What difference does it make to know about psychosocial risk factors?

The “psychosocial yellow flags” or risk factors for developing ongoing disability after a bout of acute low back pain have been promulgated in New Zealand since 1997. Introduced as part of the Acute Low Back Pain Guidelines, the yellow flags were lauded both locally and internationally and subsequently there have been many international guidelines which have adopted this kind of integration. But what exactly do we do with that information? How does it help if we find out that someone is really afraid their pain means something awful, or if they fear their life will never be the same again, or if they truly worry about doing movements that provoke their pain?

Truth to tell, although there have been a lot of studies examining the relevance of psychosocial risk factors, the uptake among clinicians has been fairly abysmal. This is particularly so among clinicians who work either mainly with acute musculoskeletal pain, or amongst those who are mainly involved in treated the body. One physio I know said she got the impression during her training that psychosocial factors “are the things we can blame when our treatments don’t work”.

I think part of the problem is the focus on assessment “technology”. There is a proliferation of questionnaires that can be used to help spot the person who’s likely to have difficulty recovering. We have STartBack, Orebro Musculoskeletal Questionnaire, Pain Catastrophising Scale, Tampa Scale of Kinesiophobia, Depression Anxiety Stress Scale, Pain Self Efficacy Scale – most of which are known by their abbreviations, so it’s like an alphabet soup! But despite knowing about these questionnaires, and perhaps even administering them to people we think might need assessment, once the results come in it’s pretty difficult to know what to do next.

So what if a person reports really high levels of catastrophising? Or that they’re very high on the Fear of Injury/Reinjury on the TSK? Or that they have the lowest ever score on the PSEQ? What on earth do you do to make an impact?

Some people are very actively engaged in “Pain Education”. It’s given to absolutely everyone because “the evidence says” it “works”. Pain reduces. People get engaged in their exercise. Life returns to normal.

Some people refer immediately on to a psychologist. Let them deal with the “difficult” patients.

Others just carry on as normal but in the back of their mind have the “out” that “Oh but they have yellow flags” – and drop their expectations accordingly.

To me that’s just not good enough, and it suggests to me that we need to learn more about what these measures mean – and what to do differently as a result.

There’s a couple of problems though:

  1. How do we choose who to give a questionnaire to? – do we rely on “intuition” or do we give them out to everyone?
  2. Which questionnaires do we use? There’s no “gold standard” – just a mix of various measures that tap into part of the picture…
  3. How much do we rely on strong RCT’s examining whole treatment packages, versus how much do we rely on principles of behaviour change and knowledge of the underlying theories relevant to pain and disability?
  4. What if our clinical reasoning models are completely silent on the work involved in supporting people who present with these risk factors – what if our clinical reasoning models suggest that this work is not all that important compared with the “real” work of tissues and muscles and movement?

Here are my thoughts on what we can do.

I think we should give screening questionnaires to everyone who comes in with an acute bout of musculoskeletal pain, and I think there are a couple that really work well – Orebro is clearly one of them, PCS or PASS are both useful, and I think it’s helpful to screen for mood problems. Why do I think everyone needs these? Well, it’s easier to give them to everyone than to rely on our terribly inaccurate intuition. The risk of failing to identify someone who needs more support is high (and the consequences of omitting this is serious). By routinely administering screening measures we can de-stigmatise the process (though there shouldn’t be any stigma associated with understanding that pain involves the whole person!). We can make the administration easy by integrating it within routine clinic entry process – and by using electronic forms of each questionnaire we can make entering and scoring them easy.

We then need to learn what the questionnaires actually measure – not just the total score, but the subscales as well. Then we need to use those subscale scores to understand what we need to ask the person when we see them face-to-face. This helps us begin to understand the person and how they came to develop these beliefs and attitudes, and in doing so we can develop greater empathy for their experience – and alter our treatments to reflect their needs.

