The Shoulder Disability Questionnaire (SDQ) demonstrated promising psychometric properties, but currently there is no evidence of an Italian version.
To cross-culturally adapt the Italian version of the SDQ (SDQ-I) and to explore its psychometric properties in patients with non-specific shoulder pain (SP).
The SDQ-I was translated according to international guidelines and administered to 105 patients with SP. The SDQ-I scores were compared with Disabilities of the Arm, Shoulder, and Hand Questionnaire (DASH), Numerical Pain Rating Scale (NPRS), and 36-item Short Form Health Survey (SF-36). Structural validity (Exploratory Factor Analysis [EFA]), internal consistency (Cronbach’s alpha), test-retest reliability (Intraclass Correlation Coefficient [ICC]), measurement error (Standard Error of the Measurement [SEM], Minimal Detectable Change [MDC]), and construct validity (hypothesis testing with correlation of the DASH, NPRS, and SF-36) were explored.
The EFA identified two subscales (activity and participation-psychosocial factors). The internal consistency was supported for activity (α = 0.78), but not for participation-psychosocial subscales (α = 0.45). Both subscales demonstrated high test-retest reliability (ICC = 0.97 [95%CI = 0.96–0.98] and ICC = 0.95 [95%CI = 0.93–0.97]), a SEM of 5.8% and 7.0%, and a MDC of 15% and 19% for the first and second subscale, respectively. Construct validity was satisfactory, as >75% of the expected correlations were met for each subscale.
The SDQ-I was successfully cross-cultural adapted into Italian and proved to be valid for patients with non-specific SP, but its reliability should be further assessed, as internal consistency of the participation-psychosocial subscale showed some psychometric flaws. Further research is needed to refine the instrument.
The Shoulder Disability Questionnaire is a widespread tool to measure disability.
No evidences are available in Italian subjects with shoulder pain.
The Italian version was successfully translated and cross-culturally adapted.
Test-retest reliability, measurement error and construct validity were supported.
Internal consistency was supported for the subscale#1, but not for the subscale#2.
The lifetime prevalence of shoulder pain (SP) has been reported as high as 7%–67% ( ), making SP one of the most common musculoskeletal complaints ( ). SP may hamper the proper movement of the upper limb, and negatively affects daily activity performance, autonomy ( ; ), and sleep quality ( ). Symptoms often persist and become chronic, as only 21%–50% of patients with SP in primary care recover within six months after treatment ( ).
The availability of reliable and valid measurement instruments is necessary in clinical practice and research ( ). Patient reported outcome measures are considered the best means to quantify health-related constructs (e.g. fatigue, physical or psychological functioning) ( ; ): because they capture the patient’s perspective of the impact of a disease on the individual ( ).
Several questionnaires investigating disability in patients with SP are available ( ; ), including the Shoulder Disability Questionnaire (SDQ). The SDQ ( ) aims to assess the restrictions of daily activities secondary to shoulder symptoms. It includes items investigating activities considered to be important by both patients and clinicians (physiotherapists and physicians), as well as questions related to psychological well-being such as irritability and loss of appetite. The SDQ is composed of 22 items, dropped from the Sickness Impact Profile ( ), which are relevant to SP and classified into 14 categories, describing impairments in activities and participation. These categories appear to reflect the range of limitations associated with SP, including domestic, recreational and psycho-social concerns. The SDQ is simple to administer in clinical practice and easy to score, making it a practical measure ( ). Furthermore, the original version of the SDQ even included questions from a group of patients that were experiencing SP ( ). The SDQ has been investigated in multiple conditions to include in patients with SP ( ), patients with adhesive capsulitis ( ), with various sized rotator cuff tears ( ), with chronic non-specific SP ( ) and has been widely used in primary care ( ).
