Measuring the environmental impact of health interventions in randomised controlled trials - a scoping review.

Abstract

Background: Healthcare accounts for 4% of the overall greenhouse gas emissions worldwide. Reducing carbon emissions is a key strategic aim for health systems in multiple countries. To achieve this, environmental data needs to be considered in health decision making. Clinical trials are a key element to inform clinical practice, but to what extent they have considered environmental outcomes in their design, analysis and interpretation is unclear. We conducted a scoping review of environmental outcomes in clinical trials, including protocols, to assess what outcomes are being collected, how and why, and to what extent environmental outcomes influence interpretation of the clinical findings by trial authors.

Methods: We developed a search strategy in Embase, MEDLINE and Cochrane library to identify eligible English language studies with no date restrictions. Inclusion criteria were healthcare randomised controlled trials (RCTs) that collected, or protocols of RCTs that planned to collect, environmental outcomes in their abstracts. Two researchers conducted abstract and full text screening independently. We single extracted data using a pre-agreed and piloted data extraction form. Any uncertainties in extraction were discussed as a group until an agreement was reached. We summarised quantitative data using descriptive statistics and analysed verbatim text of qualitative data into summary categories that were agreed by the team.

Results: We identified 1,318 abstracts, from which 27 full texts were screened, and 16 RCTs (or protocols of RCTs) were included in our review. Nine studies were published from 2022; nine reported environmental outcomes in secondary publications; seven were in the nutrition field followed by oncology (n=2) and anaesthesiology (n=2). The primary reason for including environmental outcomes was to confirm that the intervention reduced emissions (n=12), rather than focusing on broader goals of lowering carbon emissions or environmental considerations in health systems. Included trials assessed various environmental impacts of the studied interventions including carbon footprint, greenhouse gas emissions, or wider environmental impacts. Most trials used life cycle assessments, or publicly available carbon footprint data, to calculate their environmental outcomes. The analysis was not pre-specified in eight trials; all trials but one undertook separate analyses for clinical and environmental outcomes. In all but two cases, environmental and clinical outcomes were in agreement (in favour of, or against, the intervention). Patient and public involvement was rarely reported (n= 3). Five trials presented examples or contextual information to aid interpretation of the environmental outcome results.

Discussion: Including environmental outcomes in RCTs appears to be increasing in prominence, although the role of these outcomes in clinical decision making is still uncertain. It is crucial for trialists to work with environmental experts to develop appropriate design, and analysis plans to include these outcomes. Patient, public and health professionals' involvement will be key in ensuring acceptability of the use of environmental outcomes to inform clinical and policy decision making.