The PRISMA guidelines were used to report this review (13). The present systematic review was registered in the International prospective register of systematic reviews (PROSPERO) under registration number CRD42020171445.
We searched for eligible studies until 11 March 2020 in the following databases: Web of Science, MEDLINE, Cochrane Central Register of Controlled Trials (See Appendix 1 for the search strategy). We searched on the OpenGrey database and also reviewed references lists of retrieved studies.
We included all types of interventional trials. Only publications in English, French or Spanish were selected. Studies had to include participants whom were in Intensive Care Units (ICU), ventilated with an endotracheal tube, and whom were undergoing respiratory weaning or extubation. Only adults were selected.
We included trials evaluating hypnotherapy during mechanical ventilation weaning or extubating process. If the study included at least two groups, one of them would not have received hypnotherapy.
The primary outcome was self-reported dyspnea, without any restriction on the method of measurement.
Secondary outcomes were self-reported anxiety, stress and pain without any restriction on the method of measurement; and extubation success, defined as the proportion of patients free from invasive ventilatory support 48 hours after extubation.
Two authors per trial (JW, AB) assessed the titles and abstracts from reference lists for possible inclusion. These review authors read the full text of articles to make the final decision for inclusion. Disagreements were resolved by discussion (JW, AB). When inclusion was still questionable even with the full text, another reviewer was contacted (CR) to determine whether the article met inclusion criteria. Duplicate publications of the same study were noted, but each trial was counted only once.
Data Extraction and Quality Assessments
Two review authors per trial (JW, AB) extracted the characteristics for each trial. The results were compared, disagreements resolved by discussion (JW, AB, CR). We extracted information on the trial design, number included/analysed, participants’ characteristics, country of study, method of assessment, inclusion/exclusion criteria, outcomes, follow-up times, adverse events, dropouts and withdrawals, intervention method and findings as they were reported in the paper.
Two review authors per trial (JW, AB) independently assessed risk of bias. For case report or case-series studies, the evaluation was based on the tool developed by Murad et al. for evaluating the methodological quality of this specific study design (14). For case-control or cohort studies, the Newcastle Ottawa Scale was used (15). For randomized and non-randomized controlled clinical trials (respectively RCT and CCT), the evaluation was based on the PEDro scale (16). A study was considered as “high quality” if all the items were present, except those who evaluated the patient and therapist’s blinding (items 5 and 6). If any other item was not found in the publication, the study was considered as “low quality”.
We used the tool developed by van Tulder et al. for evaluating the level of evidence (17) (Table 1).
Table 1. Level of Evidence
a. Strong – consistent findings amongst multiple high quality RCTs.
b. Moderate – consistent findings amongst multiple low quality RCTs and/or CCTs and/or one high quality RCT
c. Limited – one low quality RCT and/or CCT
d. Conflicting – inconsistent findings amongst multiple trials (RCT and/or CCTs)
e. No evidence from trials – no RCTs or CCTs
Findings were considered as consistent if all studies had the same conclusions.
RCT - Randomized Controlled Trials
CCT - Nonrandomized Controlled Trials
To minimize the effect of publication bias, we searched the grey literature. To reduce the reporting bias, we consulted CT.gov looking for described protocols of the included studies.