Background: Chronic pain and depression are common comorbid conditions, but there is limited evidence-based guidance for management of the two conditions together. In recent years, there has been an increase in the number of chronic pain randomized controlled trials that collect depression outcomes, but it is unknown how often these trials include people with depression or significant depressive symptoms. If trials do not include participants representative of real-world populations, evidence and guidance generated from these trials risk being inapplicable for large proportions of the target population, or worse, risk harm. Thus, in order to identify pathways to improve the conduct of clinical trials, the aims of this study were to (1) estimate the proportion of randomized controlled trials evaluating chronic pain interventions and reporting depression outcomes that include participants with significant depressive symptoms; and (2) assess the variability of inclusion proportions by pain type, intervention type, gender, country of origin, and publication year.

Methods: Studies were extracted from an umbrella review of interventions for chronic pain that reported depression outcomes. Screening and data extraction were completed in duplicate and conflicts were resolved by a third author. Randomized controlled trials with at least 50% adult participants and validated depression scales were included, and randomized controlled trials with populations whose mean scores were at or above depression thresholds at baseline were considered to have included participants with depression.

Results: Of the 346 randomized controlled trials analyzed, 142 (41%) included participants with depression. Eight pain-type groups and nine intervention types were identified. Randomized controlled trials investigating fibromyalgia and mixed chronic pain had the highest proportion of participants with depression, whereas studies of arthritis and axial pain had among the lowest. Randomized controlled trials from the United States had a significantly lower inclusion proportion compared with non-US studies, especially for studies on arthritis. The increase in inclusion proportion by publication year was driven by the increase in fibromyalgia studies.

Discussion and conclusion: This study highlights opportunities to improve the conduct of chronic pain clinical trials. The majority of randomized controlled trials s analyzed evaluated participants without significant depressive symptoms at baseline, thus the findings synthesized in systematic reviews and subsequent guidelines are most applicable to the subset of real-world populations that do not have significant depressive symptoms. As well, systemic biases around psychological conditions and gender may be important contributors to differences in the study of depression in fibromyalgia compared with common conditions such as arthritis and axial pain. In order to better inform clinical practice, future research must intentionally include individuals with comorbid depression in trials of common chronic pain conditions, and consider methods to mitigate biases that may distort study design.