Keywords
arthroscopy - joint instability - shoulder joint - sports medicine - surveys and questionnaires
- translations
Introduction
Complaints of shoulder abnormalities are frequent in the orthopedic practice. The
shoulder can present clinical manifestations with painful complaints of varying degrees
and episodes of instability.[1] The particular biomechanics of the shoulder partly explain these abnormalities,
since daily and sporting activities require a high shoulder joint range of motion
and mechanical demands. Shoulder instability is a common issue that mainly affects
patients in the second and third decades of life.[2] The incidence of traumatic dislocation is of approximately 1.7% in the general population.[3] After the first episode, symptoms of chronic shoulder instability may appear with
variable intensity and frequency. Although pain may be present, it is not the main
symptom. The predominant complaints are apprehension and loss of confidence, resulting
in reduced participation in sporting activities and a general decrease in quality
of life.[4]
[5]
Many treatments have been proposed for different types of instability; however, few
assessment instruments prove the effectiveness of these treatments.[2] The clinical examination alone does not adequately reflect the functional impairments
experienced by patients. Thus, a subjective analysis of the patient1s quality of life
becomes a significant criterion.[4]
Initially, the best method to evaluate posttreatment outcomes was observation. However,
observation ignores the patient's perception of their health condition. Several studies[6] have demonstrated that objective functional assessments of the shoulder can be inconsistent
and susceptible to errors.
Thus, self-administered questionnaires have been recommended as an effective method
to assess the functional status and perception of symptoms by the patient, valuing
their opinion about their health condition.[7]
Different scores, indices, and scales evaluate the functional impact of shoulder conditions
and health in general. These tools comprise generic assessment (such as the 36-Item
Short Form Health Survey, SF-36),[8] limb-specific (such as the Disabilities of the Arm, Shoulder, and Hand [DASH] questionnaire
and its shortened version, the QuickDASH),[9] shoulder-specific (such as the American Shoulder and Elbow Surgeons Score [ASES],
the Constant-Murley score, the University of California-Los Angeles [UCLA] Shoulder
Scale, the Subjective Shoulder Value [SSV]),[10] and condition-specific instruments (such as the Rowe score, the Western Ontario
Shoulder Instability Index [WOSI], the Western Ontario Rotator Cuff [WORC], the Melbourne
Instability Shoulder Scale [MISS], and the Oxford Shoulder Instability Questionnaire
[OSIQ]).[6]
[11]
[12]
[13]
Specific self-administered questionnaires for shoulder instability have been developed
recently, such as the WOSI, OSIQ, and MISS. However, the MISS has shown less satisfactory
psychometric properties than the WOSI.[4]
The WOSI is simpler, more effective, and more reproducible compared with other instruments.
Due to these characteristics, many consider it the best assessment method for patients
with shoulder instability.[7]
[14] Furthermore, it is easy to use (with an estimated completion time of 3 minutes),
reliable, reproducible, sensitive to changes, and it has undergone validation in multiple
languages.[2]
[4]
[5]
[7]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
The WOSI is a quality-of-life questionnaire developed for use in English. The study
of its psychometric properties showed a strong correlation with the DASH and the UCLA
instruments. In the assessment of reproducibility, the intraclass correlation coefficient
was excellent. It was developed and validated to be applied to patients with shoulder
instability. As a specific instrument, it encompasses aspects of quality of life relevant
to this disease. It contains 21 questions covering 4 domains: 1) physical symptoms;
2) sports, recreation, and work; 3) lifestyle; and 4) emotional status.[2]
[5]
The answers to the WOSI questions are provided through the Visual Analog Scale (VAS).
All questions have the same weight value. Therefore, each item may receive a score
from 0 to 100 on the VAS, and the final result ranges from 0 to 2,100. The closer
the final result is to the lower limit, the lower the disease impact and the less
significant the change in quality of life.[5]
The development of translation and cultural adaptation methods have enabled the use
of an instrument developed in a given language and culture in another language and
another cultural context after translation and adaptation.[23]
[24] Barbosa et al.[5] performed this process in 2012, obtaining the Brazilian Portuguese version of the
questionnaire.[5]
[13]
The validation stage consists of checking whether the new instrument retained the
characteristics of the original version. This entire process is relevant so that the
new tool is equivalent to the original version and culturally accepted in the country
in question.[5]
Therefore, although the original WOSI already had its psychometric properties studied,
the objective of the present study is to evaluate its validity in the Brazilian Portuguese
version.
Materials and Methods
Patient Selection
The inclusion criteria were the following: Patients who attended the shoulder and
elbow outpatient clinic at Hospital Unimed Belo Horizonte, in the state of Minas Gerais,
Brazil, and underwent an assessment for regular follow-up after treatment for shoulder
instability. We divided these subjects into those receiving surgical and conservative
treatments. The control group consisted of patients with other shoulder conditions
but no instability and volunteers without orthopedic shoulder complaints. We excluded
patients who chose not to sign the informed consent form (ICF), and those unable to
fill out the forms due to psychiatric or psychological issues, neurological conditions,
systemic inflammatory diseases, neoplasms, or cervical radiculopathy. All patients
participating in the study signed the ICF.
