DOI: https://doi.org/10.21203/rs.2.20957/v1
Background
Endotracheal tube cuff underinflation contributes to microaspiration of contaminated oropharyngeal content, source of pneumonia. Overinflation exposes to airway damage. Intermittent manual adjustment of the cuff pressure (P cuff ) may delay the detection of under- or overinflation. Devices for automated continuous adjustment of P cuff are promising but some are inconvenient, expensive or even harmful. This prospective randomized controlled study tested whether the Tracoe Smart Cuff Manager TM reduced the rate of patients undergoing ≥1 episode of underinflation (P cuff <20 cmH 2 O), as compared with routine manual P cuff adjustment. Secondary endpoints included comparisons of the rate of patients with ≥1 overinflation episode (P cuff >30 cmH 2 O), of the incidence of under- and overinflation episodes and of their magnitude. Methods Patients likely to receive invasive mechanical ventilation for >48 hours because of acute brain injury were randomly allocated to receive, during 48 hours, automated P cuff adjustment (combined with manual adjustment) or manual adjustment alone. In each group, P cuff was measured with a dedicated manual manometer, at least every 8 hours.
Results
Sixty patients were analyzed (28 patients with automated and 32 with manual P cuff adjustment) for 506 measurements of P cuff (237 and 269, respectively). As compared with manual adjustment, automated adjustment of P cuff was associated with 1) a lower rate of patients with ≥1 episode of underinflation (63% and 18%, respectively, p<0.001), 2) a lower incidence of underinflation episodes (15% vs. 2%;p<0.001), 3) a lower rate of manual adjustments (77% vs. 56%;p<0.001) and 4) manual adjustments of lower magnitude (5.9±4.5 vs. 3.6±4.7 cmH 2 O;p<0.001). For overinflation, there was no significant between-groups differences (p>0.99).
Conclusions
The adjunction of a continuous P cuff control with the Tracoe Smart Cuff Manager TM to routine manual intermittent adjustment reduced both the incidence of P cuff underinflation episodes and their magnitude without provoking overinflation.
Keeping the endotracheal tube cuff pressure (Pcuff) between 20 and 30 cm H2O [1] is part of nurse routine care of critically ill patients requiring mechanical ventilation [2, 3]. On the one hand, cuff underinflation exposes to the microaspiration of contaminated oropharyngeal content, an important contributor to ventilator-associated pneumonia [2]. On the other hand, cuff overinflation promotes airway (mucosa or cartilage) damage[4]. However, routine intermittent Pcuff monitoring and adjustment via a manual manometer may not avoid underinflation or overinflation episodes between measurements. Hence, less than one fifth of patients have a correct Pcuff during the 8 hours between two manual adjustments [5]. Besides cuff leakage, changes in the airway tone or in patient's position can alter Pcuff between intermittent adjustments [6]. Episodes of under- or overinflation are therefore often overlooked or detected [5]
To achieve a correct Pcuff over 24 hours a day, devices for continuous monitoring of Pcuff have been developed and marketed. Encouraging findings were reported with these devices (blurry, provide a result) [7]. Yet, the amount of data is still too limited to recommend their use [7]. Moreover, these devices are bulky, inconvenient, expensive, and/or need electrical power supply. In addition, some devices may be ineffective and even potentially harmful since not only do they fail in preventing tracheal cuff underinflation as well as they provoke overinflation [8]. For all these reasons, the use of automated Pcuff regulation is still not widespread [9]. A simple elastomeric device, already on the market, may overcome these issues: the Tracoe Smart Cuff Manager™ (Tracoe medical GmbH, Nieder-Olm, Germany). Whether it is effective in keeping Pcuff within the correct range is unknown since no published study assessed this device.
