14 / >14 days following intubation) or as a continuous variable (the number of days between intubation and tracheostomy) into different models. The multivariable analysis employed forward selection of covariates for propensity score matching, as well as baseline parameters with P values < 0.2 in the bivariate analysis. Hosmer-Lemeshow and Nagelkerke R2 were used to evaluate the model’s goodness-of-fit. Probabilities are reported as odds ratios (ORs) with 95% confidence intervals (95% CI). We plotted Kaplan-Meier curves for the probability of prolonged weaning failure in the early and late tracheostomy groups and compared them using the log-rank test. Finally, we conducted sensitivity analyses by redefining early tracheostomy as occurring within 10 days of endotracheal intubation.
We performed two-tailed tests; statistical significance was indicated by P < 0.05. The analyses were conducted using MedCalc software version 23.2.7 (Ostend, Belgium).
Results
The study included 738 out of 950 (78%) screened patients between December 2011 and January 2024 (Fig. 1). This unmatched cohort comprised 507 patients (69%) who underwent early tracheostomy (≤ 14 days). A summary of the baseline characteristics of the entire study population is presented in Table S1 (see Additional file 1). Propensity score matching resulted in 220 patients per group: early (≤ 14 days) versus late tracheostomy (Table 1). The baseline characteristics of these patients were well balanced between groups (Additional file 1); however, the number of days spent on ventilators at weaning center admission was significantly higher in patients with late tracheostomy.
Table 1.
Clinical characteristics on admission to the weaning center – Comparison of propensity score-matched cohorts with early (≤ 14 days) and late tracheostomy (> 14 days)
| Clinical characteristics | All patients (n = 440) |
Early tracheostomy (n = 220) |
Late tracheostomy (n = 220) |
P valuea |
|---|---|---|---|---|
| Age (years) | 69 (61–76) | 69 (62–75) | 69 (60–76) | 0.717b |
| Female gender | 172 (39) | 88 (40) | 84 (38) | 0.696c |
| Predicted body weight (kg) | 66 (58–73) | 66 (60–74) | 66 (58–73) | 0.984b |
| Body mass index (kg/m2) | 26.0 (22.9–30.9) | 26.3 (23.1–31.1) | 25.6 (22.5–30.8) | 0.174b |
| Obesity (BMI ≥ 30 kg/m 2 ) | 126 (29) | 67 (31) | 59 (27) | 0.399c |
| Smoking history | 164 (37) | 81 (37) | 83 (38) | 0.844c |
| APACHE-II (points) | 15 (12–19) | 16 (12–19) | 15 (13–19) | 0.736b |
| Pre-existing domiciliary NIV | 22 (5) | 13 (6) | 9 (4) | 0.382c |
| Extracorporeal lung support | 44 (10) | 22 (10) | 22 (10) | 0.999c |
| Percutaneous tracheostomy | 316 (72) | 156 (71) | 160 (73) | 0.672c |
| Intubation to tracheostomy (days) | 15 (9–19) | 9 (6–12) | 19 (17–22) | – |
| Ventilator days on center admission | 26 (19–37) | 20 (14–29) | 31 (24–42) | < 0.01 b |
| Reason for intubation | ||||
| Pneumonia | 159 (36) | 81 (37) | 78 (36) | 0.766c |
| SARS-CoV-2 infection | 55 (13) | 28 (13) | 27 (12) | 0.856c |
| Surgery | 128 (29) | 63 (29) | 65 (30) | 0.834c |
| Cardiopulmonary resuscitation | 43 (10) | 20 (9) | 23 (11) | 0.631c |
| Sepsis (extrapulmonary) | 35 (8) | 21 (10) | 14 (6) | 0.218c |
| Acute exacerbation of COPD | 30 (7) | 17 (8) | 13 (6) | 0.450c |
| Acute heart failure | 11 (3) | 5 (2) | 6 (3) | 0.760d |
| Other | 34 (8) | 13 (6) | 21 (10) | 0.154c |
| Comorbidities | ||||
| Charlson index (points) | 5 (4–7) | 5 (4–7) | 5 (4–7) | 0.869b |
| Renal insufficiency | 134 (31) | 68 (31) | 66 (30) | 0.836c |
| Hemodialysis on admission | 84 (19) | 42 (19) | 42 (19) | 0.999c |
| Coronary artery disease | 112 (26) | 54 (25) | 58 (26) | 0.662c |
| Diabetes mellitus | 107 (24) | 57 (26) | 50 (23) | 0.437c |