Several international versions of this instrument have been validated and available in the literature; in Turkish ( ), Korean ( ), and Spanish ( ). Currently, an Italian version has not been developed. As the psychometric properties are influenced by social, environmental, and clinical factors ( ) that are specific for Italian patients, it is necessary to assess the SDQ psychometric properties before this useful tool can be used in Italian-speaking patients with SP. This study aims to cross-culturally adapt an Italian version of the SDQ and explore the psychometric properties in patients with non-specific SP.
In this clinimetric study, patients with SP were recruited consecutively in two physiotherapy private practices in Lecce and Bari, Italy, between March 2019 and June 2019. Patients were included in this study if they presented with non-specific SP, defined as pain not attributable to a known cause or is not consequent to a clear structural, pathoanatomic, or pathophysiologic origin ( ; ). Exclusion criteria were: if < 18 years old, presented with specific SP (i.e., symptomatic rotator cuff and labral tears, fractures, X-ray documented severe joint diseases), frozen shoulder, shoulder instability, SP due to cervical impairment, unable to speak Italian, or cognitive impairment. This study was approved by the ethics committee of ASL of Lecce (Italy) (protocol number 28, January 17th, 2019), and patients signed a written informed consent.
Shoulder Disability Questionnaire
The consent was received from the authors of the original English version of the SDQ for development of the Italian version prior to the start of this study.
The 22 dichotomous (yes = 1/no = 0) items of the SDQ ( ) explore symptoms associated with SP, referring to the last 24 h and assessing restrictions of daily activities ( ; ), within the physical, emotional, and social domains ( ). The total score is obtained by summing the score which ranges from 0 to 22, where higher values correspond to higher levels of disability. Average time needed to complete the questionnaire ranges from 2 ( ) to 3.5 min ( ).
Translation and cross-cultural adaptation
In accordance with international guidelines ( ), the SDQ was cross-culturally adapted into Italian in order to ensure consistency in the content and face validity between the original and the Italian version of the questionnaire.
Step 1: Forward translation in Italian. Two Italian-native translators developed two independent Italian versions from the original version. A translator (FB) is a physiotherapist with 14-years of clinical experience with patients with shoulder diseases. The second translator is a professional English-Italian translator with 10-years of experience, no medical background and no knowledge of the SDQ construct.
Step 2: Synthesis. The two independent translations were compared and discrepancies that might reflect either ambiguous wording in the original text or differences in the translation process were identified. Consequently, discrepancies were resolved in a final translated version, after discussion between the translators and a recording observer (LP).
Step 3: Backward translation to English. From the final Italian version, two English-native translators without a medical background and blinded to the original version, translated the questionnaire back into the English language.
Step 4: Expert committee. An expert committee was instituted with the translators, a linguistic expert, two physicians, and three physiotherapists (FB, DP, ML). The linguistic expert is a PhD with 45-years of experience as an English teacher. An orthopaedist with 13-years of experience and a physiatrist with 18-years of experience contributed to the committee. The physiotherapists participating in the expert committee had clinical expertise in treating shoulder conditions and were lecturers that teach Manual Therapy directed towards treating shoulder conditions. The committee analysed the translated document from a semantic, idiomatic, and conceptual equivalence point of view to create the pre-final Italian version of SDQ (SDQ-I).
Step 5: Pretesting. The pre-final version was administered to 30 patients who met the inclusion criteria of the study to explore if both the meaning of the items and responses made sense to the patients. Finally, the definitive SDQ-I was developed.
Other measurement instruments
Disabilities of the Arm, Shoulder, and Hand Questionnaire (DASH): The DASH ( ) is composed of 30 questions measuring physical function, social function, and symptoms in patients with musculoskeletal disorders of the upper limb. Each item is rated with a 5-points Likert scale. As the DASH Italian version ( ), used in this study, demonstrated a three-factors structure ( ), three different scores were calculated: items 1–5, 7–11, 16–18, 20, and 21 composed the Manual Functioning subscale, items 6, 12–15, and 19 composed the Shoulder Range of Motion subscale, and items 22–30 composed the Symptoms and Consequences subscale. The total score of each subscale was computed by adding the scores attributed to each item.