Sample calculation followed the principles set out by Cochran[25] and Pereira.[26] The Cochran[25] formula calculates an ideal sample with predetermined precision levels (margin of
error) and confidence intervals. For our population, we chose a 95% confidence interval
(95%CI) and a 5% margin of error. Pereira's[26] formula assesses the sample size in linear correlation studies, as in the current
study. Per this author,[26] the sample calculation employs the Spearman correlation coefficient. For the present
study, we used the coefficient reported by Perrin et al.,[4] who validated the WOSI for French using scores similar to those analyzed by our
group. We found a maximum variation of three patients for the sample calculation employing
the different variables under analysis (scores). Therefore, as there is more than
one variable in the current study, we used the highest value derived from the sample
calculation, finding a minimum number of 17 patients per group and totaling 51 subjects.
We separated the patients into three groups:
-
Conservative group, consisting of 17 patients with shoulder instability who did not
undergo surgical treatment;
-
Postoperative group, consisting of 17 patients with shoulder instability who underwent
surgical treatment with a minimum follow-up of 12 months; and
-
Control group, with 17 patients without shoulder instability.
Data Collection
Two orthopedists specializing in shoulder and elbow surgery assessed all the patients
and applied the following functional and health assessment scores: WOSI, Rowe, VAS,
SSV, DASH, and UCLA ([Appendix 1]).
Results
We evaluated 51 patients, including 34 with anterior shoulder instability, 17 undergoing
surgical treatment, and 17 undergoing conservative treatment. The other 17 subjects
from the control group did not have shoulder instability.
[Table 1] presents the demographic variables sex and age. The groups were homogeneous, with
no significant difference among them (p > 0.05).
Table 1
|
Sex: n (%)
|
p-value
|
|
Male
|
Female
|
|
Postoperative group (n = 17)
|
15 (88.24%)
|
2 (11.76%)
|
0.8553T
|
Conservative group (n = 17)
|
15 (88.24%)
|
2 (11.76%)
|
Control group (n = 17)
|
14 (86.28%)
|
3 (13.72%)
|
Total (n = 51)
|
44 (82.36%)
|
7 (17.64%)
|
|
|
Age (years): mean(± standard deviation)
|
p
-value
|
Postoperative group (n = 17)
|
26.88(± 6.33)
|
0.2391T
|
Conservative group (n = 17)
|
29.12(± 5.20)
|
Control group (n = 17)
|
26.0 (± 4.42)
|
Total (n = 51)
|
27.35(± 5.43)
|
|
The comparative score analysis occurred as follows: normality test (Shapiro-Wilk and
Kolmogorov-Smirnov), median and percentile measurements (25th and 75th percentiles
– P25 and P75 respectively), and statistical data analysis (Kruskal-Wallis test followed
by the Dunn test).
Regardless of the score, the postoperative and conservative groups showed significant
differences compared with the control group (p < 0.05), except for the postoperative group in the Rowe score.
Only the VAS score was unable to demonstrate a statistical difference when comparing
the postoperative and conservative groups (p > 0.05) ([Table 2]).
Table 2
|
Posperative
(n = 17)
|
Conservative
(n = 17)
|
Control
(n = 17)
|
Total
(n = 51)
|
|
|
Median
(25P–75P)
|
Median
(25P–75P)
|
Median
(25P–75P)
|
Median
(25P–75P)
|
p-value
|
WOSI
|
425.0*$
(112.5–97.5)
|
1000.0*
(870.0–348.0)
|
0.0
(0.0–122.5)
|
360.0
(10.0–340.0)
|
< 0.0001KW
|
VAS
|
2.0*
(0.0–4.0)
|
4.0*
(1.5–6.0)
|
0.0
(0.0–1.0)
|
1.0
(0.0–4.0)
|
0.0004KW
|
SSV
|
80.0*$
(55.0–97.5)
|
50.0*
(45.0–65.0)
|
100.0
(99.0–100.0)
|
80
(50.0–100.0)
|
< 0.0001KW
|
Rowe
|
100.0$
(85.0–100.0)
|
50.0*
(40.0–75.0)
|
100.0
(99.0–100.0)
|
95.0
(55.0–100.0)
|
< 0.0001KW
|
DASH
|
7.5*$
(1.3–13.8)
|
25.8*
(17.5–38.8)
|
0.0
(0.0–2.1)
|
7.5
(0.0–19.2)
|
< 0.0001KW
|
UCLA
|
33.0*$
(29.0–35.0)
|
29.0*
(22. –29.0)
|
35.0
(35.0–35.0)
|
33.0
(29.0–35.0)
|
< 0.0001KW
|
The correlation among scores was assessed through the Spearman correlation coefficient,
which evaluates the intensity degree in the relationship between two non-parametric
variables. This coefficient correlation ranges from -1 and +1, and the higher the
absolute value, the stronger the relationship among scores.
All instruments presented a significant relationship among themselves according to
Spearman coefficient (p < 0.05). However, the correlation level was different, as follows:
-
The correlation was perfectly positive (with an r-value close to 1) for the WOSI and the DASH instruments (r = 0.96). This result indicates that the measurements from both indicators increase
significantly at the same intensity.
