Is one question enough to screen for depression and anxiety

Psychological factors, such as depression and anxiety, are significant contributors in the transition from acute to chronic pain. A person suffering from acute low back pain with additional symptoms of anxiety and/or depression has a higher risk of becoming a chronic pain patient than a person without these additional symptoms. Furthermore, suffering from low back pain for a long time can, in and of itself, result in depression and anxiety, which again worsens the prognosis for the patient if these symptoms are not handled adequately. It is therefore important to identify easy and efficient ways to identify pain patients who have depression and anxiety. Several screening measures exist, but most of them are too long and time-consuming to be used routinely in clinical care. There is a need for shorter instruments. We investigated if two single-item questions would be enough to screen for anxiety and depression in a large group of patients with chronic low back pain.

We studied 564 patients who were on sick leave due to low back pain.[1] We compared a single-item screening question with two longer screening measures for anxiety and depression. As the “gold standard” we used a clinical diagnostic interview (The Mini-International Neuropsychiatric Interview, MINI). According to the diagnostic interview, 4% of the patients had a depressive disorder while 12% had an anxiety disorder. The results showed that the single-item screening questions were equal to or better than the two longer questionnaires in identifying depression and anxiety. The results were particularly good for depression, showing a sensitivity of 95%. That means that almost all of the patients who suffered from depression were identified with this single question. The sensitivity for anxiety was lower, at 68%, but that was still equal to or better than the longer questionnaires. The specificity of the screening questions – that is, the proportion of patients without depression/anxiety that the screening questions correctly identified – was 56% for depression and 85% for anxiety.

The findings of this study could be very helpful in clinical settings where short screening measures are needed. In most clinical settings, it should be possible to ask one question, even though time is limited. Because the specificity of the questions means that there will be some false positives – patients may screen positive without being clinically depressed/anxious – there will be a need for further evaluation of those who screen positive. We therefore suggest a 2-step process, where the single-item screening questions are used to identify potential individuals with depression/anxiety, and where those who screen positive then undergo a more detailed test with greater specificity to confirm the problem and decide on the need for treatment.

The single-item screening questions could also be very useful in research where they could replace longer screening measures to identify potentially depressed patients, particularly in cases where questionnaire length is a concern. Since both questions performed equal to or better than the longer questionnaires, they could, in fact, replace these questionnaires in epidemiological studies where time is limited.

However, before the two questions can be recommended and implemented in clinical settings, these results must be replicated in other populations. Additionally, more work is needed to improve the sensitivity of the anxiety question to obtain a sensitivity that matches that of the depression question. Nevertheless, the findings do represent a promising step towards the use of ultra-short screening instruments in clinical settings for pain patients, and that could lead to proper treatment for those patients who suffer from depression or anxiety in addition to their pain. Such treatments might involve cognitive–behavioral strategies to address unhelpful pain beliefs, strengthen coping resources, and provide instruction in pain self-management.

About Silje Reme

Silje RemeA native of Norway, Dr. Reme completed her educational studies at the University of Bergen, earning her doctorate in psychology as well as her clinical psychology and undergraduate degrees. She did a 2-year postdoctoral fellowship at Harvard School of Public Health, where she worked with the Harvard Center for Work, Health and Wellbeing. She now works as a senior researcher at Uni Research in Bergen, Norway, where she is currently co-heading the research group Stress, Health and Rehabilitation. Additionally, she works as a clinical psychologist and research fellow at The Department of Pain Management and Research at Oslo University Hospital. 


[1] Reme SE, Lie SA, & Eriksen HR (2014). Are 2 questions enough to screen for depression and anxiety in patients with chronic low back pain? Spine, 39 (7) PMID: 24480946


  1. Jonathan says

    Hello Silje,

    Interesting article. Through personal experience, when a patient ask questions about a certain problem that they’re having, the person on the retrieving end should always do the following three: 1) Ask and understand the ailment of the situation. 2) Ask and understand the severity of the ailment and how it’s negatively affecting their quality of life. And 3) Ask and understand to determine what the complexity of it is and the most appropriate solution for the patient. If you’re not asking them what are the 3 things I’m trying to understand so I can help the patient, you will never find the appropriate solution.



  2. Silje Endresen Reme says

    Thanks, Cameron!
    If you give me your email, I’ll send you the references.

  3. Silje Endresen Reme says

    Hi Mel,
    Thanks for your comment!
    Most instruments that aim to capture depression and anxiety actually use a shorter timeframe, most often 2 weeks. The reason we chose last month was to try to be as inclusive of all probable cases as possible. Of course that does lead to a high rate of false positives, as many of us would answer “yes” to that question without being clinically depressed or anxious. Still, this is preferred when developing a good screening tool, as you would want to be as inclusive as possible (i.e. sensitivity close to 100%).

    In terms of psychological factors predicting chronic pain, we did not look at that in our study (which was only cross-sectional), but other studies have addressed that issue more thoroughly.

  4. John Quintner says

    Marcel, I agree with you, even when we are observing ourselves.

  5. John, of course it is a classification, I believe it is a helpful one.
    And doesn’t everything we observe and use words for to give it thought can be seen as judgements imposed by an observer

  6. John Quintner says

    Mel, as far as I can tell, none of this is proven. Let us not pretend otherwise. We need to be very careful to avoid conceptual confusion. I have fallen into this trap many times!