Numerical Pain Rating Scale (NPRS): The NPRS assesses pain over the last week ranging from 0 (“no pain”) to 10 (“the worst experienced pain”). This instrument was deemed to be reliable and valid in patients with SP ( ).
36-item Short Form Health Survey (SF-36): The SF-36 ( ) consists of 36 questions on the general health status of patients, providing eight subscales (Physical Functioning [PF], Role-Physical [RP], Bodily Pain [BP], General Health [GH], Vitality [VT], Social Functioning [SF], Role Emotional [RE], Mental Health [MH]). For each subscale, the total score ranges between 0 and 100. To obtain the Physical Component Score (PCS) (composed by PF, RP, BP, and GH subscales) and the Mental Composite Score (MCS) (composed by VT, SF, RE, MH subscales), each subscale was converted into a z-score, that were combined into a global z-score through a weighted mean, using weights that resulted from a principal component analysis. Finally, the PCS and MCS were standardized to a normal distribution with mean ± SD of 50 ± 10. Higher values indicate better general health status. The Italian version of the SF-36 was used in this study ( ).
The DASH, NPRS, SF-36 and SDQ-I were administered to each patient. In addition, demographic and clinical characteristics were collected. To assess test-retest reliability, the SDQ-I was administered again after 5 days, during which no treatment was provided, to minimize the risk of short-term clinical changes.
The psychometric properties of the SDQ-I were explored according to the definition proposed by the COnsensus-based standards for the Selection of health Measurement INstruments (COSMIN) initiative ( ).
Descriptive statistics included mean ± standard deviation (SD) for interval variables, median with first and third quartile for ordinal variables, and relative frequencies categorical variables. Floor and ceiling effects were considered when more than 15% of the patients obtained the lowest or highest score, respectively ( ). Acceptability was considered as the time required to answer questions and the presence of multiple and/or missing responses ( ).
Validity was assessed as structural validity. As the structural validity was not previously assessed, a Parallel Analysis (PA) ( ) was run to estimate the number of factors (subscales) in the questionnaire to be extracted, and subsequently an Exploratory Factor Analysis (EFA) ( ) for ordinal data was used to determine the contribution of each item to the subscales previously identified. A representative loading of each item on the latent factor >0.4 was considered ( ). PA was performed using Factor software (Version 10.9.02 for Windows, Tarragona; http://psico.fcep.urv.es/utilitats/factor/index.html ) ( ), while EFA was run though Mplus software (version 6.0. Muthen & Muthen, Los Angeles, CA; 1998–2010; www.statmodel.com ).
Reliability was assessed as internal consistency, test-retest reliability, and measurement error. Reliability analyses were performed on each separate subscale, detected previously by factor analysis. Internal consistency was explored by adopting the following criteria: 1) Cronbach’s alpha (α), where recommended values range between 0.70 and 0.95 ( ), 2) α if an item was deleted, where values below total α are expected ( ), and 3) item-to-total correlation, based on Spearman’s correlations; values ≥ 0.25 were considered as satisfactory ( ). The test-retest reliability was assessed calculating the Intraclass Correlation Coefficient (ICC 2,1 ), based on a 2-way random effects model with 95% confidence interval (CI). Before the calculation of the ICC, a non-parametric Wilcoxon test was used to confirm the absence of systematic differences between the two assessments ( ). Reliability would be supported for ICC values > 0.75 for group measures and >0.95 for measures on a single individual ( ). The measurement error was examined computing the Standard Error of Measurement (SEM) and Minimum Detectable Change (MDC). SEM was calculated with the following formula: SD*√(1– ICC), where SD is the baseline SD of the measurements and the ICC value is that of test-retest reliability. MDC was calculated by multiplying the SEM by 1.96, i.e. the z-score associated with the 95% confidence level, and the square root of 2.
