Closed–Loop ventilation using sidestream versus mainstream capnography for automated adjustments of minute ventilation—A randomized clinical trial in cardiac surgery patients

Background INTELLiVENT–Adaptive Support Ventilation (ASV) is a closed–loop ventilation mode that uses capnography to adjust tidal volume (VT) and respiratory rate according to a user–set end–tidal CO2 (etCO2) target range. We compared sidestream versus mainstream capnography with this ventilation mode with respect to the quality of breathing in patients after cardiac surgery. Methods Single–center, single–blinded, non–inferiority, randomized clinical trial in adult patients scheduled for elective cardiac surgery that were expected to receive at least two hours of postoperative ventilation in the ICU. Patients were randomized 1:1 to closed–loop ventilation with sidestream or mainstream capnography. Each breath was classified into a zone based on the measured VT, maximum airway pressure, etCO2 and pulse oximetry. The primary outcome was the proportion of breaths spent in a predefined ‘optimal’ zone of ventilation during the first three hours of postoperative ventilation, with a non–inferiority margin for the difference in the proportions set at –20%. Secondary endpoints included the proportion of breaths in predefined ‘acceptable’ and ‘critical’ zones of ventilation, and the proportion of breaths with hypoxemia. Results Of 80 randomized subjects, 78 were included in the intention–to–treat analysis. We could not confirm the non–inferiority of closed–loop ventilation using sidestream with respect to the proportion of breaths in the ‘optimal’ zone (mean ratio 0.87 [0.77 to ∞]; P = 0.116 for non–inferiority). The proportion of breaths with hypoxemia was higher in the sidestream capnography group versus the mainstream capnography group. Conclusions We could not confirm that INTELLiVENT–ASV using sidestream capnography is non–inferior to INTELLiVENT–ASV using mainstream capnography with respect to the quality of breathing in subjects receiving postoperative ventilation after cardiac surgery. Trial registration NCT04599491 (clinicaltrials.gov).


TABLE OF CONTENTS
INTELLiVENT-ASV is a safe mode of ventilation, certainly also in patients who receive postoperative ventilation, and therefore this automated mode is preferably used in all cardiac surgery patients in the participating ICU. Of note, as all patients are under sedation during the phase postoperative ventilation is applied, burden for the patient is negligible. Whether INTELLiVENT-ASV with sidestream capnography performs as well as

INTELLiVENT-adaptive support ventilation (ASV) is a fully automated closed
INTELLiVENT-ASV with mainstream capnography with respect to quality of breathing is unknown. This study, therefore, will compare INTELLiVENT-ASV with sidestream capnography to INTELLiVENT-ASV with mainstream capnography.

OBJECTIVE AND HYPOTHESIS
The objective of this randomized noninferiority clinical trial is to compare INTELLiVENT-ASV with sidestream capnography with INTELLiVENT-ASV with mainstream capnography during postoperative ventilation in patients after elective cardiac surgery.
We hypothesize INTELLiVENT-ASV with sidestream capnography is noninferior to INTELLiVENT-ASV with mainstream capnography with respect to the percentage of breaths patient spend within the 'optimal' zone of ventilation.

STUDY DESIGN
'INTELLiSTREAM' is an investigator-initiated, investigator-sponsored, single-center, randomized, noninferiority clinical trial in patients who receive postoperative ventilation in the intensive care unit after elective cardiac surgery.

STUDY POPULATIONPopulation (base)
The study population consists of consecutive elective cardiac surgery patients who will receive postoperative ventilation in the intensive care unit (ICU).

Sample size calculation
The sample size is computed based on the hypothesis that ventilation with INTELLiVENT-ASV with sidestream capnography is similar to INTELLiVENT-ASV with mainstream capnography with respect to the percentage of breaths within the 'optimal' zone of ventilation (see below). Considering an expected percentage of breaths in 'optimal' zone with mainstream capnography of 69 ± 23% (results from a previous study, the report of which is currently under review at a peer-reviewed journal; ClinicalTrials.gov Identifier: NCT03180203), a power of 80%, a one-sided alpha level of 0.05 and a non-inferiority margin of 20% (corresponding to 14% less breaths in the 'optimal' zone in the sidestream group compared to the mainstream group), 72 patients need to be included.
Patients who are extubated before 90 minutes of postoperative ventilation will be considered drop-outs and will not contribute towards the sample size. This cut-off is chosen to have sufficient time to observe the evolution of postoperative ventilation requirements, as these will significantly change over the first 90 minutes of postoperative ventilation in the ICU. This also increases comparability of patients and reduces the likelihood of clinical and statistical heterogeneity, as patients rapidly awaking from anesthesia after surgery will likely have different respiratory needs. We will continue recruiting patients until both study arms have at least 36 patients with at least 90 minutes of postoperative ventilation per randomization group.