| COPD | 88 (20) | 43 (20) | 45 (21) | 0.812c |
| Chronic heart failure | 65 (15) | 31 (14) | 34 (16) | 0.687c |
| Malignancy | 32 (7) | 14 (6) | 18 (8) | 0.463c |
| Hepatopathy | 22 (5) | 12 (6) | 10 (5) | 0.662c |
| Interstitial lung diseases | 16 (4) | 8 (4) | 8 (4) | 0.999c |
| Neuromuscular disease | 13 (3) | 6 (3) | 7 (3) | 0.779c |
Continuous variables are presented as median (– interquartile range [IQR]); categorical variables are presented as numbers (%). Significant values (P < 0.05) are in bold
Abbreviations: BMI body mass index, APACHE-II Acute Physiology and Chronic Health Evaluation II score, NIV non-invasive ventilation, COPD chronic obstructive pulmonary disease
a: P value for differences between patients with early (≤ 14 days) and late (> 14 days) tracheostomy
b: Mann-Whitney U-test
c: Chi-squared test
d: Fisher's exact test
Study outcomes in matched cohorts
There were no differences between groups in terms of prolonged weaning failures and the number of patients experiencing domiciliary NIV or IMV following discharge from the hospital. In addition, no significant differences were found in most secondary outcomes, except for total mechanical ventilation duration and VFD-90 (Table 2).
Table 2.
Results of prolonged weaning – Comparison of propensity score-matched cohorts with early (≤ 14 days) and late tracheostomy (> 14 days)
| Primary outcome | All patients (n = 440) |
Early tracheostomy (n = 220) |
Late tracheostomy (n = 220) |
P valuea |
|---|---|---|---|---|
| Prolonged weaning failure | 147 (33) | 75 (34) | 72 (33) | 0.762c |
| HMV-NIV | 57 (13) | 32 (15) | 25 (11) | 0.321c |
| HMV-IMV | 90 (21) | 43 (20) | 47 (21) | 0.637c |
| Secondary outcomes | ||||
| Weaning duration from first SBT (days) | 14 (11–21) | 14 (11–21) | 14 (11–20) | 0.853b |
| Duration of mechanical ventilation (days)§ | 47 (36–61) | 40 (32–56) | 51 (42–70) | < 0.01 b |
| VFD-90 (days) | 43 (29–54) | 50 (34–58) | 39 (20–49) | < 0.01 b |
| Decannulation failure | 179 (41) | 85 (39) | 94 (43) | 0.383c |
| SB at weaning completion (hours per day)$ | 24 (15–24) | 24 (16–24) | 24 (14–24) | 0.962b |
| LTOT at hospital discharge* | 265 (66) | 137 (68) | 128 (65) | 0.458c |
| Weaning unit-LOS (days) | 33 (24–47) | 33 (24–48) | 33 (22–47) | 0.889b |
| Death during weaning | 15 (3) | 7 (3) | 8 (4) | 0.793c |
| Hospital mortality | 41 (9) | 19 (9) | 22 (10) | 0.623c |
Continuous variables are presented as median (– interquartile range [IQR]); categorical variables are presented as numbers (%). Significant values (P < 0.05) are in bold
Abbreviations: HMV-NIV Home mechanical ventilation – Non-invasive ventilation, HMV-IMV Home mechanical ventilation – Invasive mechanical ventilation, SBT spontaneous breathing trial, VFD-90 ventilator-free days within 90 days following endotracheal intubation, SB spontaneous breathing, LTOT long-term oxygen therapy, LOS length of stay
a: P value for differences between patients with early (≤ 14 days) and late (> 14 days) tracheostomy
b: Mann-Whitney U-test
c: Chi-squared test
§: The total duration of mechanical ventilation, the period from intubation in the referring ICU until the completion of weaning at the weaning center
$: The values of 15 patients who died during weaning were set at 0 hours per day
*: Excluding 41 patients who died during their hospital stay
Respiratory system evaluation in matched cohorts
PF ratios differed slightly between patients with early (≤ 14 days) and late tracheostomies before their first weaning trial after weaning center admission, with no significant differences in PBW-normalized tidal volumes, dynamic respiratory system compliance, ventilatory ratio, mechanical power, or power density (Cdyn-MP) (Table 3).
