Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Positive airway pressure for heart failure associated with central sleep apnoea

Authors' declarations of interest

Version published: 23 September 2017 Version history

https://doi.org/10.1002/14651858.CD012803

This is not the most recent version

view the current version 04 December 2019
Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess effects of positive airway pressure therapy for people with heart failure who experience central sleep apnoea.

Background

Description of the condition

Ischaemic heart disease including heart failure is the most common cause of death in the world, with an incidence of 8.76 million deaths in 2015 according to Global Health Observatory data reported by the World Health Organization (WHO 2017). Furthermore, global prevalence of heart failure was reported at approximately 40 million by the Global Burden of Disease study (GBD 2016). Heart failure, which is defined as a clinical syndrome characterised by symptoms and signs stemming from a structural and/or functional cardiac abnormality, occurs in approximately 2% of adults and in more than 10% of elderly people (Mosterd 2007; Ponikowski 2016). About 30% to 40% of patients survive for one year after onset of heart failure, and rates of hospital re‐admission and of cardiac events are very high (Mozaffarian 2015). Heart failure consists of two subcategories ‐ heart failure with reduced ejection fraction (HF‐REF) and HF with preserved ejection fraction (HF‐PEF). The left ventricular ejection fraction in HF‐REF and in HF‐PEF is less than 40% and greater than 50%, respectively (Ponikowski 2016). Patients with left ventricular ejection fraction in the range of 40% to 49% represent a grey area (Ponikowski 2016). Individuals with HF‐REF are reported to have a 32% higher risk of mortality compared with patients with HF‐PEF (MAGGIC 2012). The incidence of heart failure has dramatically increased over the past few decades and is expected to continue to rise over the next 20 years (Mozaffarian 2015; Okura 2008).

Heart failure is characterised by symptoms such as fatigue and breathlessness during light activity, as well as disordered breathing during sleep. In particular, approximately 30% to 40% of people with heart failure experience sleep‐disordered breathing (SDB), including obstructive sleep apnoea (OSA) and central sleep apnoea (CSA) (Javaheri 1998; Sin 1999; Wang 2007). OSA and CSA are associated with a worse prognosis in heart failure (Ponikowski 2016). OSA is caused mainly by narrowing or occlusion of the upper airway due to obesity and reportedly occurs in 38% of male and 31% of female people with heart failure (Sin 1999). CSA, the most common condition of SDB among those with HF‐REF, reportedly occurs in 25% to 40% of patients with HF‐REF (Lévy 2007; McKelvie 2011). CSA is reportedly caused by low cardiac output, high sympathetic activation, and pulmonary congestion (Naughton 2017). SDB leads to hypertension through sympathetic activation, intrinsic endothelial dysfunction, and pressor effects of vasoactive peptides that sustain elevated blood pressure (Kasai 2012; Shahar 2001; Yoshihisa 2013). Furthermore, these conditions independently increase the risk of cardiovascular mortality for people with CSA (Bradley 2005). One meta‐analysis revealed that risk of mortality among people with heart failure and CSA was 1.48 times higher than in those without SDB (Nakamura 2015).

Three previous studies reported that positive airway pressure (PAP) therapy such as continuous PAP (CPAP), bi‐level PAP (BiPAP), and adaptive servo‐ventilation (ASV) improved hypopnoea, SDB, cardiac function, and haemodynamic status (Bradley 2005; Mansfield 2004; Yoshihisa 2013). Furthermore, PAP therapy was shown to dramatically increase the survival rate of patients with heart failure who had CSA and thus could contribute to improving the prognosis of these individuals (Nakamura 2015; Yoshihisa 2013).

Description of the intervention

Positive airway pressure therapy is a physical treatment in which patients wear a nasal or facial mask during sleep. The airstream blown through the mask acts as a pneumatic splint and keeps the airway open, thus lowering the risk of blocked airways that can cause hypertension through sympathetic activation, and increasing endothelial dysfunction and pressor effects of vasoactive peptides. European Society of Cardiology (ESC) Guidelines recommend PAP therapy to improve outcomes caused by SDB, especially OSA for people with heart failure (Ponikowski 2016). However, ASV in PAP therapy is not recommended for patients who have HF‐REF and CSA (Ponikowski 2016). In one meta‐analysis by Nakamura, ASV in PAP therapy reduced all‐cause mortality (risk ratio 0.13, 95% confidence interval 0.02 to 0.95) among people with heart failure who had CSA, whereas CPAP was not associated with changes in mortality rate (Nakamura 2015).

