People with pain are not all alike: MPI classifications & outcome

As an individual with chronic pain, I know that each person with chronic pain is different from the next, and as a clinician I know that there are few ways to predict who will benefit from what treatment – but it’s like a Holy Grail to find a way to group together people who will benefit from one type of intervention compared with another.

Please bear in mind as you read the rest of this post, that I’m not suggesting questionnaires ‘sort’ people into defined groups – yet – but as part of developing a picture of what works for whom and when, this study by Verra, Angst, Briochi, Lehmann, Keefe et al., caught my eye.

The MPI (Westhaven Yale Multidimensional Pain Inventory) is a robust and comprehensive questionnaire that has been used both for research and clinically, and helps classify patients into one of three main groups: Adaptive Coper, Interpersonally Distressed and Dysfunctional (Kerns, Turk & Rudy, 1985). Now while the term ‘dysfunctional’ isn’t very politically correct, it aptly describes a group of people who are often highly distressed, depressed, feeling out of control, and have high levels of disability. Those that are interpersonally distressed tend to have more focus on lack of social support, feel isolated and helpless in their pain; while people who are identified as adaptive copers are generally ‘getting on with it’ and have the lowest level of distress and disability.

The usefulness of classifying people into similar groups is not likely to be lost on anyone who has worked with a group of patients with very different needs! Although there is little information so far as to what the experience of being in a group programme is like for participants, my observation is that people can feel excluded and unsupported if they’re in a group treatment programme where their needs are not directly met, or they have unique needs that are not addressed. I think some sort of classification might be helpful. At the very least, it would make sure that the current ‘one size fits all’ approach in group programmes might be tailored slightly because, as we know, people with chronic pain are not a homogenous group at all!

In the study by Verra, Angst, Briochi et al., a group of people with fairly high levels of disability arising from fibromyalgia were given the MPI prior to starting an in-patient pain management programme. At the same time, they were also given several other questionnaires, and these were repeated at the followup of the programme. The other questionnaires included the HADS, SF36, and CSQ. Outcomes were measured at four weeks, or completion of the programme – and most of us would argue that this is far too early for outcomes to be measured (but perhaps this point needs to be made to certain purchasers in NZ?).

As is typical for me, I won’t go into detail regarding the analysis – but cut to the chase:

  • Participants were generally young, had a long history of pain, high physical disability, high prevalence of depression and high scores of depression and anxiety, and low psychosocial functioning.
  • 116 of the 118 patients were able to be classified into the three profiles of the MPI – 24% were interpersonally distressed, 36% dysfunctional, and 38% adaptive copers. The remainder were anomolous and didn’t fit into any profile. On demographics, the only difference between these three groups was pain duration, with people in the dysfunctional category having pain for less time than others.

When the authors did their statistics(!) on the other measures used, they found the differences between each group varied according to the predictions of the theory behind the MPI, confirming its validity for this population.

  • All three groups improved in terms of their pain severity at the four week point. The Dysfunctional group demonstrated the greatest effects in terms of pain severity, interference with activities, and SF36 physical functioning, but made the least improvement in SF36 social functioning, worsening in CSQ increasing activity levels and least improvement in CSQ ability to control pain. The Adaptive Coping group showed most improvement in social support, and CSQ increasing activity level.

OK, so what does this mean?
The results are quite mixed, by comparison with other similar studies carried out in both other countries and with other pain diagnoses. Several studies have suggested that outcomes are not correlated with MPI profiles, while this study demonstrated that there were differences. The reasons for this could be many – different programmes (as usual, the content of this programme is only briefly described, although it was reported to be six days a week, up to six hours a day); different therapists; different timeframe for outcome measurement; maybe different population (this group were inpatients, while other studies have been conducted on outpatients). Whatever the reason, this study seems to show that individualising programmes might make a difference, at least for this group.

Practically, it’s very difficult to implement a pain management programme using profiling. In many clinics, it’s difficult enough to identify patients who fit the usual inclusion criteria without trying to refine selection on the basis of diagnosis, gender, or MPI profile. It has been done – many years ago, Burwood Hospital’s Pain Management Centre used the MPI, and it was found that the majority of people selected for the then Residential Programme were in the Dysfunctional group. Sadly, this approach has fallen by the wayside because, in part, of funder requirements (Get them seen sooner! Who cares if what they’re getting isn’t quite suited to them, just SEE them will ya!!).

As a clinician, I would never rely on a single instrument to classify patients and treat accordingly. I still think there is a need to develop a unique case formulation to describe and explain the causal relationships between various factors influencing the person’s disability. The MPI (and other questionnaires) add to this formulation.

I’m still struggling to find a paper that describes the content of a pain management programme in sufficient detail that it could be replicated in another setting. This really hampers development of a robust treatment for chronic pain. When even researchers fail to clearly describe what they do, is it any wonder that funders think ‘as long as it walks like a duck, talks like a duck – it might as well be one’?

As clinicians, I think we know intuitively that people are not alike, but we often use the same treatments, the same approaches, the same structure even to deliver pain management.  My main argument with a programme that has been introduced to New Zealand directly from Canada is that patients don’t fit a pre-imposed structure, and to pretend that they all have the same needs, the same readiness to do pain management, and the same response to intervention simply flies in the face of experience.  It’s almost akin to believing the old biomedical model still holds sway – if we really do hold to a biopsychosocial model, I’m pretty sure our treatments might also need to reflect the different psycho and social aspects of an individual.

Martin L Verra, Felix Angst, Roberto Brioschi, Susanne Lehmann, Francis J Keefe, J Bart Staal, Rob A de Bie, André Aeschlimann (2009). Does classification of persons with fibromyalgia into Multidimensional Pain Inventory subgroups detect differences in outcome after a standard chronic pain management program? Pain Research & Management : The Journal of the Canadian Pain Society, 14 (6) : 1929024711

The West Haven-Yale Multidimensional Pain Inventory (WHYMPI)
R. D. Kerns, D. C. Turk, T. E. Rudy. Pain. 1985. Vol. 23, Iss. 4; 345-356

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s