The primary objective of this prospective randomized controlled study was to evaluate whether the adjunction, to routine care, of automatic adjustment of Pcuff with the Tracoe Smart Cuff Manager™ reduced the rate of patients undergoing at least one episode of tracheal cuff underinflation (Pcuff <20 cm H2O), as compared with intermittent manual Pcuff adjustments. As secondary endpoints, we compared, between the two strategies, the rate of patients with ≥ 1 overinflation episode (Pcuff >30 cm H2O), the incidence of under- and overinflation episodes and their magnitude. The incidence of early respiratory infections, ventilator free days, ICU length of stay and mortality was also examined.
This prospective randomized controlled study was performed in two ICUs of Nantes University Hospital (16 and 20 beds). Adult patients admitted to the ICU with acute brain injury were included within the first 48 hours of their ICU stay if they were likely to require mechanical ventilation for at least 48 additional hours through orotracheal tube with a low pressure-high volume cuff.
Patients were not included or were excluded if they had a nasotracheal tube or tracheostomy, if there was a change in the upper airway management within the 48 hours following the inclusion (extubation, change in the tracheal tube or tracheostomy). Pregnancy and safeguarding regimen were also non-inclusion criteria.
Patients were randomly assigned to receive either solely intermittent manual Pcuff adjustments or their combination with Tracoe Smart Cuff Manager™-driven automated adjustments (a group subsequently referred to as automated Pcuff adjustment). Randomization was performed using a 1:1 computer-generated random assignment list.
In both groups, intermittent Pcuff measurements and adjustments were performed with a dedicated manual manometer (Endotest®, Rusch™Teleflex ®, Wayne, USA), every 8 hours [11] and immediately before and after intra-hospital transfer [12].
The use of the Tracoe Smart Cuff Manager™ is detailed in Fig. 1. Briefly, this elastomeric device includes a blue inner balloon. This balloon can either receive air from the tracheal cuff (if the latter is overinflated) or inflate the tracheal cuff (if underinflated). This keeps Pcuff within the 20–30 cm H2O range.
Immediately after randomization, the Tracoe Smart Cuff Manager™ was connected to the Pcuff control port of the endotracheal tube via a three way stopcock in order to avoid repeated disconnections of the device when manometer measurements of Pcuff are performed. Indeed, for a more direct measurement of Pcuff in this study, we chose to measure Pcuff upstream of the Tracoe Smart Cuff Manager™ rather than using the dedicated air inlet of the device.
The Tracoe Smart Cuff Manager™ was kept connected 24 hours a day, even during patient transfer (to the imaging facility for instance), for at least 48 hours (the observation period). Its maintenance after the 48th hour was left to the discretion of caregivers.
The primary objective was to evaluate whether the Tracoe Smart Cuff Manager™ reduced the rate of patients undergoing at least one episode of tracheal cuff underinflation (Pcuff <20 cm H2O), as compared with intermittent Pcuff adjustments only.
As secondary endpoints, comparisons between the two groups were as follow: the rate of patients undergoing at least one episode of tracheal cuff overinflation (Pcuff >30 cm H2O), the incidence of under- and overinflation episodes, the rate of Pcuff adjustments and their magnitude. Ventilator free days, ICU length of stay and mortality and incidence of early respiratory infections were also compared. The latter was defined as the prescription, by the attending physician, of antimicrobial therapy for pneumonia or tracheobronchitis, over at least a 7-day period beginning 48 hours after the randomization. Indeed, an infection occurring within the first 48 hours is likely to be related to aspiration of upper airway content prior to the inclusion rather than to failure of the allocated strategy for Pcuff control. Patients with ongoing antimicrobial therapy for respiratory infection at the beginning of this 7-day period were excluded from this analysis. After careful review of the medical charts, cases of early respiratory infections were adjudicated by a physician blind to the randomization.
A previous study reported that, with no device for automated Pcuff control, 55% of patients experienced at least one episode of underinflation within a 48-hour observation period [7]. To detect a Tracoe Smart Cuff Manager™-induced decrease, from 55 to 20%, in the proportion of patients experiencing at least one episode of underinflation, 56 patients were required (two groups of 28), with two-sided tests, an alpha risk of 0.05 and a power of 80%. A 5% margin yielded a total of 60 patients.