How the intervention might work

Use of PAP therapy for people with heart failure and SDB has been reported to decrease the frequency of episodes of sleep apnoea and hypopnoea, and to improve cardiac function and exercise capacity (Bradley 2005). For this reason, PAP improves pulmonary congestion by reopening collapsed alveoli, increasing lung volume, preventing peripheral airway occlusion, improving oxygenation and lung compliance, and reducing cardiac preload through reduced venous return (Mehta 2001; Naughton 1995; Naughton 2017; Takano 1986). Therefore, PAP therapy should be considered for people with respiratory distress (class of recommendation ‐ IIa; level of evidence ‐ B) according to the ESC Guideline (Ponikowski 2016).

Why it is important to do this review

PAP therapy is reportedly effective for people with OSA when provided according to available guidelines (Randerath 2012; Yumino 2013). Therefore, PAP therapy may be considered to treat nocturnal hypoxaemia in OSA (Randerath 2012; Yumino 2013). However, the effectiveness of PAP therapy for people with heart failure who have CSA is unclear. A trial conducted in 11 countries, known as the SERVE‐HF trial, which investigated the effects of ASV on survival and cardiovascular outcomes of people with HF‐REF and CSA, reported that the ASV group had an almost 1.3 times higher risk of all‐cause mortality and cardiovascular mortality than the control group (Cowie 2015). In light of this trial's results, the ESC Guideline has suggested that ASV is not recommended for HF‐REF and predominantly CSA. However, the SAVIOR‐C trial, conducted in Japan, which investigated effects of ASV in HF‐REF regardless of CSA, reported no significant differences in cardiovascular events between ASV and control groups (Momomura 2015). Futhermore, clinical composite response, New York Heart Association (NYHA) classification, and symptoms during daily activities showed greater improvement in the ASV group than in the control group. As results of these and other trials were different, the effectiveness of PAP for people with heart failure and CSA remains unclear. This review aims to determine the benefits and harms of PAP therapy for people with heart failure who have CSA.

Objectives

To assess effects of positive airway pressure therapy for people with heart failure who experience central sleep apnoea.

Methods

Criteria for considering studies for this review

Types of studies

We will include individual parallel and cluster‐randomised controlled trials (RCTs). We will exclude cross‐over trials and will include full‐text studies, studies for which only the abstract has been published, and unpublished data.

Types of participants

We will include participants 18 years of age or older with a diagnosis of heart failure with CSA. We will include HF‐REF and HF‐PEF. If a study includes only a subset of eligible participants, we intend to ask study authors to provide data on the subset of interest. Furthermore, when we cannot obtain data for the subset of eligible participants, we will exclude them from whole data and will distribute these studies to “Studies awaiting classification." We will define CSA by polysomnography and fulfilment of the following criteria: apnoea‐hypopnoea index ≥ 5 events per hour, and 3% of oxygen desaturation index ≥ 5 events per hour. We will exclude participants with an implantable ventricular assist device and those who have undergone heart transplantation.

Types of interventions

We will include trials comparing PAP (CPAP, BiPAP, or ASV) therapy versus usual care consisting of medical therapy based on relevant guidelines (Hunt 2009; McMurray 2012). Standard mechanical settings are used for PAP therapy, and expiratory positive airway pressure can be increased manually to manage sleep apnoea. Participants are requested to use the PAP device for three hours per night on average to ensure the efficacy of PAP therapy (Cowie 2015).

Types of outcome measures

We will assess all outcome measures and clinical events after randomisation with a minimum follow‐up of six months because this time point is clinically important according to the ESC guideline (Ponikowski 2016).

Primary outcomes

  1. All‐cause mortality

  2. Cardiovascular (cardiac‐related) mortality

  3. Cardiac‐related rehospitalisation: defined as total number of rehospitalisations

  4. Health‐related quality of life as assessed by validated questionnaires

  5. Adverse events (PAP device‐related and non‐device‐related): defined as total number of adverse events

Secondary outcomes

  1. SDB markers; apnoea‐hypopnoea index, respiratory disturbance index, and oxygen desaturation index

  2. Cardiovascular and cerebrovascular function markers: blood pressure, echocardiography, and left ventricular ejection fraction

  3. Physical function markers: exercise capacity, six‐minute walking distance, and NYHA classification

Reporting in the trial one of more of the outcomes listed here is not an inclusion criterion for this review.

