Abstract
BACKGROUND: Chronic obstructive pulmonary disease (COPD) remains a significant public health burden. Non-invasive ventilation (NIV) is a method of supported breathing used as standard care for acutely unwell patients in hospital with COPD, but there is uncertainty around the potential benefits of using NIV in the treatment of stable patients in a non-hospital setting. This is a protocol for systematic reviews of the clinical and cost-effectiveness of NIV in this context, being undertaken in support of a model based economic evaluation. METHODS/DESIGN: Standard systematic review methods aimed at minimising bias will be employed for study identification, selection and data extraction for both the clinical and economic systematic reviews. Bibliographic databases (for example MEDLINE, EMBASE) and ongoing trials registers will be searched from 1980 onwards. The search strategy will combine terms for the population with those for the intervention. Studies will be selected for review if the population includes adult patients with COPD and hypercapnic respiratory failure, however defined. Systematic reviews, randomised controlled trials and observational studies (with n >1) will be included, and quality assessment will be tailored to the different study designs. The primary outcome measures of interest are survival, quality of life, and healthcare utilisations (hospitalisation and Accident and Emergency attendances). Meta-analyses will be undertaken where clinical and methodological homogeneity exists, supported by predefined subgroup analyses where appropriate. A systematic review of the evidence on the cost-effectiveness of non-hospital NIV will be completed, and a model-based cost-utility analysis undertaken to determine the cost-effectiveness of non-hospital-based NIV compared with standard care. DISCUSSION: These reviews will attempt to clarify the clinical effectiveness of non-hospital NIV in COPD patients as well as the cost-effectiveness. The findings may indicate whether NIV in a non-hospital setting should be considered more routinely in this patient group, and what the likely cost implications will be. PROSPERO REGISTRATION: 2012:CRD42012003286.