Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation.

OBJECTIVE: Reinstitution of mechanical ventilation (MV) for tracheostomized patients after successful weaning may occur as the care setting changes from critical care to general care. We aimed to investigate the occurrence, consequence and associated factors of MV reinstitution.
METHODS: We analyzed the clinical data and physiological measurements of tracheostomized patients with prolonged MV discharged from the weaning unit to general wards after successful weaning to compare between those with and without in-hospital MV reinstitution within 60 days.
RESULTS: Of 454 patients successfully weaned, 116 (25.6%) reinstituted MV at general wards within 60 days; at hospital discharge, 42 (36.2%) of them were eventually liberated from MV, 51 (44.0%) remained MV dependent, and 33 (28.4%) died. Of the 338 patients without reinstitution within 60 days, only 3 (0.9%) were later reinstituted with MV before discharge (on day 67, 89 and 136 at general wards, respectively), and 322 (95.2%) were successfully weaned again at discharge, while 13 (3.8%) died. Patients with MV reinstitution had a significantly lower level of maximal expiratory pressure (PEmax) before unassisted breathing trial compared to those without reinstitution. Multivariable Cox regression analysis showed fever at RCC discharge (hazard ratio [HR] 14.00, 95% confidence interval [CI] 3.2-61.9) chronic obstructive pulmonary disease (HR 2.37, 95% CI 1.34-4.18), renal replacement therapy at the ICU (HR 2.29, 95% CI 1.50-3.49) and extubation failure before tracheostomy (HR 1.76, 95% CI 1.18-2.63) were associated with increased risks of reinstitution, while PEmax > 30 cmH2O (HR 0.51, 95% CI 0.35-0.76) was associated with a decreased risk of reinstitution.
CONCLUSIONS: The reinstitution of MV at the general ward is significant, with poor outcomes. The PEmax measured before unassisted breathing trial was significantly associated with the risk of reinstituting MV at the general wards.

Introduction

Liberation from mechanical ventilation (MV) in a critical care setting remains challenging. About 10% of patients with acute respiratory failure may require prolonged MV, commonly defined as longer than 21 days [1]. These patients have a very poor prognosis, with a one-year survival rate between 40% and 50% [2]. Prolonged MV also imposes a significant care burden on intensive care units (ICUs) [3]. Successful weaning with sustainable independence from invasive MV is, therefore, pivotal to the management of patients with prolonged MV.

One of the problems encountered during MV liberation is the reinstitution of MV after the patient is transferred out the critical care setting after being deemed successfully weaned by a protocoled process [4,5], commonly based on an operational definition of 5 days of unassisted breathing. A study reported that reinstitutions occurred within 28 days in 52% of patients, indicating that enduring freedom could not be established until > 28 days had elapsed [5]. However, the associated clinical and physiological factors for the reinstitution are uncertain, and few reports have investigated the reinstitution of MV support in successfully weaned patients with prolonged MV [1, 5–7].

Traditionally, observing a patient’s clinical respiratory condition is used to define being successfully liberated from the ventilator during the last few days of a continuous unassisted breathing trial (UBT), but without the routine application of regular assessments of respiratory physiologic parameters. Measuring weaning parameters has been commonly applied in the extubation of intubated patients [8]; however, few studies have investigated its application in tracheostomized patients [9], and no studies have reported its use in assessing tracheostomized patients undergoing days of an unassisted breathing trial. While physiologic measurements remain difficult in non-intubated patients, patients with a tracheostomy tube may provide an opportunity to explore the respiratory mechanics during an unassisted breathing trial when clinicians need to assess the feasibility of transferring out the patients who are considered to have been successfully weaned from the ventilator.

In this study, we aimed to investigate whether weaning parameters were associated with the reinstitution of MV in patients who were successfully weaned from a ventilator in a protocoled weaning care setting.

Materials and methods

Design and setting

This retrospective study was conducted at the Respiratory Care Center (RCC), a dedicated weaning unit of National Taiwan University Hospital, between January 2016 and December 2018. This Respiratory Care Center has 15 beds and receives patients with prolonged MV from the ICUs of the same hospital. The Research Ethics Committee B of this hospital approved this study (#201902056RINB) and waived the need for informed consent from the patients.

The decision of weaning after admission to the Respiratory Care Center was made by the attending physician. The Respiratory Care Center uses a standardized weaning protocol that is comparable to that reported by Jubran et al. [4]. Briefly, the ventilator settings were gradually reduced to pressure support of 10 cmH2O and an end-expiratory positive pressure of 5 cmH2O for at least 8 hours to assure the patient’s tolerance. After the respiratory therapist had measured the physiological variables, the patient then underwent a screening procedure that consisted of unassisted breathing for 12 hours for 2 consecutive days with humidified oxygen delivered through a T-piece oxygen tube, The patients who did not develop distress during the screening period then underwent 5 consecutive days of unassisted breathing as a direct liberation trial. The patients who failed this screening process were subjected to a stepwise liberation trial, which consisted of daily increases in the duration of unassisted breathing starting from 2 hours, then extending to 2 hours twice daily, 4 hours daily, 4 hours twice daily, 8 hours daily, 12 hours daily, 16 hours daily, 20 hours daily, and finally continuous unassisted breathing for 5 days. If the patient repeatedly failed to complete the session, they underwent a slow weaning trial which consisted of either breathing through a T-piece but supported with an external positive airway pressure device with a gradual reduction in the pressure level, or stepwise liberation as with the stepwise trial, but repeating each duration of unassisted breathing for 3 to 5 days so the care team could assess the feasibility of further weaning with a longer duration of unassisted breathing and more extended time. The patients who tolerated the liberation process and the final 5 days of continuous unassisted breathing trial were transferred out of the Respiratory Care Center to the general ward for further care and preparation for hospital discharge. Those who failed the liberation process in at least two cycles of the liberation trial were transferred to the long-term respiratory care ward or to the general to manage the medical problems in addition to MV.

Patients

We included patients aged 20 years or older who were admitted to the Respiratory Care Center and were transferred out to a general ward between January 2016 and December 2018 after successful weaning from a ventilator, defined as at least 5 consecutive days of unassisted breathing without reinstitution of MV at the Respiratory Care Center. Patients with at least one of the following conditions were excluded: no data of weaning parameters after tracheostomy measured at the Respiratory Care Center or within 7 days before transfer to the Respiratory Care Center from an ICU; those with reinstitution of MV due to surgery or other elective interventions; those who received high-flow oxygen support or any device providing non-invasive ventilation at a general ward. We did not exclude those who received MV in the operating room during surgical interventions.

