The delivery of chemotherapy at home: an evidence synthesis

Background: Recent policy and guidance has focused on chemotherapy services being offered closer to home, but the clinical and economic implications of this are uncertain. Objectives: To compare the impact of delivering intravenous chemotherapy in different settings on a range of outcomes, including quality of life, safety and costs. Design: Multimethods approach: systematic review of clinical effectiveness, qualitative and cost-effectiveness studies; description of the patient pathway and brief survey of current provision; and development of a decision model to explore aspects of cost-effectiveness. Setting: Provision of intravenous chemotherapy. Participants: Chemotherapy patients. Interventions: Setting in which chemotherapy was administered (home, community or outpatient). Outcome measures: Safety, quality of life, preference, satisfaction, opinions/experiences, social functioning, clinical outcomes, costs and resource/organisational issues. Data sources: Sixteen electronic databases (including MEDLINE, EMBASE and The Cochrane Library) were searched from inception to October 2013 for published and unpublished studies. Review methods: Two reviewers independently screened potentially relevant studies, extracted data and quality assessed the included studies. Study validity was evaluated using appropriate quality assessment tools. Clinical effectiveness and cost-effectiveness studies were summarised narratively, and qualitative studies were synthesised using meta-ethnography. Results: Of the 67 eligible studies, 25 were comparative, with nine including a concurrent economic evaluation. Although some of the 10 randomised trials were designed to minimise avoidable biases, slow recruitment rates and non-participation of eligible patients for setting-related reasons meant that trial sample sizes were small and populations were inherently biased to favour the home or community settings. There was little evidence to suggest differences between settings in terms of quality of life, clinical outcomes, psychological outcomes or adverse events. All nine economic evaluations were judged as having low or uncertain quality, providing limited evidence to draw overall conclusions. Most were cost–consequence analyses, presenting cost outcomes alongside trial results but deriving no summary measure of benefit. Poor resource use reporting and use of different perspectives across settings made results difficult to compare. Seventeen qualitative studies (450 participants) were judged as moderate to good quality, although all compared new or proposed services with existing outpatient facilities and biased samples were used. The three main lines of argument were barriers to service provision, satisfaction with chemotherapy and making compromises to maintain normality. Most patients made explicit trade-offs between the time and energy required for outpatient chemotherapy, which reduced quality of life, and an increased sense of safety. A patient pathway was described, informed by expert advice and a brief survey of NHS and private providers, which identified wide variation in the ways in which home and community chemotherapy was delivered. Considering limitations of the available data and variation in provision, cost-effectiveness modelling results were not robust and were viewed as exploratory only; the results were highly unstable. Conclusions: Primary studies comparing settings for administering intravenous chemotherapy appear difficult to conduct. Consequently, few robust conclusions can be made about the clinical effectiveness and cost-effectiveness. Qualitative studies indicate that the patient time and energy required for outpatient chemotherapy reduces quality of life. A nested randomised controlled trial within a larger observational cohort of patients is proposed to enhance recruitment and improve generalisability of results. Future economic evaluations require detailed patient characteristic, resource use, cost and quality-of-life data, although their results are likely to have limited generalisability. Study registration: This study is registered as PROSPERO CRD42013004851. Funding: The National Institute for Health Research Health Services and Delivery Research programme