To inform the development of a national clinical guideline for Chronic Obstructive Pulmonary Disease (COPD), prioritized by the National Clinical Effectiveness Committee in Ireland, a systematic review was conducted to examine the cost-effectiveness of pulmonary rehabilitation programs (PRPs), outreach programs (OPs), and long-term oxygen therapy (LTOT), compared with usual care.

Medline, Embase, the Cochrane Library and grey literature sources were searched up to 19 June 2018. Studies evaluating cost-effectiveness published post-2008 in English were included. Screening, data extraction, and quality assessment using the Consensus Health Economic Criteria and International Society for Pharmacoeconomics questionnaires were conducted independently by two reviewers. Costs were converted to 2017 Irish Euro using consumer price indices for health and purchasing power parity.

From 8,661 articles identified, seven studies (one comparing both PRPs and LTOT) were included (PRPs: five; OPs: one; LTOT: two). PRP cost-utility analyses (n = 4) reported conflicting results due to considerable heterogeneity in program and study design, with incremental cost-effectiveness ratios (ICERs) ranging between EUR 12,391 and EUR 509,122 per quality adjusted life-year (QALY) gained. The remaining study investigated hospitalizations avoided and found outpatient and community-based PRPs to be dominant, while home-based PRP produced an ICER of EUR 1,913. OPs were found to be less costly, but also less effective. However, the results of the underpinning trial were neither statistically nor clinically significant. LTOT was found to be cost-effective, with ICERs of EUR 17,603 and EUR 26,936 per QALY gained.

Applying a willingness-to-pay threshold of EUR 45,000 per QALY gained, this systematic review found that, compared with usual care, there is inconsistent but generally favorable evidence for PRPs, no clear evidence for the cost-effectiveness of OPs, and that LTOT is likely to be cost-effective. However, there was a lack of methodologically robust studies included in the review and most were not directly transferable to the Irish context.