Malpass et al, in a mixed methods study assessing PHQ-9 scores and patients' experiences, report that patients found the PHQ-9 to be helpful,1 concurring with other recent qualitative work that suggested that patients viewed such measures as an ‘objective adjunct to medical judgement’.2 Unfortunately, the value of these observations rests entirely on the assumption that the PHQ-9 is a valid measure of depression severity. Considerable doubt attends this premature notion.3,4 Indeed, the most recent of these findings is reported by Reddy and colleagues on the pages following Malpass et al's piece.5 We should not be comforted by the observation that patients' believe their depression is being better assessed by this process until it is shown that this belief matches the evidence.
A further finding of Malpass et al was of discord between symptom frequency and intensity in relation to the PHQ-9 and patients’accounts.1 This raises an important consideration for the use of the PHQ-9 in assessing depression severity and treatment responsiveness. If depression severity measures are intended to facilitate the alignment of clinical decision making to evidence-based interventions, consideration should be given to how severity of depression was measured in that evidence base. Guidelines indicate6 that largely this has been in studies where depression severity has been measured with the Hamilton Depression Rating Scale.7 With regard to how to administer this measure, Hamilton states that ‘no distinction is made between intensity and frequency of symptom, the rater having to give due weight to both of them in making his judgment’. It is not surprising that the PHQ-9, with its sole emphasis on symptom frequency, fails to probe important aspects of the patient experience of the severity of depressive symptoms.
The authors also state that they ‘are aware of only one study that considers sensitivity to change over time of the PHQ-9’ however, they may like to expand their reading to include a study of ours. We assessed the sensitivity to change over time of the PHQ-9, relative to the Hospital Anxiety and Depression Scale, Depressive subscale (HADS-D), in a sample of patients referred to primary care mental health workers.3 At end of treatment, in a sample of 491, the PHQ-9 and HADS-D demonstrated similar effect sizes (0.99 and 1 respectively). However, while the HADS-D provided a useful reference standard, in that there is evidence of the scale measuring treatment responsiveness,8 further work is required to assess the sensitivity of change over time of the PHQ-9 relative to a more stringent reference standard.