Investigational treatment
Upon arrival at the ICU, patients will be randomized to receive either ventilation with INTELLiVENT-ASV with mainstream capnography, or INTELLiVENT-ASV with sidestream capnography (Figure). Attending physicians will decide to extubate the patient based on general extubation criteria: the patient is responsive and cooperative, has a core temperature > 36.0°C, urine output > 0.5 mL/kg/h, chest tube drainage < 100 mL in the last hour, no uncontrolled arrhythmia or hemodynamic instability, and a respiratory rate of > 10 per minute without machine-controlled breaths for at least 30 minutes. T-piece weaning is not applied; patients are extubated once they reach the aforementioned extubation criteria.

Main study parameter/endpoint
Three widely accepted and previously used predefined ventilation 'zones' are used, i.e., 'optimal', 'acceptable' and 'critical' (Table 1) (Table). In case of missing data, if any of the present parameter is within the 'critical' zone, the zone is defined as 'critical'. If data is missing, but none of the parameters are in the 'critical' zone, the zone is defined as 'missing'.
When the capnography signal is lost, INTELLiVENT-ASV 'freezes' minute ventilation, and continues ventilation based on the last registered capnography signal.
Ventilation data during signal loss will be analyzed as part of the primary endpoint. and 30-day mortality, capnography equipment signal loss requiring a correction by ICU nurses, and incidence of hypoxemia, defined as percentage of breaths with SpO2 < 85% but only when SpO2 had a quality index > 50%.

Randomization, blinding and treatment allocation
Included patients will be randomly allocated in a 1:1 ratio to start ventilation with INTELLiVENT-ASV using either a mainstream or a sidestream capnography. The allocation sequence will be computer-generated using permuted blocks of random block sizes. Randomization will then be performed by local investigators patient-by patient using a dedicated, password protected, SSL-encrypted website. Patients will remain unaware of which sensor technique for capnography was used to control the ventilator. Due to the nature of the treatment, blinding of the care providers is not possible. However, the outcome assessor analyzing the results will remain blinded to treatment allocation.

Study procedures and data to be collected
Patients will receive postoperative ventilation using INTELLiVENT-ASV as part of standard care after cardiac surgery. Data will be collected while the patient is receiving postoperative ventilation in the intensive care unit (ICU). Collection of data will continue until tracheal extubation, or until 6 hours after start of ventilation in the ICU, whatever comes first. Patients randomized to sidestream capnography will be placed on mainstream capnography after 6 hours of ventilation in the ICU, if not yet extubated.
In case the capnography signal is lost, e.g. due to kinking or obstruction of the tube with sidestream capnography, or sensor disconnection or condense in the sensor-eye, INTELLiVENT-ASV 'freezes' the minute ventilation, and continues ventilation based on the last registered capnography signal, guaranteeing minute volume and patient safety. At the same moment, a visual and auditory alarm is activated, after which the nurse will solve the problem, and INTELLiVENT-ASV restarts using the capnography signal to adjust the minute volume.

Withdrawal of individual subjects
The investigator can decide to withdraw a subject from the study for urgent medical reasons. Subjects can decide to retract informed consent before start of anesthesia for surgery, and also to retract consent for use of data after tracheal extubation. In the randomization log these cases will be recorded without patient-specific data.

Follow-up of subjects withdrawn from treatment
Patients will be subjected to a 30-day follow-up after postoperative admission to the ICU. Patients will be contacted once, 30 days after inclusion, to assess for complications, readmissions, and the final date of hospital discharge.

Temporary halt for reasons of subject safety
In accordance to section 10, subsection 4, of the WMO, the sponsor will suspend the study if there is sufficient ground that continuation of the study will jeopardize subject health or safety. The sponsor will notify the accredited METC without undue delay of a temporary halt including the reason for such an action. The study will be suspended pending a further positive decision by the accredited METC. The investigator will take care that all subjects are kept informed.

AEs, SAEs
This study compares fully automated ventilation with two types of sensors that are used within their indication, and no related serious adverse events (SAEs) are expected.
Related SAEs or possible related SAEs will be reported to the METC within 7 days of first knowledge of SAE. Unrelated SAEs will be reported to the METC via line listing once per year. All SAEs and related AEs will be recorded in the eCRF.