Search methods for identification of studies

Electronic searches

We will identify trials through systematic searches of the following bibliographic databases.

  1. Cochrane Central Register of Controlled Trials (CENTRAL), in the Cochrane Library.

  2. MEDLINE (Ovid).

  3. Embase (Ovid).

  4. Web of Science Core Collection (Thomson Reuters).

We will adapt the preliminary search strategy devised for MEDLINE (Ovid) (Appendix 1) for use in searches of other databases. We will apply the Cochrane sensitivity‐maximising randomised controlled trial (RCT) filter to MEDLINE (Ovid) searches, and adaptations of it to searches of the other databases, except CENTRAL (Lefebvre 2011).

We will conduct a search of ClinicalTrials.gov (www.ClinicalTrials.gov) and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) Search Portal (http://apps.who.int/trialsearch/). We will search all databases from their inception to the present, regardless of the language of publication.

We will not perform a separate search for adverse effects of interventions used for treatment of CSA. We will consider only adverse effects described in the included studies.

Searching other resources

We will check reference lists of all primary studies and review articles for additional references. We will contact experts in the field and will ask if they know of any ongoing or unpublished trials. We will also examine any relevant retraction statements and errata for included studies.

Data collection and analysis

Selection of studies

Two review authors (SY and TY) will independently screen titles and abstracts for inclusion of all potential studies identified as a result of the search and will code them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. If disagreements arise, we will ask a third review author (KN) to arbitrate. We will retrieve full‐text versions of study reports, and three review authors (SY, TY, and KN) will independently screen them to identify studies for inclusion, and to identify and record reasons for exclusion of ineligible studies. We will resolve disagreements through discussion and, if required, will consult other review authors (CN and RM). We will identify and exclude duplicates and will 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' tables.

Data extraction and management

We will use a data collection form that has been piloted on at least one study in the review to record study characteristics and outcome data. Two review authors (SY and TY) will extract the following study characteristics from included studies.

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

  2. Participants: number of participants who were randomised, withdrew, or were lost to follow‐up, and were analysed; 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 trial, and notable conflicts of interest of trial authors.

Two review authors (SY and TY) will independently extract outcome data from included studies. We will resolve disagreements by reaching consensus or by involving a third review author (KN). One review author (SY) will transfer data into Review Manager 5.3 (RevMan 2014). We will double‐check that data have been entered correctly by comparing data presented in the systematic review against study reports. A second review author (TY) will spot‐check study characteristics against trial reports for accuracy.

Assessment of risk of bias in included studies

Three review authors (SY, TY, and KN) will independently assess risk of bias for each study, using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will resolve disagreements by discussion or by consultation with other review authors (CN and RM). We will assess 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 (e.g. industry funding).

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

When considering treatment effects, we will consider risk of bias that may have affected a given outcome. Note that given the nature of PAP therapy, study participants and staff cannot be blinded.

Assessment of bias in conducting the systematic review

We will conduct the review according to this published protocol and will report deviations from it in the 'Differences between protocol and review' section of the systematic review.

Measures of treatment effect

We will express dichotomous data as risk ratios (RRs) with 95% confidence intervals (CIs). For continuous variables, we will compare net changes (i.e. intervention group value minus control group value). For each trial, we will ascertain the mean change (and standard deviation (SD)) in outcomes between baseline and follow‐up for both exercise and control groups; if this information is not available, we will use the absolute mean (and SD) outcome at follow‐up for both groups. We will express the results as mean differences (MDs). For studies employing a different scale or measurement system, we will use standardised mean differences (SMDs).

Unit of analysis issues

We will include cluster‐randomised trials in the analyses, along with individually randomised trials. We will adjust sample sizes or standard errors according to methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions, using an estimate of the intracluster correlation coefficient (ICC) as derived from the trial, from a similar trial, or from a study of a similar population (Higgins 2011). If we use ICCs from other sources, we will report these results and will conduct sensitivity analyses to investigate effects of variation in the ICC. If we have identified both cluster‐randomised trials and individually randomised trials, we plan to synthesise the relevant information. We will consider it reasonable to combine results from both if we note little heterogeneity between study designs, and if the interaction between effect of the intervention and choice of randomisation unit is considered unlikely. We will also acknowledge heterogeneity in the randomisation unit, and will perform a sensitivity analysis to investigate effects of the randomisation unit if necessary. If results of trials are presented for several periods of follow‐up (e.g. six months, one year, two years), we plan to perform separate analyses for short‐term and long‐term periods. If each study compares exactly the same interventions (i.e. CPAP, BiPAP, or ASV), we will split the usual care group into two or more groups with smaller sample sizes and will compare intervention groups against those usual care groups (i.e. CPAP vs half the usual care group, or BiPAP vs half the usual care group).

