Shoulder pain mapping for common shoulder disorders
, MPhil thesis, University of Salford.
Pain mapping for specific disorders was described in the literature for face, back and hip
pain, but not for shoulder pain.
The aim of the study was to fill the gap in assessing patients as a whole for common
shoulder disorders, to develop a pathway from subjective experience of patients to
diagnoses of pathology, to ascertain specific patterns of pain in patients with common
shoulder disorders, to describe comprehensive shoulder pain maps and to test these.
The study was designed in three phases with prospective blinded method. The first phase
aimed to establish the pain patterns for common shoulder disorders. The patients, who
presented as new patients with shoulder pain to the outpatient department, were given a
custom-made shoulder mapping form to mark their pain, its character and severity. The
patients’ final diagnoses were coded after investigations and the codes were correlated
with the pain map patterns to achieve the aim. Later, colour-coded maps were established
for each shoulder disorder. SPSS (statistic package) was used for the first study.
The second phase was designed to test the accuracy of the previously established colourcoded pain patterns in the first phase, to assess sensitivity and specificity of the maps for Levent Bayam @00279526
the disease groups and each individual disorder, to improve the previous pain mappings
and to establish an algorithm. This was achieved by collecting the maps from a larger
number of the patients than the first phase’s number and the researcher, who was blind to
the diagnoses, gave his estimations for each map immediately after collecting the maps
from each patient. After all the investigations, treatments and follow-ups, the final
diagnoses were coded.
The final phase was to assess inter-tester reliability. The third phase was to test inter-tester
reliability of the maps by estimating the shoulder diagnoses using algorithm and colourcoded maps by three raters. This test was used to observe a score of how much consensus
or homogeneity there was for the algorithm of the shoulder pain maps. Another aim in the
third phase was to examine if the mapping system is easy to use or requires a lot of
Statsdirect and VassarStats were used to analyse statistical data in the second and the third
phases of the study.
Ethical opinion was sought from the local R&D department and obtained. There was no
conflict of interest.
The first phase of the study included 94 patients and showed that there were definite
patterns for each shoulder disorder and it described colour-coded shoulder pain patterns
according to the radiation of the pain around and beyond shoulder for six shoulder Levent Bayam @00279526
disorders; acromioclavicular joint pathology, instability: Bankart’s, SLAP etc, calcific
tendonitis, rotator cuff pathology, impingement syndrome, gleno-humeral joint arthritis.
This showed a range from a very localised pain such as ACJ pathology to a very widespread pain such as GHJ arthritis.
The second phase of the study included 194 patients and it tested the mapping patterns
from the first phase. The accuracy for the first estimation for individual disorders was
45.4% and the overall accuracy for both estimations was 62.4. The sensitivity was high
especially for instability and it was good for ACJ pathology and impingement syndrome.
This phase clarified the pain patterns further and detailed three groups of pain patterns.
The first group of diseases showed a localized pain around the shoulder and second group
showed radiation of the pain beyond shoulder. Later, the second group was subdivided
into two. Group 2A showed the pain radiation down to elbow level, whereas group 2B
showed radiation below the elbow level.
Third phase was to test inter-tester reliability of the maps by estimating the diagnoses
derived from the maps by 3 raters. It tested the reliability for each disease group and
individual disease. It showed a substantial agreement between the raters (Kappa (κ) =
A definitive pattern of pain distribution and specific types of pain were demonstrated for
common shoulder pathologies. Testing the established maps indicated that the colourcoded maps were reliable and the algorithm was easy to understand. The study advocates
the use of pain maps as an adjunctive diagnostic tool in general practice clinics and
orthopaedic / shoulder clinics.
Actions (login required)
||Edit record (repository staff only)