… a little more about Pain Catastrophising subscales

I’ve been writing about the Pain Catastrophising Scale and how to use this instrument in clinical practice these last two posts here and here because the construct of catastrophising (thinking the worst) has become one of the most useful to help identify people who may have more distress and disability when dealing with pain. Today I want to continue with this discussion, but looking this time at a large new study where the subscales magnification, rumination and hopelessness have been examined separately to understand their individual impact on pain severity and disability.

Craner, Gilliam and Sperry looked at the results of 844 patients with chronic pain prior to taking part in a group programme (a heterogeous sample, rather than a single diagnosis, so this group probably look at lot like those admitted to high intensity tertiary chronic pain management services such as Burwood Pain Management Centre here in Christchurch).  Most of the participants were female, European/white and married, and had chronic pain for an average of 10.7 years. Just over half were using opioid medication to manage their pain.

Along with the PCS, participants also completed some very common measures of disability (Westhaven-Yale Multidimensional Pain Inventory – MPI) and quality of life (SF-36), and the CES-D which is a measure of depression.

Now here comes some statistical analysis: multiple hierarchical regression! Age, sex, duration of pain and use of opioids were entered into the equation and found to account for only 2.0% variance of the pain severity subscale of the MPI – but once the PCS was added in (subscales entered separately) an additional 14% of the variance was accounted for, but the helplessness subscale was the only one to contribute significantly to the overall variance.

When Pain Interference was  entered as the dependent variable, all the same demographic variables as above contributed a meagre 1.2% of the variance, but when the Pain Severity subscale scores were added, 25.5% of the variance was explained – while the combined PCS subscales contributed 6.5% of the variance. Again, helplessness was the only subscale to contribute to Pain Interference.

Moving to quality of life – the physical subscale of the SF-36 was used as the dependent variable, and once again the demographic variables accounted for only 1.5% variance in physical QOL, with Pain Severity accounting for 23%. PCS subscales contributed only 2.6% of the variance, with only the magnification subscale being identified as a unique contributor. When the mental health subscale was used, again demographics only accounted for 1.2% of variance, with pain severity accounting for 12.4% of the variance. This time, however, the PCS subscales contributed 19.5% of the variance with both Magnification and Helplessness contributing to the variance.

Finally, examining depression, demographics contributed a small amount of variance (3.3%), with pain severity additing 9.8% of variance. The PCS subscales were then entered and contributed a total of 21% to the prediction of depression with both Magnification and Helplessness contributing to the overall depression variance.

The so what factor

What does this actually mean in clinical practice? Well first of all this is a large group of patients, so we can draw some conclusions from the calculations – but we need to be a little cautious because these participants are a group who have managed to get into a tertiary pain management facility. They’re also a group with a large percentage using opioids, and they were pretty much all European – and from North America, not New Zealand. I’m not sure they look like the people who might commonly come into a community-based facility, or one where they’d be referred directly from a GP or primary care centre.

At the same time, while this group may not look like the people most commonly seen for pain management, they share some similar characteristics – they tend to magnify the “awfulness” of pain, and then feel helpless when their pain is bothering them. Surprisingly, I thought, ruminating or brooding on pain wasn’t a unique contributor and instead the helplessness scale contributed the most to pain severity, pain-related interference (disability associated with pain), poor mental health quality of life, and low mood, while magnification scale contributed to poorer physical health quality of life, mental health quality of life and low mood.

What this means for practice

The authors suggest that the construct measured by the helplessness subscale might be a factor underlying poor adaptation to life’s difficulties in general, leading to passivity and negative emotions. They also suggest that magnification might be a unique contributor to perceiving obstacles to doing the things we need to do every day, while hopelessness might mean people are less likely to participate in enjoyable activities and then in turn contribute to feeling low.

Importantly, the authors state: “We offer that simply collapsing the 3 dimensions of this phenomenon (ie, rumination, magnification, helplessness) may actually conceal nuanced relationships between specific dimensions of catastrophizing and outcomes that would might inform treatment approaches.” Looking at the overall scores without thinking about the subscales is going to give you less information to use for individualising your treatment.

In a clinical setting I’d be reviewing the individual subscales of the PCS alongside both disability and mood measures to see if the suggested relationships exist in the scores this person has given.