Data collection and variables

The following data were retrieved from the electronic medical records of the hospital: age, gender, dates of hospitalization, ICU and Respiratory Care Center admissions, dates of Respiratory Care Center discharge to a general ward and hospital discharge, in-hospital outcomes including survival status and ventilator liberation status, dates of initiation of MV, tracheostomy, initiation of the first and last session of the unassisted breathing trial weaning process, and reinstitution of MV, co-morbidities, main condition related to MV at the ICU, main intervention related to respiratory failure, acute physiology and chronic health evaluation (APACHE)- II score, number of extubation before tracheostomy, method of weaning for the session with weaning success, and data of weaning parameters. The weaning parameters included maximal inspiration pressure (PImax), maximal expiration pressure (PEmax), tidal volume, minute ventilation, respiratory rate, and rapid shallow breathing index (RSBI), and were measured by the respiratory therapist in the care team at the Respiratory Care Center. Measuring weaning parameters after tracheostomy is part of the usual care process at the Respiratory Care Center, and is routinely performed after the patients have been transferred to the Respiratory Care Center if not done within 7 days before the transfer. If the tracheostomy was performed at the Respiratory Care Center after the patient had been admitted, the measurements were performed within 3 days after the procedure before initiation of the active weaning process. The weaning parameters were measured every 14 days during the Respiratory Care Center stay and also before the patients were transferred to a general ward upon successful weaning from MV depending on the clinical condition of the patients. In this study, as the timing of initiating unassisted breathing trial was determined by clinicians who may not necessarily have recorded the measurements of weaning parameters, we only selected the data of weaning parameters closest to the beginning of unassisted breathing trial. This real-world data may have included measurements of weaning parameters for tracheostomized patients before they were transferred out of the Respiratory Care Center after days of un-interrupted unassisted breathing trial, however the decision to take the measurements was based on the judgement of clinicians and respiratory therapist, who may have provided expert opinions about further care at the general wards. The measurements were performed based on the standards of care by a respiratory therapist, as reported previously [10], including using a device with a unidirectional valve [11]. As only 12 patients had missing weaning parameter data, we decided to exclude them from sensitivity analysis. The primary outcome of this study was in-hospital MV reinstitution within 60 days. A patient was allocated to the ‘non-reinstitution’ group if the they died in the hospital without reinstitution of MV.

Statistical analysis

Demographic and clinical characteristics are expressed as mean ± standard deviation for continuous variables and number (percentage) for categorical variables. Comparisons between the reinstitution and non-reinstitution groups were performed using Fisher's exact test for categorical variables, and the independent sample t-test for continuous variables. Comparisons of paired data of weaning parameters were performed using the paired sample t-test. To investigate the associated factors of reinstitution of MV, we performed a multivariable Cox proportional hazard model within 60 days during the hospitalization. Variables entered in the multivariable Cox model were those variables whose significance less than 0.1 in the univariate analyses. To facilitate the clinical use of this multivariable analysis, several continuous variables were dichotomized in the multivariable analysis, including APACHE II score on ICU admission (by 25 scores), APACHE II score on Respiratory Care Center admission and at Respiratory Care Center discharge (by 15 scores) and PEmax (by 30 cmH2O). A two-sided p-value <0.05 was considered to be statistically significant, and no adjustment of multiple testing (multiplicity) was made in this study. All statistical analyses were performed using SPSS version 22.0 (SPSS, Chicago, IL, USA).

Results

Demographic and clinical characteristics and patient outcomes

From January 2016 to December 2018, a total of 1,069 patients were discharged from the Respiratory Care Center. Of them, 620 (58.0%) were successfully liberated from the ventilator and transferred to general wards, including 154 who were successfully extubated from an endotracheal tube and 466 who were liberated with a tracheostomy tube. Of the tracheostomized patients with successful weaning, 454 had at least one set of valid data of weaning parameters, and they constituted the analysis cohort of this study. Fig 1 depicts the patient inclusion flow diagram for this study (Fig 1). Of the included patients, MV was reinstituted within 60 days in 116 (25.6%) during their stay at a general ward, with a median interval of 8 days (range, 0–58 days). Of these 116 cases, 57 (49.2%) were reinstituted in less than 7 days after transfer to the general ward, 83 (71.6%) in less than 14 days. T hospital discharge, 42 (36.2%) of these 116 patients were eventually liberated from MV, 51 (44.0%) remained MV dependent, and 33 (28.4%) died. Of the 338 patients without reinstitution of MV within 60 days, only 3 (0.9%) were later reinstituted before discharge (on day 67, 89 and 136 after transfer to general wards, respectively), and 322 (95.2%) were discharged without the need for MV, while 13 (3.8%) died.

Fig 1

Flow diagram of patient inclusion and outcomes.

Table 1 summarizes the demographic and clinical characteristics of the 454 patients and comparisons between those with and without MV reinstitution. Overall, the patients were generally older, with 64.1% being older than 65 years and 13.9% being older than 85 years. Most of the patients were male, and co-morbidities were common. The most common conditions related to the use of MV in the ICU were pneumonia (55.3%), post-operative status (32.6%), and cerebral hemorrhage/head injury (31.7%). Almost 30% had experienced at least one episode of extubation failure before tracheostomy was performed. Nearly half of the patients were successfully weaned from MV using the direct liberation methods.

Table 1

Demographics and clinical characteristics of the 454 patients in this study.