For management, I think we have to, at this stage, step beyond the RCT for evidence. There’s a few reasons for this: one is that RCTs naturally omit individual responses to the treatment meaning we lost the detail as to who responds to which aspect of the treatment. Another is that RCTs often group patients together to ensure power is reached – but in doing this, omit important individual differences. And finally, each person we see is a unique individual with a unique interaction between the various factors influencing their presentation – and there are simply insufficient RCTs to account for these differences. Does this mean we stop using evidence? Oh no!! It just means we need to look at the principles behind many treatments – what are the guiding principles and why might they apply to this person at this time? Finally we need to monitor outcomes so we can establish whether our approach actually helped.

Finally, I think our clinical reasoning models need to include important aspects of treatment that we vary, often without being aware that we do.

For example, if we see someone who scores very high on the PCS and tends to ruminate or brood on the negative, we can’t go ahead and give that person the same set of exercises or activities we’d give someone who is quite confident. We’ll need to lower the physical demands, give really good explanations, take the time to explain and de-threaten various sensations the person may experience, we’ll probably need to move slowly through the progressions, and we’ll definitely need to take time to debrief and track progress.

These “invisible” aspects of treatment are, I think, often the most important parts – but they’re often not mentioned in clinical protocols, and perhaps our skill in titrating the challenges we give our patients is not well developed. These factors incorporate psychological techniques of behaviour change – things like reinforcement, motivational interviewing, problem solving, Socratic questioning, how to fade support, how to bolster confidence, how to vary the environment, and how to avoid pliance and tracking (or going along with things rather than truly integrating the learning). If we want to work with people and help them change their lives, we need to learn how people change behaviour. That means, I’m afraid, learning some psychology…!


  1. Bronnie,

    This is a question I’ve asked for a long time. Having used a handful of questionnaires in the clinic I sometimes wonder what does it matter. What evidence do we have saying it really matters. So we understand an individual is at high risk for chronicity, fearful, depressed, anxious,etc. Is there any evidence that shows addressing these factors actually makes a difference?

    1. Yes there is good evidence showing that addressing catastrophising, depression, sleep and avoidance can reduce disability – even though it may not do a lot for pain intensity.

      I’ll be discussing more practical strategies in subsequent posts, but if you search within my blog you’ll find many posts discussing studies showing that it’s definitely worth assessing and then actively managing these modifiable factors.

      What’s more challenging is developing ways to integrate those factors into physiotherapy/chiro/osteo/massage therapy practice because as I’ve pointed out, most clinical reasoning models don’t incorporate the psychosocial/behavioural elements.

      I’m conjecturing here, but within my own practice I draw extensively on cognitive and behavioural learning paradigms, with often quite subtle changes to how I go about modifying these factors. I use a clinical case formulation approach – there are a few out there and again if you use the search function in the blog you’ll be able to locate several of the posts I’ve written. I think this kind of individually tailored approach to using the results from psychosocial assessment gives better results than attempting to triage people into high or low risk and treating everyone the same within that group.

      Thanks for taking the time to comment – and keep me honest if I don’t ultimately answer your question!

  2. Bronnie,

    thanks for the reply. I struggle to articulate my thoughts, but here goes.

    It seems to me the evidence suggests physical/occupational/CBT/ACT have little effect on pain severity/intensity for those with chronic pain. We may see improvements for disability. That said, is the improvement enough to have a relevant effect on society in the way of gainful employment, social participation and avoidance of healthcare utilization?

    Just today I had two individuals with chronic low back pain. Both had seen and read every horror story for low back pain. They were able to recite all the “malalignments” of their spine. One has “flexibly persisted” while the other has removed themselves from everything outside of work. For one, the simple message of “hurt is not harm” provided reassurance to continue flexibly persisting.

    For the other, “hurt is not harm” made sense, but the continued pain was simply unacceptable. The recommendation to return back to their personal goal of exercise seemed unacceptable with pain. The idea of a potential flare up is stopping them from attempting it. For #2 there were obvious work and life stressors present, some anxiety history, etc.

    These two cases, which appear so similar yet turn out so different make what we do so interesting and frustrating at the same time….

