January 2010
Reviewer: Jeff C. Huffman, MD
An ultra-brief screening scale for anxiety and depression: the PHQ-4
Background: Depression and anxiety in patients with medical illness are common, underrecognized, and often associated with negative medical outcomes. Finding accurate and brief tools for screening patients in medical settings for depressive and anxiety disorders is a vital first step in improving patient outcomes.
Methods: The authors have developed a pair of measures to assess depression and anxiety (the PHQ-9 for depression and the GAD-7 for anxiety, respectively). They have also adapted these measures to develop very brief screening tools for depression and anxiety (the PHQ-2 assesses depressed mood and anhedonia; the GAD-2 assesses anxiety and difficulty controlling worry). The authors have now created a combined brief scale, the PHQ-4, that combines the two items from the PHQ-2 and the two items from the GAD-2, to screen for both depression and anxiety disorders.
The authors describe an evaluation of the potential utility of this PHQ-4, by reviewing information gathered when the GAD-7 was validated and tested.
In this trial, the GAD-7, PHQ-8 (PHQ-9 items except for the suicidal ideation question), and the SF-20 (to assess functional status) were administered to patients at 15 primary care clinics; a random sample of subjects was also contacted by phone and received structured psychiatric interviews from assessors (blinded to GAD-7/PHQ-8 scores) to assess patients for formal anxiety diagnoses. In this report, the authors describe the apparent utility of the PHQ-4 in this setting (i.e., how well the scores on the first two items of the GAD-7 and PHQ-8 corresponded to anxiety diagnoses by formal interview).
Results: In sum, 2149 patients received the screening, with 965 undergoing subsequent standardized interview by phone. In the total sample, 19% of patients had at least moderate depressive sx (PHQ-8=10) and 20% had at least moderate anxiety (GAD-7=10). Mean PHQ-4 score was 2.5, with 38.5 of subjects having a score of 3 or greater out of a total of 12 points, and increasing PHQ-4 scores were strongly correlated with poorer function in multiple domains as measured by the SF-20. With regard to anxiety disorder diagnoses, the two anxiety items on the PHQ-4 (i.e., the GAD-2) served as an excellent screening tool for GAD (area under the curve [AUC]>.90), and a good screening tool for panic disorder, social phobia, and PTSD (AUC>.80). Importantly, the two anxiety and two depression items appeared to measure different constructs/dimensions of illness, with the PHQ-2 items loading heavily on depression and GAD-2 on anxiety.
Discussion: This paper confirms the utility of short screening tools in the identification of depression and anxiety in medical settings. The PHQ-4 as a combined tool makes excellent sense, as the PHQ-2 has become the depression screening tool of choice in many settings, and the GAD-2 seems a parallel and useful anxiety screening tool. The ability of the GAD-2 to identify cases of non-GAD anxiety disorders was also surprisingly good, though if one works in a population with high rates of panic disorder or PTSD, it may be worthwhile to add a third question specifically focused on that disorder. This evaluation was limited by the fact that it did not seem originally designed to be a prospective trial of the PHQ-4 per se (but rather a post hoc evaluation of the 4 items), and that the responsiveness of the PHQ-4 to treatment has not been identified. Despite this, the PHQ-4 now seems to be the screening tool of choice in those settings in which one wishes to screen for both depression and anxiety, given its ease of administration and strong operating characteristics. One caveat: screening alone (or combined with simple feedback to treaters) does not improve patient outcomes, so the screening must be combined with a systematic and more comprehensive intervention, such as a triggered psychiatric consultation or further evaluation from a care manager.