Characteristic	Total (n = 454)	Reinstitution within 60 days		p-value
		Yes (n = 116)	No (n = 338)
Age, mean±SD	68.9 ± 16.2	71.2 ± 14.8	68.1 ± 16.6	0.08
Sex, male (%)	299 (65.9)	77 (66.4)	222 (65.7)	0.89
Co-morbidity
Hypertension	157 (34.6)	34 (29.3)	123 (36.4)	0.17
Congestive heart failure	147 (32.4)	41 (35.3)	106 (31.4)	0.43
Diabetes mellitus	123 (27.1)	33 (28.4)	90 (26.6)	0.70
Neurologic disease	114 (25.1)	34 (29.3)	80 (23.7)	0.23
Chronic kidney disease	113 (24.9)	39 (33.6)	74 (21.9)	0.01
Malignancy	112 (24.7)	38 (32.8)	74 (21.9)	0.02
Liver cirrhosis	46 (10.1)	14 (12.1)	32 (9.5)	0.42
COPD	36 (7.9)	16 (13.8)	20 (5.9)	0.01
Main conditions related to MV use
Pneumonia	251 (55.3)	73 (62.9)	178 (52.7)	0.06
Post-operation MV use	148 (32.6)	34 (29.3)	114 (33.7)	0.38
Cerebral hemorrhage/head injury	144 (31.7)	26 (22.4)	118 (34.9)	0.01
Sepsis	72 (15.9)	25 (21.6)	47 (13.9)	0.05
Heart failure	48 (10.6)	15 (12.9)	33 (9.8)	0.34
ARDS	20 (4.4)	7 (6.0)	13 (3.8)	0.32
Spinal cord injury	12 (2.6)	5 (4.3)	7 (2.1)	0.31
Admitted from medical ICU	242 (53.3)	62 (53.4)	180 (53.3)	0.97
Intervention at the ICU
Tracheostomy	444 (97.8)	115 (99.1)	329 (97.3)	0.25
Renal replacement therapy	88 (19.4)	38 (32.8)	50 (14.8)	<0.01
Inhaled nitric oxide	10 (2.2)	4 (3.4)	6 (1.8)	0.46
Prone positioning	7 (1.5)	2 (1.7)	5 (1.5)	1.00
ECMO	6 (1.3)	3 (2.6)	3 (0.9)	0.35
Extubation failure before tracheostomy	132 (29.1)	42 (36.2)	90 (26.6)	0.05
Extubation times before tracheostomy	1.32 ± 0.55	1.41 ± 0.60	1.29 ± 0.53	0.05
ICU days before RCC transfer	25.7 ± 17.0	25.6 ± 19.2	25.8 ± 16.3	0.91
RCC length of stay	15.5 ± 8.1	16.3 ± 8.6	15.2 ± 7.9	0.21
Total MV days	38.7±27.9	38.8 ± 24.2	38.6 ± 29.1	0.95
MV days at ICU	28.6±26.8	27.9 ± 23.2	28.9 ± 28.0	0.73
MV days at RCC	10.1±8.1	10.9 ± 8.7	9.8 ± 7.9	0.19
APACHE II score on admission to the ICU	23.9±7.4	25.6 ± 7.1	23.2 ± 7.4	<0.01
APACHE II score on admission to the RCC	14.9 ± 5.0	15.7 ± 4.9	14.6 ± 5.0	0.03
APACHE II score at RCC discharge	17.0 ± 4.4	18.0 ± 4.8	16.6 ± 4.2	<0.01
Heart rate at RCC discharge	87.0 ± 14.5	87.2 ± 15.7	86.9 ± 14.1	0.88
Respiratory rate at RCC discharge	20.7 ± 4.7	20.6 ± 5.0	20.7 ± 4.6	0.85
Mean blood pressure at RCC discharge	88.3 ± 12.4	87.8 ± 12.5	88.4 ± 12.4	0.63
Fever (BT>38.3 degree) at RCC discharge	2 (0.4)	2 (1.7)	0 (0.0)	0.07
Mode of weaning success				0.21
Direct liberation	199 (43.8)	43 (37.1)	156 (46.2)
Stepwise protocol	245 (54.0)	71 (61.2)	174 (51.5)
Slow weaning	10 (2.2)	2 (1.7)	8 (2.4)
In-hospital mortality	47 (10.4)	33 (28.4)	14 (4.1)	<0.01

APACHE II: acute physiology and chronic health evaluation II; ARDS: acute respiratory distress syndrome; BT: body temperature (Celsius); COPD: chronic obstructive pulmonary disease; ECMO: extracorporeal membrane oxygenation; ICU: intensive care unit; MV: mechanical ventilation; RCC: respiratory care center.

Many of the demographic and clinical characteristics were similar between the patients with and without MV reinstitution as shown in Table 1, however significant differences were observed in univariate analysis, including malignancy (p = 0.02), chronic obstructive pulmonary disease (p = 0.01), cerebral hemorrhage/head injury (p = 0.015), renal replacement therapy (p<0.01), and APACHE II scores on admission to the ICU (p<0.01), on admission to the Respiratory Care Center (p = 0.03) and at discharge from Respiratory Care Center (p = 0.004). The reinstituted group also had a significantly higher in-hospital mortality rate (28.4% vs. 4.1%, p<0.01).

Comparisons of clinical and physiologic variables

Table 2 summarizes the weaning parameter data of the included patients and comparisons between groups. Measurement of parameters was performed 9.7±7.9 days before the first day of continuous unassisted breathing trial. The patients without in-hospital reinstitution of MV within 60 days had similar intervals of parameter measurements and initiation of unassisted breathing trial (9.6±8.0 days vs. 9.9±7.7 days, p = 0.70, data not shown). In general, the 454 patients had satisfactory results before weaning, with 404 (89.0%) having a PImax better than -20 cmH2O, 306 (67.4%) with a PEmax better than 30 cmH2O, and 328 (72.2%) with a rapid shallow breathing index better than 105 (refer to Additional S2 File). The patients with MV reinstitution had a significantly higher level of PEmax than those without MV reinstitution (43±20 vs. 37±17; p<0.01). However, there were no significant differences in PImax, tidal volume, minute ventilation, respiratory rate, and RSBI between the two groups (Table 2).

Table 2

Comparisons of weaning parameters between the patients with and without mechanical ventilation reinstitution at a general ward.

Parameter	Total (n = 454)	Reinstitution within 60 days		p-value
		Yes (n = 116)	No (n = 338)
PImax, cmH2O	39 ± 13	37 ± 12	39 ± 14	0.19
PEmax, cmH2O	42 ± 19	37 ± 17	43 ± 20	<0.01
VT, mL	341 ± 124	334 ± 110	344 ± 128	0.49
VE, L	8.5 ± 3.0	8.3 ± 2.6	8.5 ± 3.1	0.46
RR, breaths/min	25.8 ± 7.3	25.6 ± 6.7	25.8 ± 7.5	0.78
RSBI, breaths/min/L	89.3 ± 52.2	88.6 ± 48.9	89.5 ± 53.3	0.88

PImax: maximal inspiratory pressure; PEmax: maximal expiratory pressure; VT: tidal volume; VE: minute ventilation volume; RR: respiratory rate; RSBI: rapid shallow breathing index.

Of the 454 patients, 290 (63.9%) also had additional measurements of weaning parameters upon transfer out to a general ward (S1 Table in S1 File). S2 Table S1 File shows comparisons of weaning parameter data before the unassisted breathing trial and after successful weaning of the 290 patients using the paired t-test. There were no significant evolutional changes in the data for each parameter (S2 Table in S1 File); therefore, we decided to include only weaning parameters before the unassisted breathing trial for consideration in further multivariable regression analysis.

Multivariable analysis for factors associated with time to reinstitution of MV

Table 3 summarizes the results of multivariable Cox regression analysis, in which the time to reinstitution of MV within 60 days during the hospitalization was the outcome variable. The Cox regression analysis showed that fever at RCC discharge (hazard ratio [HR] 14.00, 95% confidence interval [CI] 3.2–61.9), chronic obstructive pulmonary disease (HR 2.37, 95% CI 1.34–4.18), renal replacement therapy at the ICU (HR 2.29, 95% CI 1.50–3.49) and extubation failure before tracheostomy (HR 1.76, 95% CI 1.18–2.63) were associated with increased risks of reinstitution, while PEmax > 30 cmH2O (HR 0.51, 95% CI 0.35–0.76) was associated with a decreased risk of reinstitution.

Table 3

Multivariable Cox regression analysis of factors associated with time to reinstitution of MV within 60 days during the hospitalization.