Table 3.
Respiratory system evaluation at first SBT following center admission – Comparison of propensity score-matched cohorts with early (≤ 14 days) and late tracheostomy (> 14 days)
| Ventilatory variables and indexes (pre-SBT) |
All patients (n = 440) |
Early tracheostomy (n = 220) |
Late tracheostomy (n = 220) |
P valuea |
|---|---|---|---|---|
| P/F ratio (mmHg) | 279 (221–332) | 272 (211–325) | 280 (232–343) | 0.048 b |
| P/F ratio < 250 mmHg | 160 (36) | 88 (40) | 72 (33) | 0.113c |
| VT/PBW (mL/kg) | 8.3 (7.4–9.7) | 8.4 (7.5–9.5) | 8.3 (7.4–9.7) | 0.910b |
| Dynamic driving pressure (cmH2O) | 16 (15–19) | 16 (15–19) | 15 (15–18) | 0.227b |
| Ventilatory ratio | 1.90 (1.59–2.37) | 1.87 (1.59–2.31) | 1.93 (1.60–2.42) | 0.756b |
| Cdyn (mL/cmH2O) | 32 (27–41) | 32 (27–41) | 32 (28–40) | 0.734b |
| Mechanical power (Joule/min) | 20.8 (17.5–21.2) | 20.8 (17.5–25.3) | 20.9 (17.7–24.9) | 0.820b |
| Cdyn-MP (cmH2O2/min) | 6600 (5054–8400) | 6704 (5054–8519) | 6384 (5070–8320) | 0.642b |
|
Arterial blood gas analysis (pre-SBT) | ||||
| PaO2 (mmHg) | 81 (71–92) | 81 (72–93) | 81 (71–89) | 0.632b |
| PaCO2 (mmHg) | 35 (31–40) | 35 (32–40) | 35 (30–39) | 0.350b |
| pH | 7.48 (7.45–7.52) | 7.48 (7.45–7.52) | 7.48 (7.45–7.52) | 0.503b |
Continuous variables are presented as median (– interquartile range [IQR]). Significant values (P < 0.05) are in bold
Abbreviations: SBT spontaneous breathing trial, P/F ratio the ratio of partial pressure of oxygen to fraction of inspired oxygen, VT/PBW tidal volume normalized to the predicted body weight, Cdyn dynamic respiratory system compliance, Cdyn-MP mechanical power normalized to dynamic respiratory system compliance
a: P value for differences between patients with early (≤ 14 days) and late (> 14 days) tracheostomy
b: Mann-Whitney U-test
c: Chi-squared test
Logistic regression and survival analysis
In univariable and multivariable logistic regression analyses, after adjusting for baseline clinical characteristics, the main reason for intubation, and comorbidities, there were no significant associations between early tracheostomy (≤ 14 days) and prolonged weaning failure (Additional file 1: Table S2-S3). However, including tracheostomy timing as a continuous variable in uni- and multivariable models indicated an inverse relationship between the interval between intubation and tracheostomy and the inability to wean from prolonged mechanical ventilation (Additional file 1: Table S4-S5).
Based on the Kaplan-Meier survival analysis, there were no significant differences in the probability of prolonged weaning failure between patients with early (≤ 14 days) and late tracheostomy (Fig. 2).
Fig. 2.
Kaplan-Meier curves comparing patients with early and late tracheostomy. Comparison of prolonged weaning outcomes between patients with early (≤ 14 days) and late tracheostomy (N = 738)
Sensitivity analysis
The sensitivity analysis based on the 10-day threshold for defining early and late tracheostomy provided similar results in terms of the primary and secondary study outcomes (Additional file 1: Tables S6-S9, Figure S1).
