Types of studies

We will include reports of randomised controlled trials (RCTs) conducted in adults, including studies reported in full text, those published as an abstract only, and unpublished data. No language restrictions will be applied. Cross‐over randomised clinical trials will be included for analysis of short‐term outcomes only. We will include both blinded (sham‐controlled) and non‐blinded studies. We will not include cluster‐randomised, quasi‐randomised, and non‐randomised trials.

Types of participants

We will include adults with a diagnosis of obesity hypoventilation syndrome (OHS). Most contemporaneous studies use the diagnostic criteria set forth by the International Classification of Sleep Disorders (ICSD). However, OHS only became recognised as a distinct condition in the ICSD in its third revision in 2014 (American Academy of Sleep Medicine 2014). Prior to this, OHS had been considered a subtype of 'Central Alveolar Hypoventilation Syndrome' and variations in PaCO₂ criteria thresholds are seen in earlier literature by expert authors. Our operational definition for OHS will be adults having all of the following criteria:

1. daytime PaCO₂ 45 mmHg or more;

2. body mass index 30 kg/m² or more;

3. absence of medical disorders or medications that may cause hypoventilation.

Regarding the last criterion, we will include studies only where a specific comment has been made regarding efforts to exclude other causes for hypoventilation, including chronic lung disease, neuromuscular conditions, chest wall abnormalities and sedating medications. We will exclude participants receiving non‐invasive PAP therapy for acute respiratory failure, or who are in hospital for an indication other than elective acclimatisation to PAP therapy (or both). Adults with congenital forms of hypoventilation will not be included.

We will include studies where the majority of participants are aged 16 years and over.

Uncertain or contentious cases regarding eligibility will be documented in the review and sensitivity analyses performed to assess the impact of these decisions on review findings.

Types of interventions

The intervention of main interest is non‐invasive positive airway pressure (PAP) therapy, regardless of support mode (i.e. continuous or bi‐level PAP therapy). We will consider these interventions together, in line with the current definition of OHS, which does not differentiate between those with and without obstructive sleep apnoea. We will include studies if PAP of any type is included in one or more treatment groups. We will define PAP as prescription of positive airway pressure support delivered during sleep, usually overnight, via a close‐fitting mask. Hence, studies including patients with tracheostomy devices on long‐term ventilatory support will not be included unless as a comparator group to non‐invasive PAP therapy.

We will categorise studies according to types of comparisons to account for intervention variability:

1. PAP therapy (any) compared to inactive control (e.g. placebo or usual care);

2. PAP therapy (any) compared to weight loss interventions by lifestyle modification (e.g. dietary or physical exercise modification, or both);

3. PAP therapy (any) compared to weight loss surgery;

4. PAP therapy (any) compared to pharmacotherapy (including oxygen);

5. PAP therapy (any) compared to tracheostomy with or without invasive ventilatory support;

6. PAP therapy (any) compared to an alternative mode of PAP therapy (e.g. continuous PAP compared to bi‐level PAP therapy).

We will include studies of PAP therapy with co‐intervention if the comparator group receives the same co‐intervention.

Types of outcome measures

Reporting one or more of the outcomes listed will not be an inclusion criterion for the review.

Electronic searches

We will identify studies from the following sources:

1. Cochrane Central Register of Controlled Trials (CENTRAL), which includes the Cochrane Airways Trials Register, through the Cochrane Register of Studies Online (crso.cochrane.org);

2. MEDLINE Ovid SP 1946 to date;

3. Embase Ovid SP 1974 to date;

The proposed Medline strategy is listed in Appendix 1. This will be adapted for use in the other databases. All databases will be searched from their inception to the present, and there will be no restriction on the language of publication. Handsearches of conference abstracts and grey literature will also be conducted through the CENTRAL database.

In addition, we will search the following trial registries:

1. US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov)

2. World Health Organization International Clinical Trials Registry Platform (www.apps.who.int/trialsearch)

Searching other resources

We will check the reference lists of all primary studies and review articles for additional references. We will search relevant manufacturers' websites for study information.

We will search for errata or retractions from included studies published in full text on PubMed and report the date this was done within the review.

Selection of studies

Two review authors (YC, HV) will independently screen the titles and abstracts of the search results and code them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. We will retrieve the full‐text study reports of all potentially eligible studies and two review authors (YC, HV) will independently screen them for inclusion, recording the reasons for excluding ineligible studies. We will resolve any disagreement through discussion or, if required, we will consult a third author (GM). We will identify and exclude duplicates and collate multiple reports of the same study so that each study, rather than each report, is the unit of interest in the review. We will record the selection process in sufficient detail to complete a PRISMA flow diagram and 'Characteristics of excluded studies' table (Moher 2009).

Data extraction and management

We will use a data collection form for study characteristics and outcome data, which has been piloted on at least one study in the review. Two review authors (YC, FG) will extract the following study characteristics from included studies.