Variable	HR (95% CI)	p-value
Age, year	1.00 (0.98–1.01)	0.96
Malignancy	1.44 (0.95–2.18)	0.09
COPD	2.37 (1.34–4.18)	<0.01
Pneumonia	1.19 (0.74–1.91)	0.47
Cerebral hemorrhage/head injury	1.07 (0.61–1.87)	0.81
Sepsis	1.09 (0.68–1.75)	0.72
Renal replacement therapy at ICU	2.29 (1.50–3.49)	<0.01
Extubation failure before tracheostomy	1.76 (1.18–2.63)	<0.01
APACHE II score on ICU admission ≥ 25	1.38 (0.93–2.07)	0.11
APACHE II score on RCC admission ≥ 15	1.40 (0.93–2.10)	0.10
APACHE II score at RCC discharge ≥ 15	1.30 (0.76–2.22)	0.35
Fever (BT>38.3 degree) at RCC discharge	14.00 (3.2–61.9)	<0.01
PEmax > 30 cmH2O	0.51 (0.35–0.76)	<0.01

APACHE II: acute physiology and chronic health evaluation II; BT: body temperature; CI: confidence interval; COPD: chronic obstructive pulmonary disease; HR: hazard ratio; MV: mechanical ventilation; ICU: intensive care unit; PEmax: maximal expiratory pressure; RCC: respiratory care center.

Fig 2 shows the Kaplan-Meier survival curves of the outcome of in-hospital MV reinstitution within 60 days. Patients with PEmax > 30 cmH2O had a significantly lower probability of reinstitution during their stay at a general ward before discharge from the hospital (log-rank test, p<0.01) (Fig 2).

Fig 2

Kaplan-Meier plot for the reinstitution of MV within 30 days, grouped by before-weaning PEmax > 30 cmH2O.

Discussion

In this study, we found that the tracheostomized patients with prolonged MV had a substantial probability of MV reinstitution at a general ward despite being liberated from a ventilator according to predetermined criteria at a dedicated weaning unit and that these patients had poor outcomes at a general ward. While several clinical characteristics were associated with a higher risk of MV reinstitution at a general ward, we also found that PEmax was significantly associated with in-hospital MV reinstitution.

The reinstitution of MV can be a significant issue as it requires modification of the care setting. In this study, the definition of not requiring MV support for at least 5 days to determine successful weaning is compatible with the literature [5]. While patients requiring prolonged MV generally had a poorer long-term prognosis [12], our findings of 26% of reinstitution rate with a higher in-hospital mortality rate (29% vs. 4%) were compatible with previously reported that reinstitution within 14 days was a poor predictor for prolonged MV patients after successful weaning, and physicians should closely monitor such patients [6].

Our finding that PEmax consistently served as a significant factor associated with developing a respiratory condition requiring MV reinstitution has not previously been reported in the literature. Although multiple comorbidities and in-hospital clinical condition contribute to an increased risk of reinstituting MV, for tracheostomized patients, there might be a common respiratory condition which increases the risk of further developing respiratory failure. While a previous study suggested that a more normal rapid shallow breathing index and static compliance were associated with better weaning outcome and prolonged survival [13], other researchers showed controversial results [14]. Further reported potential clinical and physiologic predictors for successful weaning from prolonged MV might include hypercapnic ventilatory response [15], duration of stay at the weaning unit, blood urea nitrogen levels, modified Glasgow coma scores, serum albumin, and PImax levels [16], trans-diaphragmatic pressure and tension-time index of the diaphragm [17]. A possible explanation why PEmax has rarely been described as a prognostic factor for MV reinstitution may be that most studies on ventilator weaning have focused on extubation success, with weaning parameters being measured before the endotracheally intubated patients proceed with a spontaneous breathing trial for extubation.

Assessments of weaning parameters traditionally focus on PImax and rapid shallow breathing index, which also focus mainly on inspiratory muscle strength and lung mechanics rather than expiratory muscle strength. Our findings suggest that the patients who succeeded in the 5-day unassisted breathing trial assessment generally had adequate inspiratory function, and that the critical factor for the reinstitution of MV at a general ward may have been expiratory strength or cough function, which may not have been identified in clinical observations of breathing condition during the 5-day unassisted breathing trial at the weaning unit. The measurement of PEmax in this study conformed to standard procedures as previously validated [10], although the reference value for PEmax has not been as frequently discussed in the literature as PImax [18]. As it may have been difficult to perform the measurements due to poor cooperation and respiratory distress during the testing, it is also possible that this suboptimal testing condition suggests a deterioration in their respiratory function such as an inability to remove lower airway secretions or to cough out aspirated material from the upper airway. PEmax, therefore, may determine the ability of airway clearance, which contributes to the outcome in patients staying in general wards, which are equipped with fewer human resources and monitoring facilities. We also found that the probability of MV reinstitution was not related to the weaning methods in this study. This also suggests that under observation during the 5-day assessment period, the strength of inspiratory muscles may have determined the weaning success in most of the patients without concerns of early transfer out. Our findings may provide further insights into the choice of weaning method based on the literature that unassisted breathing, compared with pressure support, can result in a shorter median weaning time, and that weaning mode does not affect survival at 6 and 12 months [4].

Our finding that PEmax contributed to MV reinstitution has several clinical implications. Understanding the presence of suboptimal PEmax and lowered reserve during continuous unassisted breathing trial may allow the general ward for the initiation of proactive care processes such as intermittent manual hyperinflation or intermittent positive pressure breathing without significantly scaling up the care setting. Furthermore, the detection of suboptimal PEmax may allow for the initiation of management strategies to stabilize the patient’s condition and prevent future reinstitutions. Compared with inspiratory muscle training [19, 20], few studies have explored expiratory muscle training in terms of prolonged weaning success [21, 22]. As this study is retrospective, we recommend further prospective investigations about the clinical usability of serial and even continuous monitoring of respiratory mechanics during the period before the patients are transferred out of a critical care setting.

There are several limitations to this study. First, there is potential bias due to the single-center retrospective study design, such as the exclusion of patients without weaning parameter data, who may have had different clinical and physiologic features from the analysis cohort. Second, the decision to reconnect MV at the general ward was made mainly by attending physicians or residents rather than intensivists or respiratory therapists. However, as we showed, these reconnected patients had a poor prognosis compared to those who were not reconnected. Therefore, we believe this routine care practice reflected actual patient conditions in that the need for MV reinstitution was made from an intensivist’s point of view. Third, we only assessed tracheostomized patients with prolonged MV; therefore, the condition of the patients with reinstitution more than 5 days after they had been extubated could not be analyzed. However, measurements of physiologic parameters in non-tracheostomized patients could be more complicated than in tracheostomized patients once they have been liberated from a ventilator. Fourth, we did not explore the specific clinical conditions of these patients at general wards after they had been transferred out of the weaning unit; therefore, it was difficult to assess the effect of newly developed non-respiratory events on recurrent respiratory failure. Fifth, we did not perform a follow-up study of the patients who were discharged MV free; therefore, the possibility of recurrent MV could not be excluded in the outpatient setting. Sixth, as this was a retrospective study and we retrieved real-world data from the medical records, we were not able to obtain rigorous measurement data within a fixed short interval before the patients proceeded to whole-day unassisted breathing. Although we consider that the weaning parameters over several days might be able to represent the respiratory muscle condition through the course of unassisted breathing trials, prospective studies examining the evolution of weaning parameters before and during the weaning process is needed for the generalization of our findings. Last, the definition of reinstitution might require further study, based on further evidence relating to the relevance of patient status upon transfer to general ward and the event leading to MV reinstitution. The criteria of 60-day reinstitution, despite including most except for three patients, still need further validation.