Intention-to-treat analyses were performed. Normally and non-normally distributed quantitative parameters were expressed as mean ± SD or median [interquartile range], respectively. Their comparisons relied on Student’s t-test or Mann-Whitney U test, respectively. Qualitative parameters were expressed in count (%) and compared with Fisher’s exact test. The Epi info™ version 7.2.2.6 (CDC, Atlanta, GA, USA, 2011) and the RStudio version 1.2.1335 (R Foundation for Statistical Computing, Vienna, Austria) softwares were used. A p < 0.05 was considered significant.
From March 2018 to July 2019, 65 patients were included. Informed consent could not be obtained in 5 patients. Therefore, 60 patients were analyzed: 28 were allocated to the adjunction of automated Pcuff adjustment, 32 to routine manual adjustments only (Fig. 2).
Overall, 506 measurements of Pcuff were performed: 237 and 269 in patients with automated and with manual adjustments, respectively (mean of 8.5 ± 3.5 and 8.4 ± 2.4 measurements per patient).
Most patients’ characteristics were similar in the 2 groups except for a few of them, including some airway pressures which were higher in patients with manual Pcuff adjustment (Tables 1 and 2). Prone position was not performed in any of the included patients during the observation period.
Total (n = 60) | Automated Pcuff adjustment (n = 28) | Manual Pcuff adjustment (n = 32) | p | |
---|---|---|---|---|
Age (years) | 55 [42–64] | 57 [44–66] | 52 [42–60] | 0.24 |
Female gender (n [%]) | 25 (42%) | 10 (36%) | 15 (47%) | 0.44 |
Height (cm) | 170 [162–175] | 168 [162–180] | 170 [163–174] | 0.79 |
Weight (kg) | 69 [59–80] | 70 [61–77] | 68 [58–81] | 0.86 |
BMI | 23.4 [21.6–26.7] | 22.9 [19.8–24.8] | 24.0 [22.5–28.0] | 0.14 |
SAPS II | 47 [34–60] | 49 [39–43] | 43 [32–58] | 0.53 |
SOFA score | ||||
admission | 8 [5–9] | 8 [6–9] | 7 [5–10] | 0.75 |
inclusion | 9 [7–10] | 9 [6–9] | 9 [7–11] | 0.15 |
Reason for ICU admission (n [%]) | ||||
SAH | 26 (43%) | 12 (43%) | 14 (44%) | > 0.99 |
ICH | 10 (17%) | 5 (18%) | 5 (16%) | > 0.99 |
Post-operative admission after elective surgery | 9 (15%) | 4 (14%) | 5 (16%) | > 0.99 |
TBI | 6 (10%) | 2 (7%) | 4 (13%) | 0.67 |
Stroke | 5 (8%) | 2 (7%) | 3 (9%) | > 0.99 |
Other | 4 (7%) | 3 (11%) | 1 (3%) | 0.33 |
Comorbidities (n [%]) | ||||
Hypertension | 17 (28%) | 5 (18%) | 12 (38%) | 0.15 |
Immunocompromised | 1 (2%) | 1 (4%) | 0 | 0.47 |
Diabetes mellitus | 3 (5%) | 2 (7%) | 1 (3%) | 0.59 |
Antimicrobial therapy started within 48 h after tracheal intubation (n [%]) | 15 (25%) | 5 (18%) | 10 (31%) | 0.37 |
Endotracheal tube model (n [%]) Profile™ Soft-Seal® Cuff* Other | 37 (62%) 22 (38%) | 19 (68%) 9 (32%) | 18 (56%) 14 (44%) | 0.43 |
Intra-hospital transport during the 48 hour observation period (n [%]) | ||||
1 | 44 (77%) | 18 (64%) | 26 (81%) | 0.02 |
