How do researchers select usual care comparators when designing primary care trials of complex interventions?

Problem

Primary care trials assessing complex health interventions often evaluate new or modified interventions against “usual care”. Typically, usual care refers to the care provided in everyday clinical practice. However, usual care can differ for the same condition and across clinical sites, countries and over time. This variability may raise ethical concerns in terms of the care trial participants allocated to usual care will receive, and can affect the design of a trial, e.g. what sample size is needed to detect clinically significant differences between treatments. Where usual care for a specific condition is highly variable, researchers may choose to specify at the start of their trial what usual care will entail. This, however, may affect the trial’s overall design and delivery, and reduce its relevance to clinical practice.The impact the content of usual care can have on a trial’s design has led to researchers and reporting guidelines, such as TIDieR, requesting that researchers fully detail both the intervention and the comparator arm when designing a trial. To establish whether primary care researchers are doing this, and to identify on what basis they decide its content, we conducted a systematic review of recent trial protocols describing primary care trials of complex interventions.

Approach

We searched MEDLINE, Embase, the Cochrane Library and PsycINFO from 1 July 2020 to 20 June 2022. We identified 6063 records from searching databases. After de-duplication, 4077 records were screened. 293 articles were included for full-text screening and 87 were included in the review. Data were analysed using a narrative synthesis approach.

Findings

Majority of the included studies were from the UK (n=15), USA (n=11), Australia (n=7), Netherlands (n=7) and Canada (n=5). Majority of the included studies (n=45) used the term usual care/care as usual/usual treatment. Other studies commonly used treatment as usual (n=11), standard care/practice/standard of care (n=11) and control arm/control group/condition (n=9). Only 37 studies provided justification of choosing usual care as comparator. Nearly all the included studies provided some definition or description of usual care. However, the information on the content and delivery of usual care including who delivered it and where was often vague and patchy. Only 14 of the 87 studies provided a justification of choice of usual care content including some information on actions that informed this.

Consequences

While the importance of describing usual care prior to a trial starting has been highlighted within the literature and by reporting checklists, such details continue to be missing from published trial protocols.