  7. Mel Colgar says

    John, it is proven with studies that chronic pain sufferers have a different cortex mapping for pain. it is also well accepted that psychological factors do contribute chronic pain development.
    You are right, obviously, it is too hard to even study the transformation of psychological factors turning into chronic pain but it is also not easy to claim the otherwise.

  8. John Quintner says

    Mel, I wonder what you mean when you say that “psychological factors turn into chronic pain”? Can you please enlighten me as to how such a transformation might take place?

  9. John Quintner says

    Silje, all that might be required is that we look at this problem from a different viewpoint. I fail to see how ongoing research can solve it.

    In the words of Ludwig Wittgenstein: “The existence of the experimental method makes us think we have the means of solving the problems which trouble us; though problem and method pass each other by.” [Wittgenstein 2001: Part II, sec xiv, p.197]

  10. Silje Endresen Reme says

    John: I agree – your suggestion of a third possibility appears very plausible, and we can only hope that all the ongoing research will bring us closer to a better understanding of this important issue.

  11. Thanks Silje.

    So simple – I love it. “Have you felt sad or anxious lately?” Great example of how “thin slicing” can be used to keep it simple and save time.

    I’d be interested in the reference list relating to the cause/effect question if you could email please.

  12. Mel Colgar says

    Hi John, Silje and everyone
    it is lovely to see that these concepts are being discussed so openly and deeply.
    I think that research and clinical practice has its own big problems when it comes to interrelation and I am only a clinician not a researcher what I say is likely to be biased.
    About the discussion above, the phrase “in the last month” can be questionable. I believe psychological factors do take time to turn into chronic pain and do the suggested changes in the cortex. One month is very likely to be too short for this to happen.

    Silje was there a specific reason for you to choose the time period because as John pointed out, I could, like many others out there, easily answer “yes” to the question?

  13. John Quintner says

    Silje, thank you for your thoughtful response. But have you considered that there might be a third possibility?

    Could the mood disturbances that we call “anxiety” and “depression” be inextricably bound to that which we call “pain”?

  14. Silje Endresen Reme says

    Dear Marcel and John,
    I agree that the distinction between acute and chronic pain is somewhat nonsensical, especially since the pain in most cases show a cyclical pattern. Perhaps it is time for a revision of this terminology?

    When it comes to the role of psychological factors, however, it has now been fairly well documented that these factors involve an increased risk for chronic pain.

    Still, the nature of the relationship between psychological factors and chronic pain (i.e. cause/effect) is not yet fully understood. Studies have shown that some develop psychological disorders, particularly depression, as a consequence of suffering from pain for a long time, while other studies have shown that psychological disorders in some cases precede the pain. It therefore appears to be empirical support in favor of both a cause and effect relationship between psychological factors and pain.

    I would be happy to send you some references if you are interested – just let me know.

  15. Silje Endresen Reme says

    Hi Marcel,
    In some clinical settings, asking a few more questions would certainly be possible (and preferable), but in other settings (e.g. GP settings where the average time per patient is 4-5 minutes), ultra short screening tools, such as this one, would be the only realistic option to implement.

  16. Silje Endresen Reme says

    Hi Stuart,
    There are definitely some overlap between the two questions we used from the SHC Inventory and the PHQ-4, except for the obvious number or items and the timeline (PHQ asks for symptoms the last 2 weeks while we asked over the last month). In our paper we actually suggest that a comparison between our 2 questions and the PHQ should be an avenue for further research, given that the PHQ has shown good promise in identifying depression an anxiety in different medical populations.

    Also, if you want to read the full paper, it is actually available as open access:

  17. John Quintner says

    Marcel, surely the descriptors “adaptive” and “maladaptive” are no more than value judgements imposed by an observer on the lived experience of another human being.

    I agree with you that the concept of there being a transition from acute to chronic pain is nonsensical. To then incriminate psychological factors as being important factors in such a transition seems to be far fetched. Silje, could there be some confusion here between cause and effect?

    In all honesty, I would answer positively to both questions but I do not experience pain (yet).

  18. Is it that difficult to ask a few questions?

    I gave the idea “Transition from acute to chronic pain” proposed earlier on the site , quite some thought.
    I only can conclude there is No transition only than in the mind of the beholder not in the sufferer. Why, it is mere classification in terms of time. It might be a better idea to use terms as adaptive and maladaptive pain suggested by the late L.Gifford.

    Like someone asked me on the other thread, ” does acute pain feels different than chronic pain, for the sufferer?”

  19. Silje Endresen Reme says

    Hi Cameron,
    The questions were taken from the Subjective Health Complaints Inventory and reads “have you been affected by sadness/depression during the last month” and “have you been affected by anxiety during the last month”

  20. stuart miller says

    Hi, how does this compare to the PHQ-4 ? I realize there are 4 questions (and it doesn’t target suicidal ideation and bipolarity) but it is quick (just like PHQ-2 and GAD-2 it is derived from). I wasn’t able to open the full article either so I am curious. Thanks for this important work…

  21. Thanks for sharing this information! What were the questions used?

  22. Hi Silje, What were the questions?