2 | 16 (27%) | 6 (21%) | 10 (31%) | 0.55 |
3 | 6 (10%) | 4 (14%) | 2 (6%) | 0.40 |
4 | 2 (3%) | 1 (4%) | 1 (3%) | 1 |
MV duration (days) | 14 [7–23] | 17 [8–28] | 13 [7–20] | 0.45 |
Ventilator Free Days | 4 [0–8] | 5 [0–7] | 3 [0–10] | 0.96 |
Length of stay (days) | 20 [11–34] | 25 [11–35] | 17 [11–28] | 0.39 |
ICU Mortality (n [%]) | 22 (37%) | 10 (36%) | 12 (38%) | 0.99 |
Legend: BMI: body mass index; SAPS II: simplified acute physiology score; SOFA: sequential organ failure assessment; ICU: intensive care unit; SAH: subarachnoid hemorrhage; ICH: intra-cerebral hematoma; TBI: traumatic brain injury. *Profile™ Soft-Seal® Cuff, Portex, Smith Medical, MN, USA. |
Automated adjustment of Pcuff | Manual Adjustment of Pcuff | p | |
---|---|---|---|
Assisted Controlled Ventilation Peak Pressure (cm H2O) Plateau Pressure (cm H2O) | 179/246 (73%) 27.4 ± 5.9 13.6 ± 5.6 | 225/277 (81%) 28.3 ± 7.8 15.4 ± 5.8 | 0.02 0.18 0.002 |
Pressure Support Ventilation Pressure support (cm H2O) | 67/246 (27%) 12.0 ± 4.1 | 52/277 (19%) 13.4 ± 3.3 | 0.02 0.052 |
Enteral Feeding | 150/246 (62%) | 200/277 (72%) | 0.01 |
Vomiting since the previous Pcuff determination | 5/246 (2%) | 8/277 (3%) | 0.59 |
Head of bed elevation, degrees | 29 ± 3 (n = 246) | 29 ± 5 (n = 277) | 0.97 |
PEEP | 5.5 ± 1.6 | 5.6 ± 1.4 | 0.46 |
RASS | -5 [–5, –4] | -5 [–5, –5] | 0.08 |
Infusion of, n (%) | |||
Propofol | 21/28 (75%) | 24/32 (75%) | > 0.99 |
Midazolam | 16/28 (57%) | 24/32 (75%) | 0.18 |
Sufentanil | 23/28 (82%) | 28/32 (87%) | 0.72 |
Sodium thiopental | 4/28 (14%) | 6/32 (19%) | 0.74 |
Neuromuscular blocking agent | 6 (21%) | 12 (38%) | 0.26 |
Legend: Pcuff: tracheal cuff pressure; PEEP: positive end-expiratory pressure; RASS: Richmond agitation sedation scale. |
The rate of patients with at least 1 underinflation episode within the 48 hour observation period was significantly lower with automated Pcuff adjustment than with manual Pcuff adjustment only 5/28 (17.9% [95%CI 7.4–36.0]) vs. 20/32 (62.5% [95%CI 45.2–77.1]); p = 0.006)(Fig. 3).
The rate of patients with at least 1 overinflation episode was similar (p > 0.99) in the 2 groups: 3/28 (10.7% [95%CI 2.9–28.0]) vs. 4/32 (12.5% [95%CI 4.4–28.7]) with automated Pcuff adjustment and with manual Pcuff adjustment only, respectively.
Incidence of under- and overinflation episodes. During the 48-hour observation period, mean Pcuff was higher with the automated Pcuff adjustment (25.3 ± 5.0 vs. 22.9 ± 4.3 cm H2O; p < 0.001). The incidence of underinflation episodes was significantly lower in the automated Pcuff adjustment (237 measurements) than in the manual adjustment group (269 measurements): 2.1% [95%CI 0.3–3.9] vs. 14.9% [95%CI 10.3–19.4]; p < 0.001 (Fig. 4). The incidence of overinflation episodes was similar in the two groups: 1.7% [95%CI 0.5–4.4] vs. 1.9% [95%CI 0.7–4.4], p > 0.99 (Fig. 4).