1. Methods: study design, total duration of study, details of any 'run‐in' period, number of study centres and location, study setting, withdrawals and date of study.

2. Participants: N, mean age, age range, gender, severity of condition, diagnostic criteria, baseline lung function, smoking history, inclusion criteria and exclusion criteria.

3. Interventions: intervention, comparison, concomitant medications and excluded medications.

4. Outcomes: primary and secondary outcomes specified and collected, and time points reported.

5. Notes: funding for studies and notable conflicts of interest of trial authors.

Two review authors (YC, FG) will independently extract outcome data from included studies. We will note in the 'Characteristics of included studies' table if outcome data were not reported in a usable way. We will resolve disagreements by consensus or by involving a third review author (GM). One review author (YC) will transfer data into the Review Manager file (RevMan 2014). We will double‐check that data are entered correctly by comparing the data presented in the systematic review with the study reports. A second review author (FG) will spot‐check study characteristics for accuracy against the study report.

Assessment of risk of bias in included studies

Two review authors (YC, HV) will assess risk of bias independently for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will resolve any disagreements by discussion or by involving another author (GM). We will assess the risk of bias according to the following domains:

1. random sequence generation;

2. allocation concealment;

3. blinding of participants and personnel;

4. blinding of outcome assessment;

5. incomplete outcome data;

6. selective outcome reporting;

7. other bias.

We will judge each potential source of bias as high, low or unclear and provide a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We will summarise the 'Risk of bias' judgements across different studies for each of the domains listed. We will consider blinding separately for different key outcomes where necessary. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the 'Risk of bias' table.

When considering treatment effects, we will take into account the risk of bias for the studies that contribute to that outcome.

Measures of treatment effect

The outcome of survival will be measured using hazard ratios, and hospital utilisation using admission rate ratios, to be combined using the generic inverse variance method. Dichotomous data will be analysed using odds ratios (ORs) and continuous data using the mean difference (MD) or standardised mean difference (SMD). We will describe skewed data narratively. Where multiple trial arms are reported in a single study, we will include only the relevant arms. Data from cross‐over studies will be analysed using paired t‐tests and the generic inverse variance outcome in RevMan.

We will undertake meta‐analyses only where this is meaningful; that is, if the treatments, participants and the underlying clinical question are similar enough for pooling to make sense. If two comparisons are combined in the same meta‐analysis, we will either combine the active arms or halve the control group to avoid double‐counting. If data from rating scales are combined in a meta‐analysis, we will ensure they are entered with a consistent direction of effect.

We will use intention‐to‐treat (ITT) or 'full analysis set' analyses where they are reported (i.e. those where data have been imputed for participants who were randomly assigned but did not complete the study) instead of completer or per protocol analyses.

Unit of analysis issues

We will use participants as the unit of analysis.

Dealing with missing data

We will contact investigators or study sponsors in order to verify key study characteristics and obtain missing data where possible. If this is unsuccessful, we will impute missing values using replacement values, multiple imputation, or statistical models to allow for missing data, and then perform sensitivity analyses to assess the impact of changing the assumptions made. Where missing data are thought to introduce serious bias, we will take this into consideration in the GRADE rating for affected outcomes.

Assessment of heterogeneity

We will use the I² statistic to measure heterogeneity among the studies in each analysis. If we identify substantial heterogeneity we will report it and explore the possible causes using prespecified subgroup analysis. 

Assessment of reporting biases

If we are able to pool more than 10 studies, we will create and examine a funnel plot to explore possible small study and publication biases.

Data synthesis

We will combine data when the interventions, patient groups, and outcomes are sufficiently similar as determined by consensus. We will use a random‐effects model and perform a sensitivity analysis with a fixed‐effect model.

Subgroup analysis and investigation of heterogeneity

We plan to carry out the subgroup analyses for the following groups.

1. Presence of OSA (Apnoea‐Hypopnoea Index less than 5 versus 5 or more events/hour)

2. Presence of severe OSA (Apnoea‐Hypopnoea Index less than 30 versus 30 or more events/hour)

3. Type of non‐invasive PAP therapy (e.g. continuous PAP versus bi‐level PAP therapy)

4. Previous episode of decompensation (acute or acute on chronic hypercapnic respiratory failure)

5. Use of sham continuous PAP for control versus unblinded study

We will use the formal test for subgroup interactions in Review Manager (RevMan 2014).

Sensitivity analysis

We plan to carry out the following sensitivity analyses, removing the following from the primary outcome analyses.

1. Studies assessed as 'high' for overall risk of bias. We will consider studies to be at overall 'high' risk of bias if they meet criteria for high or unclear risk of bias in at least two domains in the risk of bias assessment, with at least one domain unrelated to blinding procedures

2. Studies considered contentious for inclusion in the review based on criteria set forth in Types of participants above

3. Industry‐sponsored trials

We will compare the results from a fixed‐effect model with the random‐effects model.