I’d be taking a look at the repertoire of coping strategies the person can identify – and more, I’d be looking at how flexibly they apply these strategies. Extending the range of strategies a person can use, and problem-solving ways to use these strategies in different activities and contexts is an important part of therapy, particularly occupational therapy and physiotherapy. Another approach you might consider is helping people return to enjoyable activities that are within their tolerance right here, right now. By building confidence that it’s possible to return to things that are fun we might counter the effects of helplessness, and help put pain back where it belongs – an experience that we can choose to respond to, or not.

I’d also be taking a look at their tendency to avoid feeling what their pain feels like, in other words I’d like to see if the person can mindfully and without judging, complete a body scan that includes the areas that are painful. This approach is intended to help people notice that alongside the painful areas are other nonpainful ones, and that they can successfully be with their pain and make room for their pain rather than attempting to block it out, or over-attend to it. The way mindfulness might work is by allowing people to experience the sensations without the judgement that the experience is bad, or indicates some terrible catastrophe. It allows people to step back from the immediate reaction “OMG that’s BAD” and to instead take time to view it as it actually is, without the emotional halo around it.

Pain catastrophising is a useful construct – but I think we need to become more nuanced in how we use the scores from the questionnaire.

Craner, J. R., Gilliam, W. P., & Sperry, J. A. (2016). Rumination, magnification, and helplessness: How do different aspects of pain catastrophizing relate to pain severity and functioning? Clinical Journal of Pain, 32(12), 1028-1035.


  1. Hi Bronnie,
    How do you actually identify which subscale the person is highest on – Rumination has 4 items vs helplessness 6 items – so is it sufficient to just sum them? I do understand that looking at individual responses to individual items is important but just wondering if I have missed something with the subscale calculations?

    1. The full scoring instructions are found on the PCS manual available from Prof Mick Sullivan’s website – http://sullivan-painresearch.mcgill.ca/pcs1.php
      If you download the manual you’ll find the calculations for subscales and also at the end of the manual in Appendix 1 is a table giving percentiles from the original study. If your population looks like the ones included in the original study you can use those scores, or over time you can collect your own and calculate your own norms (I think this is the better long-term strategy).
      You can check the items with your client, but I always score the questionnaire and check percentiles as well. I then use that when interviewing the person – for example, I’ll say “It looks like you might have some trouble getting your mind off your pain at times, can you tell me about that?” and let them confirm or disconfirm from their experience. Rather than going through each item with the person I suggest paraphrasing the extreme scores (either very high or very low) and checking in with the person to establish their thoughts. Questionnaire responses are a snapshot in time, and can’t be treated like x-rays of the mind because people do misinterpret the questions, or may be having a particularly good or bad day, so it’s always worth checking this with them if the questionnaire doesn’t seem to “fit” with how they’re presenting. Hope this helps!

  2. Bronnie, I cannot help but feel that the Questionnaire does not adequately consider the relative magnitude of the other side of the equation, i.e. the various major life stressors (no job, no money, no home, no family etc) that people in pain commonly experience. Please correct me if I am wrong about this.

    1. John, I agree with you – but it’s not intended to be used as a single instrument in isolation. There may well be good reasons for the person scoring high on any of the sub-scales, and any good clinician would explore these dimensions within an interview (as I’ve explained in the comment above). What we do know is that if a person has multiple life stressors they may be more likely to interpret their pain in a catastrophic way – and that by addressing those other stressors, the person’s scores on this questionnaire may well change. I can’t think of any questionnaire that can adequately represent the impact of all the multiple dimensions of our experience of pain, and neither should they. They’re simple a convenient way of measuring a construct (or more than one) in a consistent way to help researchers, clinicians and the person completing the questionnaire understand part of “what it is like to be this person experiencing this pain”. Used sensitively, used with due regard to the other dimensions and the limitations of attempting to measure abstract constructs, they can be useful. If they’re not useful, not adding to our understanding, not helping measure change – then they often fail to be used. Just like the old MMPI (not the WHYMPI – but the https://en.wikipedia.org/wiki/Minnesota_Multiphasic_Personality_Inventory) which used to be used frequently by psychologists has eventually faded out of common use in pain psychology, measures like the PCS have to keep earning their way into practice, or they’ll also eventually be abandoned.
      BTW I’d hope that no person ever attempts to use a measure like this, or a measure of mood or anxiety, as a “diagnostic” tool on their own – they’re not developed for that kind of purpose, but act as another source of information to be pulled together along with taking a good history, careful clinical observation, other clinical testing and most importantly, a conversation with the person experiencing pain.

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.