The Pain Science Division of CPA

The last class of the new pain elective course that I designed and delivered for the first time this year, happened today.  I saved one of my favourite topics for last, and that is pain assessment.  We spent a considerable amount of time discussing the numeric rating scale, and more importantly, the anchors we should assign to the ends of the scale, and what they might be telling us.  We have much discussion over this, as I have had with others in the past, and some really neat ideas came up from the MPT students in the class.  I want to share one in particular because I think it's very insightful.  The student suggested that zero should represent where the patient would like to be, or the ideal state (in terms of pain intensity in this case), and 10 represents something like extremely far away from ideal, or as far away from your preferred state as you can get.  I thought this was a very interesting take (I don't like the traditional 'worst pain ever' or 'worst pain imagineable'), and I plan to give this a try with some of my patients later on this week.

First, much to my dismay, the deadline for abstract submissions to the IASP congress in Glasgow this year has come and gone, without anything submitted from me.  If the deadline were a month or two from now, I would have had a few things to submit, but the timing was just off.  Since my research funding will only cover me if I'm going to a conference to present, I will not be able to afford to go this year :-(  The program looks absolutely brilliant, and I encourage those of you with the means to attend.  Another option would be to attend the pain, mind and movement satellite symposium the following week in Dublin, should be another brilliant conference.  So many conferences, so little time.

Speaking of conferences, I did have the opportunity to attend and present some work at the World Congress on Neck Pain in LA two weeks ago.  The biggest news that came out of that conference, besides my being interviewed and quoted in a story by WebMD (neato!), was the release of the BJD Neck Pain Task Force work.  This group released several documents, which have been published in this month's supplementary issue of spine.  I will not dwell on all of them, but I will present a few musings around the proposed new model for classification of neck pain, which I will say right now is by and large an excellent leap forward for the field.

My thoughts (written while at the conference):

Overall, I find the model to be an encouraging improvement over the previous symptom-based classification system of the QTF.  The recognition of neck pain as a condition with a multifactorial etiology will, hopefully, influence policy-makers when it comes to the acceptance of neck pain as a distinct entity, regardless of the presence or absence of structural pathology.  I appreciate the challenge in constructing a model that is comprehensive yet parsimonious, and I applaud the effort, and in large part, I applaud the result also.

The classification of neck pain along a series of 5 axes will help systematic reviewers in identifying homogenous subsets of the population.  However, this will only be helpful if researchers in the field adopt this classification scheme, which will be dependent on the forthcoming dissemination efforts of the task force, and I wish them luck in this endeavour.

Below are some of my immediate reactions in terms of limitations to the model.  Some of these are more problematic than others, and I’m sure that yet more limitations will be realized once the model is put into practice.

·         The extrapolation of Beaton et al.’s qualitative work to another population is problematic in constructivist paradigms.  While I am thrilled to see this model incorporated, the problem with generalizing this work should at least be addressed.

·         Non-interfering neck pain as neck pain that a) does not interfere with activity or participation, b) the person does not seek care from a physician and c) the person does not file a claim.

Interfering neck pain as neck pain that a) leads to seeking care b) leads to filing a claim or c) interferes with activity or participation.

On the surface, this is problematic because it assumes that the reason a person will seek care is due to interference in activity or participation.  However, it is entirely possible that a patient with neck pain, who is able to perform necessary activities, but might be fearful of possible future consequences of the neck pain will seek care.  This person does not appear to fall into either classification.  However, I will admit that on deeper reading, the inclusion of the word ‘or’ in the criteria for interfering neck pain, whether intentional or not, will account for the person who visits a health care provider but does not report interference with activity and participation.  The use of the word ‘interfering’, then, is somewhat misleading as it suggests that the primary criteria for being included in this classification is interference.  Perhaps a better term would have been ‘bothersome’ or ‘distressing’ or something of that nature.  To its credit,  It appears that the Task Force recognizes this limitation (pg. 7 paragraph 2).

·         Grade III neck pain assumes that decreased DTRs, weakness and sensory disturbance are in fact diagnostic of neurologic compromise.  This literature is extremely sparse on this, I’m aware of Wainner (2003) that found adequate diagnostic accuracy when testing deltoid and biceps brachii, and biceps DTR.  Rainville (2007) found that weakness of forearm pronation appeared to be reasonably diagnostic of C6 or C7 radiculopathy, but the statistical presentation of the findings precludes calculation of Sn/Sp, and the lack of a control group diminishes the internal validity of this study.  Beyond that, I am unaware of well-controlled studies of the diagnostic accuracy for clinical signs of radiculopathy.