Follow-up of adverse events
Follow-up of AEs and SAEs is 30 days after the intervention.

STATISTICAL ANALYSIS
1 8.1 Primary study parameter(s) 2 The primary outcome, the percentage of breaths spent in the predefined 'optimal' zone, 3 will be analyzed for non-inferiority, considering the margin of non-inferiority of 20%. 4 Data will be collected breath-by-breath, but the percentage of breaths in the 'optimal' 5 zone will be summarized per patient according to the formula below (Eq.1). 6 Data will be presented as median (quartile 25% -quartile 75%) and mean ± 7 standard deviation, compared as a mean ratio (as described in the Eq. 2), tested for 8 non-inferiority considering a margin of 20%, and presented as a one-sided 95% 9 confidence interval. Thus, non-inferiority will be established if lower boundary of the 10 one-sided 95% confidence interval was higher than 0.80 (20% decrease in percentage 11 of breaths in optimal zone). A one-sided p value for non-inferiority will be calculated. 12 Results will be presented in a table of outcomes and also in a forest plot. Statistical 13 uncertainty will be expressed by 95% confidence intervals. interaction of time and randomization group as fixed effects. Two p values will be 1 reported: 1) p value for the group difference, reflecting the overall test for difference 2 between groups across the three hours; and 2) p values for the group x time interaction, 3 evaluating if change over time differed by group. In addition, since it is expected that 4 the baseline values will be similar between the groups, these will be exposed in the 5 graphs but excluded from the models. 6 For outcomes assessing proportions of breaths and incidence of hypoxemia, 7 the denominator will be the total number of breaths. Secondary binary outcomes, 8 including the proportions of breaths and incidence of hypoxemia, will be assessed with Patients will be recruited at pre-assessment by the anesthesiologist. Subjects will be 7 informed verbally by local researchers, and by a patient information letter. The patient 8 will be given sufficient time to consider their decision and to discuss the decision with 9 their relatives or the independent expert. Written informed consent will be obtained 10 prior to surgery, with the possibility to ask any remaining questions about participation.   18 The sponsor/investigator has an insurance for liability, which is in accordance with 19 article 7 subsection 6 of the WMO. All research subjects are insured to cover damage 20 through injury or death caused by the study. The insurance applies to the damage that 21 becomes apparent during the study or within 4 years after end of the study.

Handling and storage of data and documents
2 All patients will be assigned a patient identification code. The codebook will be stored 3 digitally and in paper. The paper version will be stored in a locked cabinet and the 4 digital form will be encrypted with a double password. All data will be stored for the 5 length of the study and for 15 years afterwards. All handling of personal data will 6 comply with the GDPR. Amendments are changes made to the research protocol after a favorable opinion by 15 the accredited METC has been given. All amendments will be notified to the METC 16 that gave a favorable opinion. All substantial amendments will be notified to the METC 17 and to the competent authority. Non-substantial amendments will not be notified to the 18 accredited METC and the competent authority, but will be recorded and filed by the 19 sponsor.  3 The investigator will notify the accredited METC of the end of the study within a period 4 of 8 weeks. In case the study is ended prematurely, the investigator will notify the 5 accredited METC within 15 days, including the reasons for the premature termination. 6 Within one year after the end of the study, the investigator/sponsor will submit a final 7 study report with the results of the study, including any publications/abstracts of the 8 study, to the accredited METC. 9 10.6 Public disclosure and publication policy 10 The results of the study will find their way into (inter-) national scientific journals and 11 guidelines. We will submit the analyses to scientific journals in the field of intensive 12 care medicine as well as anesthesiology, since both ICU physicians and 13 anesthesiologists apply ventilation in the ICU setting. 14 11. STRUCTURED RISK ANALYSIS 1 11.1 Potential issues of concern 2 Chapter 12.1 is skipped because both types of etCO2 sensors are used within their 3 indication, and are alternately used in the participating unit. 4

5
There is no additional risk to mechanical ventilation with INTELLiVENT-ASV with 6 sidestream or mainstream capnography, other than the risks associated with 7 ventilation per se. Patients included in this study will receive mechanical ventilation as 8 part of standard care. Therefore, this intervention does not lead to an increase of any 9 risk for the patient. Extubation will not be delayed because of this study. Furthermore, 10 patients will be sedated during the study, without exception.