Standing the test of time: diagnostic accuracy of the Edmonton Symptom Assessment System-Revised (ESAS-r) for anxiety...

Depression and anxiety are often undertreated in patients with cancer, and brief, robust screening measures are needed. The objective was to assess the sensitivity and specificity of the Edmonton Symptom Assessment System-Revised (ESAS-r) anxiety and depression items against the Patient Health Questionnaire (PHQ-9) and the Generalized Anxiety Disorder 7-item scale (GAD-7) (reference measures).

We conducted a analysis of baseline data from the evaluation of a real-world patient-reported outcome measure (PROM) screening program. Patients with cancer completed the ESAS-r, PHQ-9, and GAD-7. Receiver operating characteristic analysis assessed ESAS-r depression and anxiety item accuracy by calculating area under the curve (AUC) and 95% confidence intervals (95% CI). Youden's index guided suggested screening cutoffs. Analyses were conducted for the entire sample, followed by exploratory analyses stratified by sex, age, income, and education.

Of 375 eligible participants, 177 with complete data were included. The AUCs were excellent against the reference measures: 0.81 (95% CI: 0.73-0.90) for depression and 0.83 (95% CI: 0.70-0.96) for anxiety. Optimal cutoffs were ≥ 2 for depression (sensitivity: 0.83, specificity: 0.66) and ≥ 4 for anxiety (sensitivity: 0.83, specificity: 0.79), with low positive predictive values (0.32 and 0.31, respectively). Exploratory stratified analyses suggested comparable performance across subgroups, anxiety prevalence in males (2%) was too low for meaningful analysis.

The AUCs for the depression and anxiety ESAS-r items were excellent, at 0.81 and 0.83, respectively. Findings support ESAS-r-D ≥ 2 and ESAS-r-A ≥ 4 as thresholds to identify those in need of further assessment, but professional judgment and previous findings should also be considered.