    1. I wonder what the alternative to nonpharmacological and nonsurgical or procedural approaches are? Given the poor response most people living with chronic pain have to these, and the length of time before people are seen before they’re seen for nonpharma approaches, I’m not surprised there are limited changes to pain intensity, and disability changes involve far more than what’s going on in the tissues (truly biopsychosocial).
      I think it takes time, caring clinicians who’re not afraid to say that there’s limited likely change to pain intensity once pain is persistent (particularly neuropathic and “nociplastic”), the offer of hope that life can be good despite pain, and some strategies to reduce the “cognitive fusion” with the idea that it’s impossible to enjoy life when pain is present – and each individual will also have family and friends, cultural expectations, their own life history, their own perspective on what pain means to them in the context of their life. All of these factors will undoubtedly influence how someone responds to the challenge of persistent pain. We do know that there are some therapeutic approaches that can help – graded exposure to avoided activities irrespective of pain intensity fluctuations has been shown to reduce disability, distress AND pain intensity in single subject designs.
      In the case of your two individuals, I’d be looking at Pain Catastrophising and the Pain Anxiety Symptoms Scale because these might give insight into what the person is most afraid of and help identify some of the cognitive patterns that might be making it hard for the person to shift towards being more able to tolerate pain. The thing is, the injury/mechanism involved is only part of the picture – the most important part is, I think the meaning the person places on his or her pain.

    2. “We may see improvements for disability. That said, is the improvement enough to have a relevant effect on society in the way of gainful employment, social participation and avoidance of healthcare utilization?” The individual doesn’t matter then?

      1. In my experience, people living with pain (and I’m one) want to have what everyone has – opportunities to contribute, to grow, to have balance in life, and to be independent from requirements to attend healthcare unnecessarily. Since much healthcare treatment seeking is driven by distress, and seeing as distress is about feeling unhappy with what’s happening and often results in withdrawing from social engagements and feeling part of society, helping people feel less distressed is an integral part of my approach to managing pain. It’s always dependent on the individual and his or her goals, naturally, but my experience suggests that people who are able to participate in work, enjoy social participation and avoid having to seek healthcare do feel more confident to manage their lives in a way that’s healthy for them.

      2. This was a comment directed at Matt, who seems to be taking a different view to you in your post. He says “We may see improvements for disability. That said, is the improvement enough to have a relevant effect on society in the way of gainful employment, social participation and avoidance of healthcare utilization?”

  3. This was so interesting and really hits a personal note for me. Being chronically ill has caused severe anxiety and stress not just from the actual pain I live with, but also from the constant worries about things getting worse, wondering if I’ll forever be in a physical decline and all sorts of other thoughts that keep me up at night.

    I am always thankful to the doctors who explain to me what I should worry about and what I should not. I am honestly amazed at how much pain I can be in without it being something horrible (not that I want something horrible, but, you know). One thing that helps me the most is when I am clearly (and repeatedly) explained to how exactly my body works and why I feel the way I do. That way later on, when anxiety strikes, I can have some tools to calm it down!

    1. Hi! And thank you for taking the time to comment – I’m so pleased that this post rings true for you. There are some clinicians who are worried about “labelling” someone as having worries and fears about their pain, but to me it’s perfectly normal to be anxious when you don’t know what the pain means. I’ll keep your feedback in mind when I write some more posts on how to handle “psychosocial stuff” (as I put it) in the clinic – please keep coming back and keeping me on the straight and narrow!

  4. I just discovered your website and am thrilled to check out some of the resources you recommend. I work in a SNF where the average age is younger then most facilities. Many of my patients have multiple chronic health issues and I will check out some of the pain scales you have listed. Thank you!

  5. Hi Bronnie,
    I completely agree with your sentiments. We know from decades of research that the premorbid nature of people presenting with chronic pain varies widely, and that this personality/history/experience will impact on: the development of the pain, symptom presentation, the persons engagement in treatment, and outcomes. One size will not fit all in treatment, and professionals in this field will be well served to develop an understanding of this fact, and the skills (or team) to support this diversity of people, in practice.
    One minor edit to your post: I believe the DASS is Depression Anxiety Stress Scales… is this what you had intended?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.