Discussion
Study results can be summarized as follows: Based on propensity score matching, there was no difference between prolonged ventilated patients with early (≤ 14 days) and late tracheostomy regarding weaning failures and other secondary outcomes, including weaning duration, decannulation failures, frequency of LTOT at hospital discharge, and mortality rates. Similarly, in binary logistic regression analysis, tracheostomy timing – introduced as a categorical variable into the models – was not independently associated with the primary outcome. However, the groups differed significantly in the total duration of mechanical ventilation and VFD-90, primarily due to the additional days spent on ventilators before admission to the weaning center in cases of late tracheostomy.
Based on recent meta-analyses of non-COVID-19 (including 17 randomized clinical trials (RCTs)) and COVID-19 patients (observational studies only), early tracheostomy was associated with shorter mechanical ventilation duration and decreased ventilator-associated pneumonia (VAP) risk but not with a reduction in mortality [5, 9]. These data contradict other reports involving non-COVID-19 patients (meta-analyses of RCTs only), particularly regarding VAP and mortality rates [6–8]. With early tracheostomy, the only outcome measure consistently improving across trials was total mechanical ventilation duration [5–9], which aligns with our findings. Interestingly, none of these meta-analyses and systematic reviews presented data concerning weaning failure rates. It is essential to note that studies examining the effects of tracheostomy timing on clinical outcomes reveal substantial heterogeneity regarding the definition of early versus late tracheostomy [6, 7] and patient characteristics. As a result of the heterogeneity of patients, it is difficult to formulate definitive recommendations, signifying that specific patient subgroups, such as those suffering from traumatic brain injury [16], may still benefit from early tracheostomy. By contrast, our findings may also point to potential differences in ventilator weaning procedures following tracheostomy placement, which may be equally important as the timing of the procedure. Despite extensive literature on the decision to insert a tracheostomy and the methods for performing a tracheostomy, questions remain about the optimal treatment following the procedure [17], and most studies do not detail the management of these patients [8]. To the best of our knowledge, the present study is the first analysis to evaluate subjects treated exclusively in a specialized weaning facility. Given this, our results may not apply to tracheotomized patients in medical or surgical intensive care units [17]. In addition, this is the only study to evaluate the transfer rate to home mechanical ventilation and other clinical outcomes, such as decannulation failures or the frequency of long-term oxygen therapy at hospital discharge, in relation to the timing of tracheostomy placement, displaying no apparent differences between early and late tracheostomies.
According to the present results, the increased total mechanical ventilation duration and less VFD-90 were primarily attributed to significantly more days spent on ventilators at the referring ICUs within the late tracheostomy group. One explanation for this finding could be that patients with late tracheostomies may have experienced a more extended period of unreadiness for weaning following intubation [18]. In such cases, most clinicians do not perform tracheostomies on patients with severe conditions associated with deteriorated gas exchange and ventilatory failure necessitating more aggressive mechanical ventilation (e.g., higher FiO2 or PEEP). Upon admission to the weaning center, there was no longer a difference between the groups in weaning duration or other clinical outcomes. Moreover, no clinically relevant differences were found in gas exchange variables or parameters reflecting lung dimensions, respiratory mechanics, and ventilatory efficiency at the first SBT following center admission. Some of these indexes, particularly dynamic respiratory system compliance and power density, have been shown to predict prolonged weaning outcomes reliably [14]. Measuring these parameters provides an indirect indicator of respiratory muscle load during unassisted, spontaneous ventilation, which is crucial for sustained autonomic breathing following prolonged ventilation. In other words, weaning readiness before and weaning trajectories after the first SBT were comparable between the two groups. Contrary to expectations [19], the net reduction in total ventilation duration for early tracheostomy did not benefit primary or secondary outcomes. Instead, logistic regression analysis revealed a significant inverse relationship between the primary outcome and the number of days from intubation to tracheostomy (Additional file 1: Table S4-S5), evidence that may reflect selection bias, as patients with evidently worse prognoses regarding the ability to wean off the ventilator may have been tracheotomized very early in the course of their illness.