·         It is unclear as to how ‘interference’ should be assessed.  My opinion is that the optimal method would be patient self-report based on their individual statements about the effect of the pain on their lives, but I know that it will only be a matter of time before the insurance industry, or epidemiologists, start to use things like scores on the NDI to determine whether the patient is experiencing ‘interfering’ neck pain or not, which will subsequently, necessarily, lead to patients exaggerating reports of interference in order to receive the care they feel they deserve (and in most cases justifiably so).  I feel like the task force missed an important opportunity to drive the assessment of interference here, and I fear what will happen once the insurance industry takes hold of this.

·         Finally, there is still the problem of defining and operationalizing ‘chronic pain and disability’.  In my meta-analysis, out of 11 cohorts there were 13 different criteria applied for determining the presence or absence of chronic pain and/or disability.  Even this weekend, there were a number of different methods used for determining whether or not the patient group had ‘recovered’, including the functional rating index, the neck disability index, and an NRS score of less than 5, which was the most problematic in my view.  I would have liked to see the task force recommend a standardized operational definition of the problem, whether we choose to call it chronic pain, late whiplash syndrome, persistent WAD , or anything similar.  This oversight will mean that systematic review and meta-analysis continues to be hampered by heterogenous outcomes.

I look forward to seeing how the future unfolds for the findings of the BJD Neck Pain Task Force.

There are three widely used self-report tools for post-traumatic stress. The most widely used, and oldest one, is the Impact of Events Scale, designed by Hurwitz (1979). It is a 15-item self-report scale that has been show to possess two factors – avoidance and intrusion. The development of this scale pre-dates the DSM-III entry for PTSD, which conceptualized PTSD as a tripartite condition, involving avoidance, intrusion and physiological hyperarousal. Perhaps for this reason, the IES has only moderate discriminative ability when it comes to diagnosing PTSD. The IES-Revised (Weiss and Marmar 1997) was developed to provide a better match between the tripartite DSM criteria and the scale. The IES-R saw the addition of 6 more items to the scale, pertaining to physiological hyperarousal. In a sample of 182 people seeking treatment post-MVA, the IES-R had Sn 74.5 and Sp 63.1, when compared against the Clinician-Administered PTSD Scale (CAPS, Blake et al. 1990) as the gold standard, indicating it is best used as a screening tool rather than a diagnostic tool (Beck et al. 2007). Using the original IES, Coffey et al. (2006) found Sn .91 and Sp .72 in a sample of 229 MVA patients, with overall accuracy of 80%.

A second widely used scale is the PTSD Symptom Scale – Self Report (PSS-SR, Foa et al. 1993). The PSS-SR contains 17 items reflecting the diagnostic criteria of the DSM-IV for PTSD, scored on a Likert scale of 0-3. In what appears to be the only test of the diagnostic validity of the PSS-SR in an MVA population, Coffey et al. found Sn to be good at .91, but a very modest Sp of .62, when compared against the CAPS as gold standard. This suggests that those who have PTSD are likely to score over threshold (cut-off of 14 in this case), but there will be a high number of people without formal PTSD that will be labeled positive by the tool.

The third tool is the PTSD Checklist (PCL, Weathers et al. 1993), a 17-item self report measure that, like the PSS-SR, contains an item pertaining to each of the 17 diagnostic criteria for PTSD from the DSM-IV. Each item is scored on a Likert scale ranging from 1 (not at all) to 5 (Extremely), meaning scores can range from 17 to 85. Using a mixed sample of 27 MVA patients and 13 victims of sexual assault, and again using the CAPS as gold standard, Blanchard et al. (1996) demonstrated Sn 94.4 and Sp 86.4, with overall accuracy of 90% when the cut-off score was set at 44. Ruggiero et al. (2003), in a sample of 392 college-student participants, 63.6% of which reported a transportation-related accident as their key trauma, found Sn of .90 and Sp of .95, accuracy 95%, with a cut-off of 44, although the gold standard in this case was not through direct assessment by a professional but through correlation with a number of other scales and factors, that makes comparison between these findings and other findings difficult to perform. An attractive component of the PCL and the PSS-SR is that the interpreter of the scale can determine not only an overall score, but can determine whether or not the responses in fact reflect a diagnosis of PTSD as set out in the DSM-IV, which require the endorsement of at least one ‘intrusion’ symptom, 3 ‘avoidance/numbing’ symptoms, and 2 ‘hyperarousal’ symptoms. It could be argued that the appropriate endorsement of symptoms, rather than the total overall score, is more important to the diagnosis of PTSD (Ruggiero et al. 2003). Unfortunately, it is as yet unknown what the cut-off score of each individual item should be before one determines the symptom to be present at a threshold adequate enough to be considered in the diagnosis. Blanchard et al. (1996) evaluated individual items, and found that while most items maximize their diagnostic efficiency with a cut-off score of 3 out of 5, items 1,2,9,10,12 and 15 demonstrate better efficiency when the cut-off for ‘present’ is set at 4 out of 5. The IES allows no such procedures to be performed.