The rate of P cuff adjustments performed -after Pcuff measurement was lower with the automated Pcuff adjustment: 55.7% [95%CI 49.4–61.8] vs. 76.6% [95%CI 71.1–81.3]; p < 0.001. In addition, these adjustments were of lower magnitude in patients with automated Pcuff adjustment (3.6 ± 4.7 vs. 5.9 ± 4.5 cm H2O; p < 0.001).
Other outcomes. In each group, 8 early respiratory infections occurred in 8 patients yielding an incidence of 29% and 25% in the automated and the manual Pcuff adjustment group, respectively (p = 0.78). No significant difference was found between the two groups with regard to the following outcomes at the ICU discharge: duration of mechanical ventilation, ventilator free days, length of ICU stay, and ICU mortality (Table 1).
The main finding of this randomized controlled study ̶ coordinated by nurses of two surgical ICUs of the University Hospital of Nantes ̶ is that implementing an automated adjustment of Pcuff by the Tracoe Smart Cuff Manager™ in combination with intermittent manual adjustments was more effective to regulate Pcuff than a strategy with manual adjustments alone. Indeed, automated adjustment of Pcuff markedly decreased the rate of patients experiencing at least one underinflation episode and the incidence of detected episodes of underinflation. Importantly, the automated Pcuff adjustment strategy was not associated with a higher incidence of detected episodes of tracheal cuff overinflation.
Interestingly, despite the randomization, airway pressures were or tended to be higher in patients allocated to undergo manual Pcuff adjustments solely (Table 2). Since airway pressure is transmitted to Pcuff [13], this imbalance could have reduced the incidence of underinflation episodes among patients with manual Pcuff adjustments solely. In other words, assuming strictly identical airway pressure between the two groups, the protective effect of the Tracoe Smart Cuff Manager™ against underinflation episodes could have been even more prominent.
To avoid both microaspirations and tracheal damage, achieving a 24-hour balance between under- and overinflation of the tracheal cuff is desirable. The issue is not new. Indeed, from the 1970s, some modified endotracheal tubes were released onto the market for this purpose. One of these tubes housed a valve connected to a larger balloon than the standard pilot balloon. This regulating valve aimed at maintaining Pcuff close to 30 cm H2O [14]. This so-called Lanz system is the forerunner of the Tracoe Smart Cuff Manager™. One advantage of the latter relies on its ability to be connected to the standard port of most tracheal tube models (contrary to the Lanz system which is natively-embedded into one model of tracheal tube). To the best of our knowledge, no published study evaluated the Tracoe Smart Cuff Manager™. Overall, devices automatically regulating Pcuff were surprisingly poorly assessed. Indeed, besides some very preliminary reports of “home-made” devices [15], a dozen studies evaluated whether an automated device connected to the endotracheal tube effectively maintains Pcuff within an acceptable range (7,8,15,18–24). All these previous studies were single-center studies. They were uncontrolled [22], not randomized [17], of limited size (18 patients or less)[8, 16, 20, 22], experimental [13, 21] and/or tested non-commercialized devices [19, 22]. Although several very different devices have been tested, most of these studies have concluded that maintenance of Pcuff within the acceptable range was better achieved with an automated rather than an intermittent manual control of Pcuff, especially via the reduction of underinflation episodes. In their hallmark study, Nseir et al. went even further by demonstrating that implementing an automated Pcuff control reduced both the tracheal level of pepsin (suggesting a reduction of gastric content microaspiration) and the rate of ventilator-associated pneumonia [7]. Importantly, the reduction in the rate of underinflation episodes must not be at the expense of a potentially harmful increase in the rate of overinflation episodes. Unfortunately, one device [8] and possibly another one [19] were associated with a higher rate of Pcuff episodes above 30 cmH2O than routine care. Other drawbacks applying to some devices are worth mentioning: they are bulky and/or inconvenient (which encourages their disconnection before patient transport, a procedure at-risk of aspiration of oropharyngeal content), relatively expensive, need electric [16] or gas [23] supply. This may explain the lack of wide clinical use of automated Pcuff control. The Tracoe Smart Cuff Manager™ may overcome all these issues.