In the present study, secondary outcomes did not differ by tracheostomy timing, including failure to decannulate, LTOT frequency at hospital discharge, and mortality rates. A high rate of unsuccessful decannulations was observed, reaching 40%, consistent with previous findings [20, 21]. Common reasons for failure to decannulate include persistent, severe ICU-acquired dysphagia (resulting in excessive salivation and aspiration) and long-term ventilator dependence following prolonged weaning, neither of which was evaluated in the present study.
This study has limitations. First, external validity is questionable due to the study’s monocentric design and the restriction of findings to patients admitted to specialized long-term weaning facilities, which limits generalizability. Second, as this was a retrospective study, despite conducting propensity score matching and logistic regression analyses, we may have overlooked confounding factors that led to imbalances in covariates between groups. Accordingly, the inability to control for factors before transfer to the weaning facility, such as variability in ICU practices (e.g., sedation and mechanical ventilation management) and health care resources, could have biased the results. Third, given the limited information on managing patients in the referring ICU, we could not assess whether tracheostomy timing would affect other important outcomes, such as VAP. Similarly, patient-centered outcomes such as swallowing, speech, or mobility impairments may vary depending on the timing of the tracheostomy. Moreover, mortality rates are likely to be deflated, as only patients who survived their ICU stay were included in the study.
Conclusions
Among a large cohort of prolonged mechanically ventilated, tracheotomized patients treated in a specialized facility, early tracheostomy (≤ 14 days) did not result in higher weaning success rates – defined as transition to long-term home mechanical ventilation – or improvements in other secondary outcomes, but was associated with a shorter total duration of mechanical ventilation and more ventilator-free days at 90 days following intubation. Future research should focus on other patient-centered outcomes (e.g., dysphagia or speech and mobility impairments), which may also depend on the timing of the tracheostomy procedure.
Supplementary Information
Acknowledgements
Not applicable.
Authors’ contributions
Study concept and design: AG. Acquisition of data: AG, KT, AL. Analysis and interpretation of data: AG, KT, AL, MB, ATK, NK. First draft of the manuscript: AG. Critical manuscript revision for important intellectual content: AG, KT, AL, MB, ATK, NK. Statistical analysis: AG. All authors have read and approved the final version of the manuscript.
Funding
None.
Data availability
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
Institutional review board approval was obtained for this project (Ethics Committee of the State Chamber of Physicians of Baden-Wuerttemberg, file number F-2025–050), and the study was conducted in accordance with the Declaration of Helsinki, allowing written informed consent to be waived.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Abbreviations
ABG: Arterial blood gas analysis.
A/C ventilation: Assisted-controlled ventilation mode.
APACHE-II: Acute physiology and chronic health evaluation score (II).
Cdyn: Dynamic respiratory system compliance.
Cdyn-MP: Mechanical power normalized to Cdyn (referred to as power density).
COPD: Chronic obstructive pulmonary disease.
CPAP: Continuous positive airway pressure.
∆Paw: Dynamic driving pressure.
FiO2: Fraction of inspired oxygen.
IMV: Invasive mechanical ventilation.
LTOT: Long-term oxygen therapy.
MP: Mechanical power.
NIV: Non-invasive ventilation.
PBW: Predicted body weight.
PEEP: Positive end-expiratory pressure.
P/F ratio: The ratio of the partial pressure of oxygen to the fraction of inspired oxygen.
Pmax: Peak positive airway pressure.
RCTs: Randomized clinical trials.
RR: Respiratory rate.
SBT: Spontaneous breathing trial.
SD: Standard deviation.
VAP: Ventilator-associated pneumonia.
VR: Ventilatory ratio.
VT: Tidal volume.