Dealing with missing data

We will contact investigators or study sponsors to verify key study characteristics and to obtain missing numerical outcome data when possible (e.g. when a study is identified as abstract only). When this is not possible, and missing data are thought to introduce serious bias, we will explore the impact of including such studies in the overall assessment of results by performing a sensitivity analysis.

Assessment of heterogeneity

We will visually inspect forest plots for signs of heterogeneity and will try to explain heterogeneity qualitatively by comparing the characteristics of included studies. We will use I² and Chi² statistics to quantitatively measure heterogeneity among the trials in each analysis. If we identify substantial heterogeneity (i.e. I² score > 50% and P < 0.05), we will report this and will explore possible causes by conducting prespecified subgroup analysis.

Assessment of reporting biases

If we are able to pool more than 10 trials, we will create and examine a funnel plot to explore possible small study biases for primary outcomes. If we suspect publication bias, we will carry out a simulation to investigate possible small‐study effects.

Data synthesis

We will undertake meta‐analyses only when this is meaningful (i.e. when treatments, participants, and the underlying clinical question are similar enough for pooling to make sense). Furthremore, we will narratively report non‐device‐related adverse events. We will pool data from each study using random‐effects modelling when appropriate. To examine robustness of results, we will perform meta‐analyses using fixed‐effect models after attributing less weight to small trials and will compare fixed‐effect pooled estimates or 95% CIs versus random‐effects pooled estimates or 95% CIs. Therefore, we will use a random‐effects model and will perform subgroup analysis when we identify statistical substantial heterogeneity. If substantial heterogeneity remains after subgroup analysis, we will not perform a meta‐analysis but will only conduct a systematic review according to recommendations provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

'Summary of findings' table

We will create a 'Summary of findings' table using the following outcomes.

  1. All‐cause mortality.

  2. Cardiovascular (cardiac‐related) mortality.

  3. All‐cause rehospitalisations.

  4. Cardiac‐related rehospitalisations.

  5. Health‐related quality of life.

  6. Total adverse events.

  7. Cardiac‐related adverse events.

We will use the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to assess the quality of a body of evidence as it relates to studies that contribute data to meta‐analyses for prespecified outcomes (Guyatt 2008). We will use methods and recommendations as described in Section 8.5 and Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions, using GRADEproGDT software (GRADEproGDT 2017; Higgins 2011). We will justify all decisions to downgrade the quality of studies by using footnotes, and we will make comments to aid readers' understanding of the review when necessary.

Subgroup analysis and investigation of heterogeneity

We plan to carry out the following subgroup analyses for primary outcomes if an I2 score > 50% is obtained.

  1. Age (≤ 65 years or > 65years).

  2. Sex (men and women).

  3. Left ventricular ejection fraction (≤ 45% or > 45%).

  4. Device (CPAP, BiPAP, or ASV).

  5. Apnoea hypopnoea index (< 15/h or ≥ 15/h)

  6. Cheyne‐Stokes breathing (yes or no).

  7. Follow‐up (≤ 1 year or > 1 year).

We will use the formal test for subgroup interactions available in Review Manager 5.3 (RevMan 2014).

Sensitivity analysis

We will carry out a sensitivity analysis for primary outcomes if high risk of bias of some included studies might affect study results. We will define as 'high risk' a study having high risk in terms of random sequence generation; inadequate allocation concealment; and greater than 20% of data missing (Tierney 2005). We plan to carry out the following sensitivity analyses.

  1. Exclusion of studies at high risk of bias.

  2. Exclusion of trials with 10 or fewer events.

  3. Exclusion of cluster RCTs.

Reaching conclusions

We will base our conclusions only on findings from the quantitative or narrative synthesis of studies included in this review. We will avoid making recommendations for practice, and our implications for research will suggest priorities for future research and will outline remaining uncertainties in this area.