Short screening tools for cancer-related distress: a review and diagnostic validity meta-analysis

Clinicians are increasingly seeking efficient methods to identify distress in cancer settings, using short screening tools with fewer than 14 items that take less than 5 minutes to complete. This article examines the value of these tools for identifying cancer-related distress, defined by semi-structured interview. An updated search, appraisal, and meta-analysis, with adjustments made for heterogeneity and underlying prevalence variations, identified 45 potentially useful short and ultra-short tools, although most were intended to help diagnose depression, with few targeted at distress (or anxiety). Very few studies attempted robust validation in cancer settings. When studies were limited to those tested against distress defined by semi-structured interview, only 6 methods had been validated, namely the Hospital Anxiety and Depression Scale (HADS; 13 studies, 14 items), the Distress Thermometer (DT; 4 studies, 1 item), a single verbal question (4 studies, 1 item), the Psychological Distress Inventory (PDI; 1 study, 13 items), combined DT and an impact thermometer (1 study, 2 items), and combined 2 verbal questions (1 study, 2 items). Comparing these 6 approaches side-by-side suggests that for screening, all tools have approximately the same accuracy. Therefore, choice of a short screening tool for distress can be based on acceptability or cost-effectiveness. Here, best evidence supports use of the DT or single verbal question. Remarkably, the overall accuracy of these single-item approaches seems comparable to that of the 14-item HADS (total score), whereas their efficiency is superior. For case-finding, data are sparse but no method seems to be entirely satisfactory. Current evidence suggests that the optimal short methods for identifying distress are 2 verbal questions or PDI. Of these approaches, the 2 verbal questions has superior efficiency. All short methods may be augmented by repeated application, an assessment of unmet needs (problem list), and clarification regarding the need for professional help. No screening tool should be seen as an alternative to careful clinical assessment and management. Despite much interest in the development of short and ultra-short tools, data on validation and implementation are currently incomplete. Nevertheless, short methods seem to be at least as successful as the HADS, although substantially more efficient and hence more acceptable, and therefore may be a suitable initial method of assessment in busy clinical settings.