Date sent:      	Wed, 13 Nov 2002 

The Repeatability of Pain Site Diagrams

J of Musculoskeletal Pain, Vol. 10(3) 2002, pp. 83-90

Paul D. Triffitt

Paul D. Triffitt, MA, MD, is Senior Lecturer, The University
Department of Orthopaedic Surgery, The Glenfield Hospital, Leicester, UK.

Address correspondence to: Paul D. Triffitt, University Department of
Orthopaedic Surgery, The Glenfield Hospital, Leicester, LE3 9QP, UK.

Submitted: May 2, 2001.
Revision accepted: January 4, 2002.

ABSTRACT.

Objectives: While pain diagrams have been in frequent use for the
assessment of patients presenting with musculoskeletal symptoms, there has
been little study of their repeatability. A diagram of pain site was
accordingly investigated.

Methods: Four hundred and twenty-four patients attending a shoulder clinic
with unilateral symptoms were asked to mark the areas in which pain was
experienced, and 269 who had not received treatment in the intervening
period were asked to complete a second diagram after a delay of
approximately two weeks [median 15 days]. The diagrams were analyzed for
the repeatability of the marking of each pain site and of the overall pain
distributions.

Results: The initial diagrams showed wide variation with 124 different
distributions seen. The kappa values for repeatability for individual pain
sites were all less than 0.65, while the kappa value for repeatability of
the pain distributions was 0.32. The number of marks in contiguous regions,
assessing near agreement, differed significantly in almost all areas.

Conclusions: The analysis suggests that even a simple diagram of pain site
showed only poor to moderate repeatability in this context, and that each
diagram should be tested for repeatability before use in any particular
patient group.


KEYWORDS. Pain, body pain diagram, regional pain


INTRODUCTION

Patient pain diagrams have been studied in a variety of settings, including
in relation to the cause and psychological aspects of low back pain (1-9),
for screening and prognosis of treatment in carpal tunnel syndrome (10-12),
in the assessment of anterior knee pain (13), in patients with nonspecific
chronic pain (14), and for assessment of the effects of treatment of
fibromyalgia (15). They have also been introduced into the structured
assessment of shoulder conditions (16).

The repeatability of pain diagrams is an important consideration if they
are to be considered valid, but there have been to our knowledge only two
studies of this characteristic. Margolis et al. (17) studied 51 patients
with unspecified chronic pain selected for inpatient treatment and found
high correlations between two administrations of a pain diagram with
respect to notional surface area affected by pain, and a high percentage
level of agreement with respect to individual pain sites, with intervals
between tests averaging 71 days. Roach et al. (18) studied the number of
pain sites marked by 53 patients with low back pain and found significant
differences even between tests carried out on the same day. Repeatability
improved when patients were able to see their previous diagram.

It was considered important to study this area further for two reasons.
Firstly, the results of previous studies are conflicting and may not apply
to new classes of patients. Secondly, an opportunity would be given to
allow a statistical analysis of pain distribution, not simply of pain site
numbers or of individual pain sites.

Diagrams may be used to assess the site of pain, its quality, and its se-
verity. It is difficult to combine all three of these elements while
retaining a diagram that is easy for patients to use and for examiners to
score. Combining variables in a single diagram might also result in a
reduction in its reliability. Accordingly we have investigated a diagram
designed to assess only one of the three variables, site of pain, and so
constructed as to make scoring and analysis as simple as possible.

© 2002 by The Haworth Press, Inc. All rights reserved.