First, inherently to the nature of the tested intervention, caregivers could not be blinded to the randomization arm.
Second, the study was focused on the specific population of brain injured patients, and we mostly used one model of endotracheal tube. However, this population is of interest in the field of prevention of respiratory infections since it is prone to develop such infections because, among other reasons, prolonged median duration of mechanical ventilation [24, 25]. We therefore believe that focusing on this specific population should be seen as strength rather than limitation of this study. Therefore, relative caution should be exercised before extrapolating our findings to other settings or populations or even to other airway access devices. Furthermore, this population is also particularly prone to be deeply sedated (Table 1). Since the absence of sedation is independently associated with increased risk for cuff underinflation [5], one can hypothesize that the adjunction of the Tracoe Smart Cuff Manager™ to routine care in less sedated patients would have been associated with an even more marked reduction in the incidence of underinflation episodes than that herein reported. This hypothesis remains to be tested.
Third, Pcuff measurements were not collected continuously. Therefore, undetected episodes of under- or overinflation may have been overlooked. This would have impacted our conclusions if these undetected episodes of incorrect Pcuff were significantly more frequent in the automated adjustment group than in the manual adjustment group, i.e., an imbalance totally opposite to that observed for detected events. This hypothesis is therefore highly unlikely.
Fourth, in patients with automated Pcuff adjustment but not in patients from routine care group, we used a three-way stopcock. Of note, direct connection of the manometer to the endotracheal tube may be source of air leakage [27]. The use of the three-way stopcock may have limited the air leakage by connecting the manometer before opening the dedicated outlet. Therefore, some patients may have benefitted more from the use of the three-way stopcock than from the use of the Tracoe Smart Cuff Manager™. We chose to add the three-way stopcock to the Tracoe Smart Cuff Manager™ in order to avoid its disconnection (and therefore its necessary re-inflation) each time a manometer is connected, i.e., several times a day. This allowed direct Pcuff measurements, i.e., upstream of the Tracoe Smart Cuff Manager™ rather than using the dedicated inlet of the device (Fig. 1). Of note, the manufacturer proposes the direct connection of the Tracoe Smart Cuff Manager™ to the tracheal tube with no three-way stopcock, intermittent manometer Pcuff measurements being then performed via the dedicated inlet. Interestingly, with this latter approach, manometer connection-induced possible air leakage would slightly deflate the inner balloon of the device but, importantly, not the tracheal cuff. Therefore, when using the Tracoe Smart Cuff Manager™ with a three-way stopcock or not, manometer connection-induced deflation of the tracheal cuff is theoretically reduced. We believe this is a strength of the device. In the present study, using a three-way stopcock in both groups would have been helpful to delineate the specific contribution of the Tracoe Smart Cuff Manager™ and the three-way stopcock to our findings. However, adding a three-way stopcock to the endotracheal tube when no device for automated adjustment is connected is not recommended [24, 25] and is therefore neither wide current practice nor part of our routine care.
Fifth, in this study, manual adjustments were not guided by strict specific rules except correcting a Pcuff value outside the recommended range of 20–30 cmH2O. Indeed, the decision of manual adjustment and its magnitude tightly depends on previous measurements and adjustments in a given patient. Hence, several manual adjustments have consisted in modifying a Pcuff value already within this recommended range (from 21 to 27 cmH2O for instance) in a possibly unduly manner. This may explain the somewhat high incidence of manual adjustments in both groups (56% in the Tracoe Smart Cuff Manager™ group vs. 77% in the routine care group, p < 0.001). Since the Tracoe Smart Cuff Manager™ is able to inflate or deflate the tracheal cuff, these manual adjustments could have been unnecessary in patients equipped with this device. Indeed, less than 4% of Pcuff measurements lied outside the recommended range in this group.