Conclusions

In conclusion, there was a high probability of the reinstitution of MV at a general ward despite successful protocoled weaning at the weaning unit in this study, with poor weaning outcomes at a general ward. PEmax measured before the unassisted breathing trial was significantly associated with the risk of reinstituting MV at a general ward. Further studies are needed to confirm the relevance of expiratory muscle strength and cough function with regards to patient outcomes after they have been successfully weaned based on the operational criteria.

Supplementary information.

Additional information regarding the results of the study.

(DOCX)

Click here for additional data file.

Patient data.

Data of the cases included in the analysis of this study.

(XLSX)

Click here for additional data file.

30 Oct 2019

PONE-D-19-28222

Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation

PLOS ONE

Dear Dr. Jerng,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Both reviewers raised several concerns, especially regarding the statistical analysis, data reporting, and interpretations. The authors need to effectively respond to these comments in their revised manuscript.

We would appreciate receiving your revised manuscript by Dec 14 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Yu Ru Kou, PhD

Academic Editor

PLOS ONE

Journal Requirements:

1.  When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Manuscript: D-19-28222

Title: Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation

Abstract:

The authors retrospectively investigated the occurrence and consequence of MV reinstitution and its associated factors in tracheostomized patients after successful unassisted breathing trials and discharged from RCC. The authors compared the study group (with MV reinstitution) with the control group (without reinstitution before hospital discharge)

Major comments:

1. In Abstract, the statement of “…similar PImax, tidal volume, minute ventilation, respiratory rate, and rapid shallow breathing index before unassisted breathing trial with those without reinstitution.” need to be reedited to “compared with those without reinstitution”.

2. In abstract, the authors reported the results of Cox regression analysis for detecting the factors associated with MV reinstitution within 60 days in hospital. There are two criteria (MV reinstitution before discharge and within 60 days in hospital) to categorize the study patients into the study and control group. Which one is the end-point for Cox regression analysis?

3. In conclusion, the authors declare that “Interventions to enhance expiratory strength in tracheostomized patients are warranted.”. However, in the current study, the study design is retrospective observational and there is no intervention to prove this concept.

4. In Data Collection, the timing for weaning parameters measurement is not clear and consistent during daily protocolized weaning process. Is it measured for (1). initiating unassisted breathing for 12 hours for 2 consecutive days during screening period, (2).direct liberation trial with UBT for 5 days, (3). stepwise liberation, (4). every 14 days during the RCC stay, (5). before patients transferred to ordinary ward. If patients had several measurements of weaning parameter for different purpose, which one is selected to be analyzed?

5. In Materials and Methods, there is no primary outcome measurement mentioned. There is no study design for how to group patients. If the study patient is expired, how to define the group for the case?

6. In Results, the authors reported that the median interval for reinstitution of MV was 8.0days, and 43.2% reinstitutions occurred in <7days after transfer to the general ward. The mean length of stay in RCC was 15.5days, and the authors investigate the reinstitution of MV after RCC discharge. The conditions on RCC discharge have a critical determinant on MV reinstitution in ordinary ward. However, only ICU conditions are showed on Table 1 and analyzed for reinstitution of MV in ordinary ward.

7. In Results, the authors reported Table 1S for 290 cases with weaning parameters measured after successful weaning upon transfer out to a general ward. What is the reason to do it in clinical practice?

8. In Table 3, the method 1 and method 2 are performed for MV reinstitution before discharge and 60 days before discharge. If the study case is connected to MV after RCC discharge, but liberation from MV before hospital discharge. How to define this patient? This classification in Table 3 (within 60 days) for the study patients is different from that in Table 1 (Reinstitution and non-reinstitution before discharge).

9. In table 1, the MV days before RCC transfer is not presented.

10. In Discussion, paragraph 3, the authors mentioned the factors associated with unsuccessful weaning. However, in the current study, the targeted study population is the ones with the reinstitution of mechanical ventilation after successful unassisted breathing trials.

11. In Discussion, paragraph 4, the authors mentioned that “although the reference value for PEmax has not been as frequently discussed in the literature as PImax “. In this study, the values of PEmax in non-reinstituion and reinstitution group are 43.5cmH2O and 37.4 cmH2O (p = 0.001). Is the 6 cmH2O difference of PEmax clinically meaningful?

12. In Discussion, the paragraph 5 and 6 are redundant, Please reedit the content of discussion to focus on the main findings (factors associated with reinstituion of MV) in the current study compared to previous literature.

13. In figure 2, it seems rapid reinstitution of MV after RCC discharge in Pemax <=30cmH2O compared to Pemax >30cmH2O. why is the cut-off value of 30cmH2O is selected and how to explain it?

Reviewer #2: I have read with interest the article entitled “Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation” by Lin et al. This is an interesting topic, which has been poorly explored in the current literature. However, I have comments that need to be addressed before further proceeding.

Major comments.

1) Abstracts. Check the instructions for authors. I don’t think it is relevant to describe the univariate analysis results in this section. Please provide data about outcome, which are missing.

2) Statistical analysis. Cox Model. Can you add on which parameters you adjusted the model, and how these factors were selected? Since there is an important imbalance between groups, this aspect is paramount.

3) Can you provide SAPS II or APACHE at ICU admission? If baseline severity between groups is different, this point could explain in itself the difference in the outcomes between groups.

4) Methodology. Can the authors describe their selection of patients with paired sets of data? I don’t exactly understand what is meant by “paired data sets”? This issue appears only in the results and is not described elsewhere. There is a potential selection bias that needs to be clearly addressed in the methodology: there are less cardiac co-morbidities and neurologic insults in Group B. This issue must be further explained and detailed in the statistical analysis section: how did the authors cope with such issue?

5) Please provide OR, confidence of Interval and p value for multivariable analysis

6) Authors must describe the variable selection process regarding logistic regression and Cox model. I believe that a methodologist should help in the process

7) The MV duration > 30 days as a protective factor of weaning failure is counter-intuitive and not in line with previous data (Beduneau AJRCCM, Funk ERJ) showing that the duration of MV is a risk factor of weaning failure. Can the authors attempt to explain this finding?

8) I am unsure that the logistic regression model obeys the parcimonious rule (ie 1 variable in the model per 5-10 events maximum). The same question arises for the Cox model. Again, a statistical reviewing seems mandatory. Moreover, these models seem redundant (same risk factors)? What is the new information provided by the Cox model? Did the authors study the same outcome?

9) In the end of the results section, the authors provide new data in a 290 patients’ sample with paired data. What is the difference with the previous sample? These data must be clearly expressed in the Statistical analysis section.

10) An effort should be made, in order to reorganize the results section

Minor comments.

