Low-quality of patient-reported outcome reporting in randomized clinical trials of major depressive disorder—a meta-epidemiological review

Abstract

ObjectiveOur goal was to review current peer-reviewed articles in which the BDI (Beck Depression Inventory), PHQ-9 (Patient Health Questionnaire), or QIDS-SR16 (16-Item Quick Inventory of Depressive Symptomatology) was used as the primary or secondary outcome measure and to evaluate the quality of PRO (Patient-Reported Outcome) reporting in RCTs (Randomized Controlled Trials) according to the 2013 PRO-specific CONSORT (Consolidated Standards of Reporting Trials) extension.MethodsWe systematically searched in electronic databases. A study would be included if it included patients diagnosed with major depressive disorder according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Classification of Diseases, version 10 (ICD-10) as participants, was a randomized controlled trial, included the BDI, PHQ-9, or QIDS-SR16 as the primary or secondary outcome measure, published between 1990 and 2013, and was in English. Two of the authors evaluated the quality of PRO reporting according to the 2013 CONSORT-PRO. Logistic regression were used to evaluate the association between reporting completeness and trial characteristics.ResultsA total of 116 studies were included. These studies were conducted in 25 countries. Sample sizes ranged from 12 to 750. The CONSORT-PRO was not cited in any one of the included studies. Among the 116 studies, 2 (1.72%) studies introduced the rationale for PRO assessment, 60 (51.72%) studies explicitly stated statistical approaches for dealing with missing data, 87 (75.00%) studies reported PRO outcome data at baseline and at subsequent time points. The mean score of reporting completeness was 66.24%. Significantly higher reporting completeness was found for RCTs published after 2013 (OR, 95%CI: 3.81, 1.32–10.99). Studies with a higher sample size were more completely reported than studies with a lower sample size (OR, 95%CI: 1.01, 1.00–1.02).ConclusionThe CONSORT-PRO guidance was rarely cited. The quality of PRO reporting in depression studies requires improvement. This result may be meaningful for the promotion of PRO reporting in RCTs.