-
The comparison of the WOSI and UCLA (r = 0.87), DASH and UCLA (r = 0.86), SSV and Rowe (r = 0.80), VAS and DASH (r = 0.75), VAS and UCLA (r = 0.74), and WOSI and VAS (r = 0.72) also showed a trend towards positive linearity between measurements.
-
The correlation of the WOSI and SSV, WOSI and Rowe, DASH and Rowe, SSV and UCLA (r = -0.83), SSV and DASH (r = -0.79), Rowe and UCLA (r = -0.78), VAS and SSV (r = -0.68), and VAS and Rowe (r = -0.60) was negative.
Discussion
The assessment and analysis of the outcomes of certain medical treatments have evolved
in recent times. The patient's perception of their problem has become more valued
in the current context. Thus, scores relying on a more subjective analysis have become
the most recommended methods.
Although there are other assessment methods for shoulder instability, the WOSI has
been considered the best one,[7]
[14] and it has already been validated in several languages.[2]
[4]
[5]
[7]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22] The choice of the analysis method for the current study relied on previously published
articles validating this score.
The selected scores considered the potential comparison with other WOSI validation
studies and encompass different assessment instruments, whether generic (VAS), limb-specific
(DASH), shoulder-specific (UCLA, SSV), or condition-specific (Rowe). The DASH has
been validated in Brazilian Portuguese, and the UCLA[10] and Rowe[11] scores have undergone translation and cultural adaptation to Brazilian Portuguese.
The use of the VAS use in its Brazilian Portuguese version is widespread despite the
lack of a published specific validation. We attribute this to the fact that it is
easily understandable, since it is a visual scale, with no complex questions that
would depend on cultural adaptation. Likewise, the SSV is a subjective assessment
performed by the patient and expressed as a percentage, using 100% as a reference
value for a completely normal shoulder.
All patients in the postoperative and conservative groups presented anterior instability
(to reduce bias in result interpretation). The average age of the study patients (of
27.35 ± 5.43 years) and the sex distribution (44 men [82.36%] and 7 women [17.64%])
were consistent with that of previous studies[4]
[7]
[18] and homogeneous at group comparison ([Table 1]).
The analysis of the validity of the WOSI in Brazilian Portuguese, according to Spearman
coefficient, revealed a significant good or excellent correlation among all scores
(presenting values close to -1 or 1). This coefficient is classified as excellent
(> 0.91), good (0.90–0.71), fair (0.70–0.51), poor (0.50–0.31), and null (<c0.31).[4]
Individually, the WOSI presents an almost a perfect correlation with the DASH score
(0.96) and a high correlation with the UCLA (0.87) and VAS (0.72) scores ([Table 3]). These results are consistent with those of the original article (DASH = 0.77;
UCLA = 0.65)[2] regarding the French (QuickDASH = 0.78; VAS = 0.83),[4] Turkish (DASH = 0.67),[7] German (UCLA = 0.61),[21] Dutch (DASH = 0.81),[20] Swedish (VAS = 0.80),[15] Italian (DASH = 0.79),[18] and Japanese (QuickDASH = 0. 63) validations.[17] All showed a positive linearity for such scores, especially the DASH and UCLA.
Table 3
|
|
WOSI
|
VAS
|
SSV
|
Rowe
|
DASH
|
UCLA
|
WOSI
|
r-value
|
|
0.72
|
-0.83
|
-0.83
|
0.96
|
0.87
|
|
p-value
|
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
VAS
|
r-value
|
0.72
|
|
-0.68
|
-0.60
|
0.75
|
0.74
|
|
p-value
|
< 0.0001
|
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
SSV
|
r-value
|
-0.83
|
-0.68
|
|
0.80
|
-0.79
|
-0.83
|
|
p-value
|
< 0.0001
|
< 0.0001
|
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
Rowe
|
r-value
|
-0.83
|
-0.60
|
0.80
|
|
-0.83
|
-0.78
|
|
p-value
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
|
< 0.0001
|
< 0.0001
|
DASH
|
r-value
|
0.96
|
0.75
|
-0.79
|
-0.83
|
|
0.86
|
|
p-value
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
|
< 0.0001
|
UCLA
|
r-value
|
0.87
|
0.74
|
-0.83
|
-0.78
|
0.86
|
|
|
p-value
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
< 0.0001
|
|
The Rowe score correlation (-0.83) presented good negative linearity, similar to the
French (-0.54)[4] and Turkish (-0.57)[7] analysis, and the original (0. 61),[2] German (0.62),[21] and Swedish papers (0.59),[15] but with positive linearity. The SSV presented good negative linearity (-0.83),
but no comparison with other studies.
The positive points of the present work include the number of analyzed scores (five,
higher than other validation studies using three or four scores), group homogeneity,
the representation of shoulder instability epidemiology, and the high significant
correlation among the scores. The negative points include the small but significant
sample size and the failure to perform test-retest and interobserver analysis.
Conclusion
We concluded that the Brazilian Portuguese version of the WOSI presents good validity.