1) I believe an English editing, by a native English speaker is mandatory

2) The introduction should be more straightforward. The discussion section must be shortened.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Hsin-Kuo Ko

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

28 Dec 2019

Responses to Reviewer 1 Comments

Major comments:

1. In Abstract, the statement of “…similar PImax, tidal volume, minute ventilation, respiratory rate, and rapid shallow breathing index before unassisted breathing trial with those without reinstitution.” need to be reedited to “compared with those without reinstitution”.

Response: We have revised this sentence by adding ‘compared’ to it as suggested. Please refer to Abstract of the revised manuscript for the change.

2. In abstract, the authors reported the results of Cox regression analysis for detecting the factors associated with MV reinstitution within 60 days in hospital. There are two criteria (MV reinstitution before discharge and within 60 days in hospital) to categorize the study patients into the study and control group. Which one is the end-point for Cox regression analysis?

Response: To avoid confusion, we have revised the manuscript to focus on “in-hospital MV reinstitution within 60 days” (referred as ‘reinstitution’) as the primary end-point of this study. Therefore, all of the results were rearranged for this end-point. For Cox regression, we only show the result related to this “in-hospital MV reinstitution within 60 days” end-point. Please refer to page 9 of Methods, Table 3 and Results sections the revised manuscript for the changes.

3. In conclusion, the authors declare that “Interventions to enhance expiratory strength in tracheostomized patients are warranted.”. However, in the current study, the study design is retrospective observational and there is no intervention to prove this concept.

Response: We have revised this section and replaced the original sentence with ‘Further studies are needed to confirm the relevance of expiratory muscle strength and cough function with regards to patient outcomes after they have been successfully weaned based on the operational criteria.’ Please refer to Conclusions on Page 21 of the revised manuscript for the change.

4. In Data Collection, the timing for weaning parameters measurement is not clear and consistent during daily protocolized weaning process. Is it measured for (1). initiating unassisted breathing for 12 hours for 2 consecutive days during screening period, (2).direct liberation trial with UBT for 5 days, (3). stepwise liberation, (4). every 14 days during the RCC stay, (5). before patients transferred to ordinary ward. If patients had several measurements of weaning parameter for different purpose, which one is selected to be analyzed?

Response: The timing for weaning parameter measurement was always before the initiation of unassisted breathing trial (UBT). We have added a description to explain that in this study, as the timing of initiating UBT was determined by the clinicians not necessarily linked to the measurement of weaning parameters, we only selected the data of weaning parameters most close to the beginning of UBT. We have provided the information in the description of the Results section that the measurement of parameters was performed on 9.7±7.9 days before the first day of continuous UBT. Please refer to page 8 of Methods section and page 13 of Results section for the change.

5. In Materials and Methods, there is no primary outcome measurement mentioned. There is no study design for how to group patients. If the study patient is expired, how to define the group for the case?

Response: As we have above responded, we have revised the manuscript to focus on “in-hospital MV reinstitution within 60 days” as the primary end-point of this study. A patient was allocated to the “non-reinstitution” group if the patient died in the hospital without reinstitution of MV. Please refer to page 9 of Methods, Table 3 and Results sections the revised manuscript for the changes.

6. In Results, the authors reported that the median interval for reinstitution of MV was 8.0days, and 43.2% reinstitutions occurred in <7days after transfer to the general ward. The mean length of stay in RCC was 15.5days, and the authors investigate the reinstitution of MV after RCC discharge. The conditions on RCC discharge have a critical determinant on MV reinstitution in ordinary ward. However, only ICU conditions are showed on Table 1 and analyzed for reinstitution of MV in ordinary ward.

Response: We have retrieved relevant physiologic data for the calculation of APACHE II scoring for patients on the day they were transferred out of the RCC. The calculated data for APACHE II at RCC discharge as well as vital signs are added to Table 1. With the multivariable Cox regression, we have included the APACHE II scores on ICU admission, on RCC admission and at RCC discharge for the analysis, and the results, as shown in Table 3, did not show a significant association with the risk of MV reinstitution. Please refer to Table 1 and page 15 in Results for the changes.

7. In Results, the authors reported Table 1S for 290 cases with weaning parameters measured after successful weaning upon transfer out to a general ward. What is the reason to do it in clinical practice?

Response: The real-world practice in our institution might include the measurement of weaning parameters for tracheostomized before they are transferred out the RCC after days of un-interrupted UBT, in case there is a concern for reconnection to MV at the general ward; however, the decision for measuring was based on the judgement of clinicians and respiratory therapist, who might provide expert opinions about further care at the general wards. We have added this description to the Methods section. There was no significant evolutional change in the data for each parameter of these 290 patients (Table 2S); therefore, we decided to include only weaning parameters before the unassisted breathing trial for consideration of multivariable Cox regression analysis. Please refer page 9 in the Methods section for the change.

8. In Table 3, the method 1 and method 2 are performed for MV reinstitution before discharge and 60 days before discharge. If the study case is connected to MV after RCC discharge, but liberation from MV before hospital discharge. How to define this patient? This classification in Table 3 (within 60 days) for the study patients is different from that in Table 1 (Reinstitution and non-reinstitution before discharge).

Response: The outcomes regarding MV reinstitution at the general ward were described in the Results on pages 10-11. For better display of the information, we have rearranged the Results and Tables. We have revised Tables 1 and 2 to show the comparison of patients with and without in-hospital MV reinstitution within 60 days, whereas the patients who had reinstitution of MV were allocated to the ‘reinstitution’ group. In Table 3, we now only show the results for Cox regression using the time-dependent outcome of “in-hospital reinstitution within 60 days” to avoid confusion to the readers. Please refer to pages 10-11, Tables 1, 2 and 3 for the changes.

9. In table 1, the MV days before RCC transfer is not presented.

Response: To Table 1 we have added the data relevant to MV use, including total days of MV use, ICU MV days, and RCC MV days. These data were similar between the reinstitution and non-reinstitution groups of analyses. Please refer to Table 1 in the Results section for the change.

10. In Discussion, paragraph 3, the authors mentioned the factors associated with unsuccessful weaning. However, in the current study, the targeted study population is the ones with the reinstitution of mechanical ventilation after successful unassisted breathing trials.

Response: As there was limit reports in the literature regarding the factors associated with MV reinstitution in tracheostomized patients with prolonged MV, we have shortened this paragraph to avoid confusion to the readers. Please refer to the Discussion section of the revised manuscript for the change.

11. In Discussion, paragraph 4, the authors mentioned that “although the reference value for PEmax has not been as frequently discussed in the literature as PImax “. In this study, the values of PEmax in non-reinstitution and reinstitution group are 43.5cmH2O and 37.4 cmH2O (p = 0.001). Is the 6 cmH2O difference of PEmax clinically meaningful?