Based on the work done to date, the PCL-civilian version seems to be the most accurate self-report scale available for diagnosing PTSD in MVA patients, according to DSM-IV criteria. At 17 items in length, and approximately 5 minutes to complete, it is on par with the PSS-SR and the IES, and slightly shorter than the 22-item IES-R in terms of respondent burden. Assessor burden can be low if a total score is all that is desired, but the tool does allow for more in depth evaluation of the responses that should improve diagnostic accuracy further.

Just had the last class in the qualitative research methodology course that we are required to take in the PT PhD program here at Western.   It was an interesting course, but a ton of work and I'm glad to have it behind me.  The piece I included in the newsletter on pain and ontology came from what I had learned in that course.

As many of you will become aware, the CPPSG has become the steering committee for the Pain Sciences Division, as we set out to, once again, apply for official division status within the CPA.  We've already received some feedback and questions regarding our motives behind the decision, so I'll elaborate here quickly.

The moratorium on new division formation has been lifted, albeit with a new process in place for obtaining division status.  This was impetus number one.  Once we heard about this, the executive of the CPPSG quickly convened via teleconference  for a rather lengthy discussion around our future direction.  Fortunately, the vision of all of us on the executive seem to be aligned in the same direction.  While we were initially rather staunch in our desire of autonomy, we decided that, in order to fulfill our vision of getting the most information to the most people, and to raise the visibility of PT in dealing with people in pain, it would be a very wise move, politically and otherwise, to become part of the professional association that is really the public face of physio in Canada.  We wanted to make sure that, especially in the case of people from outside of Canada looking inwards to see what role PTs in Canada play in pain management, those people would see a dedicated group working in the area, and since the first place international observers would look would be the CPA, obtaining division status seems a logical approach to take.

Some of our readers have voiced concern that by becoming a CPA division, we might be forced to abandon some of our other endeavours, such as international collaborations with other pain groups (in the UK and Australia so far).  Being a division of CPA in NO WAY precludes us from pursuing these collaborations, and we hope may even help.  We will also continue to publish our newsletters, and host a website.  Whether the site will move from squarespace is something that will be determined in due time, but it's possible.

Unfortunately, CPA will once again be conducting a poll of its members to identify interest in a pain sciences division.  I say unfortunately because many of you will remember that this was done two years ago during our first application for division status.  The response was very good then, and we hope it will be at least as good if not better this go around.  Please, the steering committee encourages you to respond to this survey when it comes (probably in early January).  A minimum of 100 members expressing interest in joining is what we need for our application to move forward.  An expression of interest is certainly not binding, and the survey is anonymous, but the CPA wants to see that the interest is there.

We are entering yet another phase of our existence, and we hope many of you will come along for the ride.  I am preparing to begin the comprehensive examination part of the PhD program, which is an intensive 3-month process that will see me encapsulated firmly within the world of knowledge translation and glial cells - an interesting combination if I say so myself.  Fortunately, I am blessed to be working alongside such hard workers in the steering committee, and I know we will put forth another strong application, although I may be limited to editorial work.  

All for now. 

I have spent much of this afternoon reviewing the new issue of Physical Therapy Reviews, which is dedicated entirely to pain.  I recommend picking this one up, you can get it through IngentaConnect if you don't have institutional access.  Lormier Moseley, one of those people whose work I deeply respect and always look forward to, starts off the issue with a piece on reconceptualizing pain.  It is, as usual, a brilliant review of the relevant work in the area, but there is one passage that particularly sticks out with me.  It is in my estimation a superb reply to anyone who thinks physio, especially when dealing with pain, is becoming too much like psychology.  Here's the passage:

One implication [of the biopsychosocial model] is that to base clinical reasoning on what is currently known about the biology of pain requires that the skills and knowledge of the clinician are broader than those related to anatomy and biomechanics.  That is, the clinicain must hve a sound knowledge of diagnostic tools, tissue dynamics, healing and remodelling, peripheral and central sensitization, and psychological and social factors that might affect  the implicit perception of threat to body tissues." 

Beautiful, don't you think?