Missing and masked: equity in a systematic review of remote interventions for substance misuse.
Preethy D'Souza, EPPI Centre, University College London, United Kingdom
Helen Burchett
Irene Kwan
Katy Sutcliffe, EPPI Centre, University College London, United Kingdom
2Eppi Centre, UCL, London, United Kingdom
Background: To reduce inequity, knowing whether interventions ‘work’ is not enough; we also need to know whether vulnerable subgroups experience them differently. However, vulnerable groups may not have equal opportunity to participate in research and, hence, may be underrepresented or missing entirely. Unless study findings are reported by subgroup, any differences among those included will be masked.
Objectives: As part of a systematic review (SR) exploring the effectiveness of remote therapies for drug/alcohol misuse, we undertook equity-related work with the objectives of understanding (1) who was excluded from the research, (2) who was represented in the study populations and (3) whether sub-groups experienced outcomes differently.
Methods: For each included primary study we assessed (1) the exclusion criteria; 2) the sociodemographic characteristics reported using the PROGRESS-plus framework; and (3) whether substance use outcomes were reported by equity-relevant characteristics and, if so, whether there was any variation.
Results: A total of 52 studies were included in the review, of which 47 had eligibility criteria with the potential to result in the exclusion of vulnerable subgroups, most notably those with mental health conditions (MHCs) and those without access to digital technology. Few studies justified the exclusions or reported how many people were excluded based on each criterion. Sociodemographic characteristics of the study populations were reported for all studies, although what the characteristics were and how they were measured varied. Most commonly and consistently reported were gender, age and ethnicity but it was rare that the study outcomes were reported by these or other factors.
Conclusions: The exclusion of certain vulnerable subgroups from the research process and the lack of reporting of outcomes by potential axes of inequalities have important implications for the interpretation and generalizability of findings of SRs. Particularly troubling in our review was the frequency with which those with MHCs were excluded. Understanding how different groups experience interventions is key to allow practitioners to be able to make informed decisions about their adoption in specific vulnerable groups and to ensure that interventions do not inadvertently produce, or exacerbate, inequities.
Patient, public and/or healthcare consumer involvement: An advisory group from policy and practice inputted to all aspects of the SR.