Response: As this study was conducted retrospectively, we recognized that the patient population in this study was heterogenous that multiple factors affected the outcomes. The data in Results and Table 3 support that factors other than PEmax affected the outcome independently, such as co-morbidities, complications at the ICU, severity of organ dysfunction at the ICU, and MV duration. Therefore, the univariate comparison of PEmax between these two groups (reinstitution vs non-reinstitution) might not infer the independent significance of association, or preclude this factor for multivariable analysis.

12. In Discussion, the paragraph 5 and 6 are redundant, Please reedit the content of discussion to focus on the main findings (factors associated with reinstitution of MV) in the current study compared to previous literature.

Response: We have reduced the content of these paragraphs to make the Discussion more concise. Please refer to Discussion of the revised manuscript for the changes.

13. In figure 2, it seems rapid reinstitution of MV after RCC discharge in Pemax <=30cmH2O compared to Pemax >30cmH2O. why is the cut-off value of 30cmH2O is selected and how to explain it?

Response: There were limited reports in the literature regarding the reference PEmax value for predicting outcomes for tracheostomized patients. For clinical practice, we also chose to test the discriminating power by the nearest ten value; therefore, we have also performed Cox regression analysis to use PEmax > 20 cmH2O to replace PEmax > 30 cmH2O as the cut-off value, and found that the protective significance for MV reinstitution was lost; therefore, we chose PEmax > 30 cmH2O as the cut-off value. We also addressed this point in the ‘Statistical analysis’ subsection of the ‘Methods’ section.

Responses to Reviewer 2 Comments

Major comments.

1) Abstracts. Check the instructions for authors. I don’t think it is relevant to describe the univariate analysis results in this section. Please provide data about outcome, which are missing.

Response: We have revised the Abstract section to remove the univariate analysis and add outcome descriptions. Please refer to the revised manuscript for the change.

2) Statistical analysis. Cox Model. Can you add on which parameters you adjusted the model, and how these factors were selected? Since there is an important imbalance between groups, this aspect is paramount.

Response: As we had found that the patient population was very heterogeneous, therefore we selected the variables with a p-value of less than 0.1 into the multivariable Cox regression. As we have re-performed the Cox regression analysis, the final model now contains 13 variables, with basically PEmax > 30 cmH2O remaining as significant protecting factor for MV reinstitution. Please refer to the revised manuscript at Table 3. We also addressed how to select these factors in the ‘Statistical analysis’ subsection of the ‘Methods’ section.

3) Can you provide SAPS II or APACHE at ICU admission? If baseline severity between groups is different, this point could explain in itself the difference in the outcomes between groups.

Response: We have added the results of APACHE II to Table 1, which now contains APACHE II scores on ICU admission, on admission to RCC and at RCC discharge, and the comparisons between the reinstitution and non-reinstitution groups. Please refer to the Results and Table 1 in the revised manuscript. Although the reinstitution group had higher APACHE II scores on ICU admission, on admission to RCC and at RCC discharge scores in univariable analysis, the multivariable Cox analysis showed that none of them were independent factors of risk of reinstitution of MV.

4) Methodology. Can the authors describe their selection of patients with paired sets of data? I don’t exactly understand what is meant by “paired data sets”? This issue appears only in the results and is not described elsewhere. There is a potential selection bias that needs to be clearly addressed in the methodology: there are less cardiac co-morbidities and neurologic insults in Group B. This issue must be further explained and detailed in the statistical analysis section: how did the authors cope with such issue?

Response: The real-world practice in our institution might include the measurement of weaning parameters for tracheostomized before they are transferred out the RCC after days of un-interrupted UBT, in case there is a concern for reconnection to MV at the general ward; however, the decision for measuring was based on the judgement of clinicians and respiratory therapist, who might provide expert opinions about further care at the general wards. We have added this description to the Methods section. As further statistical analysis did not show new findings, we decided to remove the statement of statistical analysis for the weaning parameter measurement at RCC discharge. Please refer page 9 in the Methods section for the change.

5) Please provide OR, confidence of Interval and p value for multivariable analysis

Response: As in Table 3 the 95% CIs and p values have been provided, we have added hazard ratio (HR), 95%CI and p values to the Abstract for Cox regression. Please refer to the Abstract of revised manuscript for the change.

6) Authors must describe the variable selection process regarding logistic regression and Cox model. I believe that a methodologist should help in the process

Response: As has been provided in the response to Comment #2, we have provided the description of variable selection based on our initial findings. Please refer to our response to Comment #2 and the ‘Statistical analysis’ subsection of the ‘Methods’ section. To avoid confuse the readers, we decided to obtain the result of Cox model and remove the logistic regression version. We did actually consult a senior statistician and we have also acknowledged for his help in the ‘Acknowledgments’ section of the revised manuscript

7) The MV duration > 30 days as a protective factor of weaning failure is counter-intuitive and not in line with previous data (Beduneau AJRCCM, Funk ERJ) showing that the duration of MV is a risk factor of weaning failure. Can the authors attempt to explain this finding?

Response: We agree that previous reports, mainly focusing on overall weaning prognosis assessment, concluded that prolonged MV > 30 days as a negative predictor for successful weaning. However, as the revised Cox regression analysis did not show MV > 30 days as a significant factor, we decide not to include this factor in the discussion. Please refer to Table 3 of the revised manuscript.

8) I am unsure that the logistic regression model obeys the parsimonious rule (ie 1 variable in the model per 5-10 events maximum). The same question arises for the Cox model. Again, a statistical reviewing seems mandatory. Moreover, these models seem redundant (same risk factors)? What is the new information provided by the Cox model? Did the authors study the same outcome?

Response: As we have rearranged the results to focus on 60-day in-hospital MV reinstitution, we have reduced the descriptions of multivariable analysis to only Cox regression analysis. The process of choosing variables is provided in the ‘Statistical analysis’ subsection of the ‘Methods’ section.

A total of 13 candidate predictors were introduced into the multivariable Cox regression and we had 116 events (reinstitution of MV within 60 days), therefore the final presentation in Table 3 would conform to the parsimonious rule. Please refer to the Methods section and Table 3 of the revised manuscript for the changes. In addition, as mentioned at comment#7, we did actually consult a senior statistician and we have also acknowledged for his help in the ‘Acknowledgments’ section of the revised manuscript

9) In the end of the results section, the authors provide new data in a 290 patients’ sample with paired data. What is the difference with the previous sample? These data must be clearly expressed in the Statistical analysis section.

Response: We have added descriptions regarding the methods for assessing the paired data mentioned in the Supplement materials. There was no significant evolutional change in the data for each parameter of these 290 patients (Table 2S); therefore, we decided to include only weaning parameters before the unassisted breathing trial for consideration of multivariable regression analysis. Please refer to the Methods sections of the revised manuscript for the changes.

10) An effort should be made, in order to reorganize the results section

Response: We have reorganized the Results section for better readability. Please refer to the Results sections of the revised manuscript for the changes.

Minor comments.

1) I believe an English editing, by a native English speaker is mandatory

Response: The manuscript had actually been edited by a native English-speaker. We have repeated this English editing process for the revised manuscript.

2) The introduction should be more straightforward. The discussion section must be shortened.