Last, our study was not designed to assess whether the Tracoe Smart Cuff Manager™ reduces the incidence of ventilator-associated respiratory infections and therefore the ventilator free days, ICU length of stay or even mortality. Indeed, the Tracoe Smart Cuff Manager™ was not kept in place throughout the duration of invasive ventilation but only 48 hours in most patients, i.e., just the duration of the study observation period for its primary endpoint. The present study demonstrates that the Tracoe Smart Cuff Manager™ reduced the incidence of cuff underinflation. Whether this encouraging finding would translate into a reduction in ventilator-associated respiratory infections is not straightforward and deserves a dedicated study [19]. Indeed, a too rapid Pcuff correction by the automated device may interfere with the self-sealing mechanism of the cuff and may reduce its sealing characteristics, therefore exposing to microaspirations [28]. Interestingly, the Tracoe Smart Cuff Manager™ houses a valve aiming at slowing down the reaction time before Pcuff correction but its effectiveness has to be specifically assessed. Moreover, prevention of ventilator-associated respiratory infections is multifaceted and demonstrating a reduction in their incidence by acting on only one facet (reduction of underinflation episodes) may require a large study size. An ongoing multicenter study testing another pneumatic device may provide further insight on this relevant topic [29].
By showing that an approach including the Tracoe Smart Cuff Manager™ allowed a better control of Pcuff into the recommended range of 20–30 cm H2O as compared with routine care, this study paved the way for a large study specifically addressing whether this device (as part of a multifaceted preventive strategy) is effective in the prevention of ventilator-associated pneumonia.
BMI: body mass index
CDC: control disease center
ETT: Endotracheal tube
ICH: intra-cerebral hematoma
ICU: intensive care unit
MV: mechanical ventilation
Pcuff: endotracheal tube cuff pressure
SAPS II: simplified acute physiology score
SAH: subarachnoid hemorrhage
SD: standard deviation
SOFA: sequential organ failure assessment
TBI: traumatic brain injury
Ethics approval and consent to participate
A national ethics committee (Comité de Protection des Personnes, CPP 2017/62) approved the study design. Written and oral inform consent was obtained from each patient or to his next of kin if the patient was not able to consent.
Consent for publication
Not applicable
Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests
Jérôme Dauvergne has no conflict of interest in connection with the work submitted. In addition, JD received, during the past 5 years, travel fees from HALYARD (once in 2016).
Anne-Laure Geffray has no conflict of interest in connection with the work submitted.
Bertrand Rozec has no conflict of interest in connection with the work submitted. In addition, BR received, during the past 5 years, lecture fees from Fisher&Paykel, Baxter, LFB, Aspen, research grants from Baxter and consulting fees from LFB, Astra Zeneca.
Karim Asehnoune received fees from Fisher Paykel, LFB, Fresenius, Edwards (outside the submitted work).
Karim Lakhal has no conflict of interest in connection with the work submitted. In addition, KL received, during the past 3 years, lecture fees from MEDTRONIC (once, in 2017), congress registration fees from SANOFI AVENTIS (once in 2018), travel fees from MSD France (once, in 2017), NOVEX PHARMA (once, in 2016), GILEAD SCIENCES (twice, 2016 and 2017), PFIZER (once, in 2019).
Funding
This study was fully funded by TRACOE Medical GmbH, prior to its conduction. The Tracoe Smart Cuff ManagerTM was also provided by TRACOE Medical GmbH. This company has not interfered in the design, conduct and data analysis of the study. None of the investigators has any direct conflict of interest with the company. The financial relationship only involved the institutional research department.
Authors' contributions
Conception and design: JED, KL.
Collection of data: JED, ALG.
Statistical analysis: JED, KL.
Drafting and revision of the manuscript: JED, KL, BR, KA
Acknowledgements
We are grateful to Mrs. Laurence Pacaud and Mrs. Sylvie Le Guillou, research nurses, and Mrs. Emmanuelle Cartron, nurse research coordinator, for their important contribution to this work. We are indebted to the Critical Care nurses and all the medical staff of our department of Anaesthesia and Critical Care Medicine.