VT/PBW: Tidal volume normalized to the predicted body weight.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Combes A, Luyt CE, Nieszkowska A, et al. Is tracheostomy associated with better outcomes for patients requiring long-term mechanical ventilation? Crit Care Med. 2007;35(3):802–7. [DOI] [PubMed] [Google Scholar]
- 2.Durbin CG Jr. Tracheostomy: why, when, and how? Respir Care. 2010;55(8):1056–68. [PubMed] [Google Scholar]
- 3.Putensen C, Theuerkauf N, Guenther U, et al. Percutaneous and surgical tracheostomy in critically ill patients: a meta-analysis. Crit Care. 2014;18(6):544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Brass P, Hellmich M, Ladra A, et al. Percutaneous techniques versus surgical techniques for tracheostomy. Cochrane Database Syst Rev. 2016;7(7):CD008045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chorath K, Hoang A, Rajasekaran K, Moreira A. Association of early versus late tracheostomy placement with pneumonia and ventilator days in critically ill patients: a meta-analysis. JAMA Otolaryngol Head Neck Surg. 2021;147(5):450–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Deng H, Fang Q, Chen K, Zhang X. Early versus late tracheotomy in ICU patients: a meta-analysis of randomized controlled trials. Med (Baltim). 2021;100(3):e24329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hosokawa K, Nishimura M, Egi M, Vincent JL. Timing of tracheotomy in ICU patients: a systematic review of randomized controlled trial. Crit Care. 2015;19:424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Andriolo BNG, Andriolo RB, Saconato H, et al. Early versus late tracheostomy for critically ill patients. Cochrane Database Syst Rev. 2015;1(1):CD007271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ji Y, Fang Y, Cheng B, et al. Tracheostomy timing and clinical outcomes in ventilated COVID-19 patients: a systematic review and meta-analysis. Crit Care. 2022;26(1):40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Boles JM, Bion J, Connors A, et al. Weaning from mechanical ventilation. Eur Respir J. 2007;29(5):1033–56. [DOI] [PubMed] [Google Scholar]
- 11.Beduneau G, Pham T, Schortgen F, et al. Epidemiology of weaning outcome according to a new definition. The WIND study. Am J Respir Crit Care Med. 2017;195(6):772–83. [DOI] [PubMed] [Google Scholar]
- 12.Ghiani A, Paderewska J, Sainis A, et al. Variables predicting weaning outcome in prolonged mechanically ventilated tracheotomized patients: a retrospective study. J Intensive Care. 2020;8:19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ghiani A, Paderewska J, Walcher S, et al. Mechanical power normalized to lung-thorax compliance indicates weaning readiness in prolonged ventilated patients. Sci Rep. 2022;12(1):6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ghiani A, Walcher S, Lutfi A, et al. Mechanical power density, spontaneous breathing indexes, and prolonged weaning failure: a prospective cohort study. Sci Rep. 2024;14(1):16297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sinha P, Fauvel NJ, Singh P, Soni N. Analysis of ventilatory ratio as a novel method to monitor ventilatory adequacy at the bedside. Crit Care. 2013;17:R34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Holevar M, Dunham JCM, Brautigan R, et al. Practice management guidelines for timing of tracheostomy: the EAST practice management guidelines work group. J Trauma. 2009;67(4):870–4. [DOI] [PubMed] [Google Scholar]
- 17.Whitmore KA, Townsend SC, Laupland KB. Management of tracheostomies in the intensive care unit: a scoping review. BMJ Open Respir Res. 2020;7(1):e000651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Gendreau S, Benelli B, Deliere M, et al. Partitioning mechanical ventilation duration in COVID-19-related acute respiratory distress syndrome. Am J Respir Crit Care Med. 2022;206(1):114–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Windisch W, Dellweg D, Geiseler J, et al. Prolonged weaning from mechanical ventilation. Dtsch Arztebl Int. 2020;117(12):197–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Park C, Ko RE, Jung J, et al. Prediction of successful decannulation of tracheostomized patients in medical intensive care units. Respir Res. 2021;22:131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ghiani A, Tsitouras K, Paderewska J, et al. Incidence, causes, and predictors of unsuccessful decannulation following prolonged weaning. Ther Adv Chron Dis. 2022;13:20406223221109655. [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.
Supplementary Materials
Data Availability Statement
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.