A study published today in The BMJ assessed the diagnostic accuracy of the Patient Health Questionnaire-9 (PHQ-9) for screening to detect major depression. The researchers found that the PHQ-9 questionnaire was more accurate than other screening tools.
The detail of this paper may be useful for data analysts, statisticians, and associated technicians, and in a research setting where clear cut-offs are needed for validation purposes. As a patient whose last encounter with the PHQ-9 was in the context of being screened for eligibility to participate in a clinical trial, I would say that this tool might still benefit from end-user co-production. For example, patients such as myself can feel marginalised by being rejected by an externally derived scale. It is rather off-putting when you offer yourself and your time to a clinical trial in order to help further medical research, and hopefully, in the long run, benefit other patients, only to be told you are ineligible, because “your score is too low”. Perhaps in these circumstances researchers may communicate to prospective participants that their score simply didn’t fall within the range pertinent to the research study.
In the more serious and clinically important context of having depression and needing care and treatment, many patients will be acutely aware of their levels of depression. They will not want access to the support that they need to be more dependent on a score derived from a screening tool, than on their own lived experience of distress and difficulty.
I recognise that the paper’s findings, which show that there is evidence to indicate the PHQ-9 is a more sensitive measure than previously thought, are positive and highly relevant to the deployment and value of PHQ-9 in research, and perhaps as a supportive tool in the clinical context. However for me, as an individual, it is preferable and more effective when my clinician listens when I express my thoughts, feelings, and impressions regarding my depressive symptoms, rather than delegate the diagnostic evaluation solely to an inflexible measure.
I can accept being excluded from a clinical trial because my PHQ-9 score is “too low”, i.e. less than the cut-off, as this doesn’t affect my quality of life or “wellness.” I would not find it helpful to be denied treatment, just because a score given by a screening tool failed to fall within a certain value range.
However, I write this as someone who has lived with depressive symptoms for many years and over that period of time I have learned how to gauge and manage this. I recognise when I need to seek help. For others who may be relatively new to the experience of depression, an evidence-based and sensitive tool can be an asset. So, this paper’s findings are of importance and could add value when augmented by clinician expertise and when combined with an informed shared decision process between the clinician and the patient.
The researchers say that patients and the public were not involved in this study. This is a shame, for although I imagine that for a non-specialist trying to understand the methodological intricacies of the study might persuade them against engagement and involvement, it is patients and the public on whom such screening tools as the PHQ-9 are used. Their views as to the relevance of the study and its possible outcomes are pertinent. Perhaps even more importantly, the lived experience of being screened using the PHQ-9 might also have illuminated more fully the significance of the various methodological choices made by the authors, and thereby served to inform the design and implementation of the study.
Sarah Markham is an academic mathematician and patient representative currently pursuing a second PhD in theoretical computer science. She is a member of the BMJ Patient Advisory Panel. Twitter: @DrSMarkham
Competing interests: None declared.