Construct validity was assessed through Hypothesis Testing. As the factor analysis revealed two subscales (activity and participation-psychosocial factors – see results), the assumptions were made separately for each subscale. To this purpose, as the data is ordinal, Spearman’s rank correlation (r s ) coefficients were calculated between subscale#1 and the other questionnaires (i.e. three DASH subscales, NPRS, and MCS and PCS SF-36 subscales). The following assumptions were considered: 1) the correlations between the subscale#1 with the three DASH subscales, NPRS and PCS were expected to be ≥ 0.60, because DASH subscales and PCS measure similar constructs (i.e., related to the activity) to that measured by subscale#1, and because pain (expressed by NPRS) showed a strong correlation with disability in patients with SP ( ); 2) the correlation between the subscale#1 and MCS was expected to be ≥ 0.30 and < 0.60, as MCS and patient-reported function correlates in patients with shoulder conditions ( ); 3) the correlations between the subscale#2 with three DASH subscales, NPRS and PCS were expected to be ≥ 0.30 and < 0.60 because the variables, measured by these instruments, represent related, but not identical constructs. Preliminary evidence suggests that shoulder function and pain correlate with psychological distress in patients with shoulder conditions ( ); 4) the correlation between the subscale#2 and MCS to be ≥ 0.60 because they investigate similar constructs (psychosocial and mental factors). The cut-offs of r s ≥ 0.60, 0.30≤r s < 0.60, and r s < 0.30 represent strong, moderate, and weak correlations, respectively ( ). Construct validity was considered satisfactory, moderate, or low if ≥ 75%, ≥50% but <75%, and <50% of expected hypotheses were met, respectively ( ).
SPSS package, version 21 for Windows (SPSS Inc., Chicago, IL; 2004) was used for all statistical analyses.
The full adaptation procedure took approximately 20 days to achieve a culturally adapted Italian version of the SDQ. The SDQ was forward and backward translated into Italian without any issues ( Appendix ). The translation process did not result in disagreement between translators. Complete agreement was achieved among the expert committee for the analysis of the two translated versions from a semantic, idiomatic, experiential, and conceptual equivalence point. The pre-final version of the SDQ-I was administered to 30 patients with SP, who stated that it did not contain unclear words or awkward sentences, confirming the comprehensibility and the cognitive equivalence of the translation.
One-hundred and five patients (mean age = 52.5 ± 13.6 years; 52.4% male) with non-specific SP participated. Detailed demographic and clinical characteristics of the sample are reported in Table 1 .
|Variable||Mean (SD)||Frequency (%)||Median (1st, 3rd quartile)|
|Age, years||52.5 |
|Weight, kg||72.1 |
|Height, cm||170.9 |
|Primary schools||3 |
|Secondary School||15 |
|High school||43 |
|College degree||44 |
|< 3 months||32 |
|Between 3 and 6 months||40 |
|> 6 months||33 |
|Pain irradiating on upper limb|
|Respiratory diseases||14 |
|Endocrine disorders||13 |
|Gastrointestinal diseases||12 |
|Depressive anxiety disorders||3 |
|Nephrological pathologies||3 |
|Musculoskeletal disorders||2 |
|Hearth Diseases||2 |
|Subscale#1 – Activity a||5.0 |
|Subscale#2 – Participation and psychosocial factors b||2.0 |
|Manual functioning c||31.0 |
|Shoulder range of motion d||16.0 |
|Symptoms and consequences e||22.0 |
|NPRS f||4.0 |
|Physical Activity||80.0 |
|Role Physical||50.0 |
|Bodily Pain||42.0 |
|General Health||61.0 |
|Social Functioning||75.0 |
|Role Emotional||100.0 |
|Mental Health||68.0 |
|Physical Component Summary||41.7 |
|Mental Component Summary||42.9 |