Response: We have thereby revised these sections to provide more readability. Please refer to the Introduction and Discussion sections of the revised manuscript for the changes.

Submitted filename: Responses to Reviewer Comments.docx

Click here for additional data file.

14 Jan 2020

PONE-D-19-28222R1

Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation

PLOS ONE

Dear Dr. Jerng,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Both reviewers continued to raise some concerns. The reviewer 1 is very critical and the authors must effectively respond to his/her comments. Please note that the final acceptance of the submission needs to get the approval from both reviewers.

We would appreciate receiving your revised manuscript by Feb 28 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Yu Ru Kou, PhD

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: 1. The authors revised the in-hospital MV reinstitution within 60 days” (referred as ‘reinstitution’) as the primary end-point of this study. Why is the 60 days determined to evaluate the ventilator outcome? The PEmax before unassisted breathing trial predicts the 60-day outcome of hospitalization. Is it reasonable?

2. About the issue of weaning parameters measurement, the authors replied that “We have added a description to explain that in this study, as the timing of initiating UBT was determined by the clinicians not necessarily linked to the measurement of weaning parameters, we only selected the data of weaning parameters most close to the beginning of UBT.”. The timing of UBT is determined by physicians. The parameters measurement and UBT are two separately events during ventilator care, and a variable and prolonged period was noted (9.7±7.9 days). It is quite difficult for physicians to apply the results of the current study on patient care. The main problem is which one of measured PEmax close to UBT is unable to be determined prospectively. Before the initiation of UBT determined by physicians, whether the PEmax measured by RT today would be the one to predict the future 60-day ventilator outcome remains unknown.

Reviewer #2: I believe my comments have been adequately addressed. I have minor 3 comments:

-There are numerous abbreviations (PMV, UBT etc...). This is a bit misleading for the reader and I don't thinnk it is mandatoy ot have so many. Please check this aspect in order to improve the lisibility.

-Please chekc the instructions for authors. In the Tables, the p-values have many numbers after the coma. I believe that false accuracy should be avoided.

-I believe the Discussion is still too long and should be shortened.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

8 Feb 2020

Responses to Reviewer 1 Comments

1. The authors revised the in-hospital MV reinstitution within 60 days” (referred as ‘reinstitution’) as the primary end-point of this study. Why is the 60 days determined to evaluate the ventilator outcome? The PEmax before unassisted breathing trial predicts the 60-day outcome of hospitalization. Is it reasonable?

Response: Thank you for the comments. Our rationale for the primary outcome was that if a patient was reconnected to the ventilator before discharge, the patient would then need a care setting especially for the use of mechanical ventilation, and this might have a significant influence on the burden of care and patient prognosis. In addition, there was a gradual reduction in the chance to reinstitute MV (page 10 of the Results section), but it was difficult to find a time earlier than 60 days to determine the non-institution group, as this could have resulted in bias due to dichotomized grouping. Therefore, we originally preferred to include all cases of in-hospital reinstitution of MV as the primary outcome. However, as some of the patients in this study cohort actually stayed in the hospital for a period beyond 60 days, in whom the influence of physiologic status upon RCC discharge may have been uncertain even if they were later reinstituted, we chose 60 days as a censor time point. In fact, only three patients reinstituted MV beyond 60 days as they remained hospitalized (who were allocated as the non-reinstitution group in this study, see Page 11), and we considered that this may be reasonable given that the 60-day censor time included most of the patients based on our rationale, and that this cut-off time may have resulted in the least influence on grouping bias of reinstitution. Please refer to the Discussion section of the revised manuscript for this limitation.

2. About the issue of weaning parameters measurement, the authors replied that “We have added a description to explain that in this study, as the timing of initiating UBT was determined by the clinicians not necessarily linked to the measurement of weaning parameters, we only selected the data of weaning parameters most close to the beginning of UBT.”. The timing of UBT is determined by physicians. The parameters measurement and UBT are two separately events during ventilator care, and a variable and prolonged period was noted (9.7±7.9 days). It is quite difficult for physicians to apply the results of the current study on patient care. The main problem is which one of measured PEmax close to UBT is unable to be determined prospectively. Before the initiation of UBT determined by physicians, whether the PEmax measured by RT today would be the one to predict the future 60-day ventilator outcome remains unknown.

Response: It is our understanding that weaning parameters in tracheostomized patients with prolonged MV can be used for multiple purposes, such as the decision on whether to initiate active lowering of MV settings, and the speed and tempo for protocoled weaning involving unassisted breathing trials interspersed with MV support. Therefore, the measurement of weaning parameters usually took place several days before the patient actually started the whole-day unassisted breathing trials for 5 successive days. As this was a retrospective study and we retrieved real-world data from the medical records, we were not able to obtain rigorous measurement data within a fixed short interval before the patients proceeded to whole-day unassisted breathing. In addition, we compared the weaning parameters before and after whole-day unassisted breathing trials in 290 patients (64% of the study cohort) with valid paired data, as describe on Page 14, and found that the available paired weaning parameters did not show significant changes even after the patients were deemed to have been successfully liberated from the ventilator. Although we consider that the weaning parameters over several days might be able to represent the respiratory muscle condition through the course of unassisted breathing trials, we agree that a prospective study examining the evolution of weaning parameters before and during the weaning process is needed for the generalization of our findings. Please refer to the Discussion section of the revised manuscript for the limitations of this study.

Thank you so much for your very constructive opinions and comments, these points have greatly enhanced the quality of our manuscript.

Responses to Reviewer 2 Comments

1. There are numerous abbreviations (PMV, UBT etc...). This is a bit misleading for the reader and I don't think it is mandatory to have so many. Please check this aspect in order to improve the lisibility.

Response: Thank you for your comments. We have revised the manuscript to reduce the use of abbreviations. Please refer to the revised manuscript for the changes.

2. Please check the instructions for authors. In the Tables, the p-values have many numbers after the coma. I believe that false accuracy should be avoided.

Response: We have revised the Table and Results section to reduce the p-values from 3 digits to 2 digits after the comma. Please refer to the revised manuscript for the changes.

3. I believe the Discussion is still too long and should be shortened.

Response: We have further shortened the Discussion section (from 1,305 to 1,101 words) to improve the readability and clarity. Please refer to the Discussion section of the revised manuscript for the changes.

Thank you for your constructive suggestions. We think that they have greatly enhanced the quality of our manuscript.

19 Feb 2020

Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation

PONE-D-19-28222R2

Dear Dr. Jerng,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Yu Ru Kou, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have completely responded to my comments. No further revision is needed to improve the quality for publication.

Reviewer #2: Dear authors,

I have no further comments about this article. I believe all issues have been addressed.

Best regards

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

21 Feb 2020

PONE-D-19-28222R2

Maximal expiratory pressure is associated with reinstitution of mechanical ventilation after successful unassisted breathing trials in tracheostomized patients with prolonged mechanical ventilation

Dear Dr. Jerng:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Yu Ru Kou

Academic Editor

PLOS ONE