Disability scales have become prominent and complementary to the traditional outcome measures such as survival or physical assessment in musculoskeletal cancer evaluation. Multiple–language versions of the existing validated questionnaires allow us to standardize the outcome assessment and to increase the statical power of clinical studies. The English version of the MSTS questionnaire had been translated, validated and culturally adapted by other countries including Japan, Turkey, Brazil, Denmark and China in patients with sarcoma of the upper and/or lower extremity. (4) (5) (6) (7) (8) However, it was never translated to the Hindi language and this study was conducted to test whether this tool could evaluate health status outcomes and its psychometric properties. To the best of our knowledge, this is the first time the lower extremity version of the MSTS questionnaire has been translated into the Hindi language following a standardized guideline.
This study revealed that the Hindi versions of the MSTS questionnaires for lower limbs are suitable and adequate tools for measuring functional outcomes in the Indian population. The Hindi version of the MSTS questionnaire showed good internal consistency, inter and intra-observer reliability and construct validity during the postoperative evaluation of patients with lower extremity sarcoma in our study. MSTS is usually the standard measurement tool to evaluate the functional outcome, which is evaluated by the physician. Considering the fact that almost all of the Indian population are Hindi speaking, some questions of the English version of the score were not appropriate for Indian patients, so we developed the Hindi version of MSTS for use in Indian patients. To the best of our knowledge, the current study is the first to test the factor structure of the MSTS rating scale for lower limbs and its reliability properties in the Indian population.
An existing questionnaire must undergo a proper cross-cultural translation to ensure that it measures the same concept as the original measurement while using it in a different subset of population. (11)In this study, we present a cross-cultural adaptation of the original MSTS rating for lower limbs in the Indian population.
The good internal consistency and inter-rater reliability found in this study are comparable to those found by Rebolledo et al (Cronbach's alpha = 0.84) and reliability (test-retest reliability and interobserver agreement of 0.92 and 0.98, respectively), (4) Iwata et al (Cronbach’s alpha coefficient was 0.87 correlation coefficient (0.92; 95% CI, 0.88–0.95) (5)and Xu et al (Cronbach's α of 0.86 correlation coefficient of 0.85–0.96) (6). It is noted that the original English version did not report a Cronbach’s α although it reported good inter-observer reliability. The translation and back translation in our study resulted in minimal discrepancies that were resolved by consensus. The resultant Indian rating score reflects both the semantic and conceptual equivalence to the original English version. Analysis from our study indicated an excellent internal consistency (0.9). This result is similar with that found by Davis et al. (3) (0.91) in a study with 83 patients with lower extremity sarcoma and better than that of Rebolledo et al. (4) In their study, Lee et al. (12)reported Cronbach’s alpha of 0.88 in a study with 49 patients with musculo- skeletal tumors, thus reflecting the internally valid nature of our score – MSTS Hindi.
Measurement errors in the MSTS has been studied previously by Saebye CKP et al. (8) where they demonstrated low mean bias on all plots, however, with wide limits of agreement, which indicated a possible high measurement error. In the present study, there was low mean bias as demonstrated in Figs. 1 and 2 with a narrow limit of agreement indicating a low measurement error. No other study has included measurement error into the analysis. A change in the MSTS score greater than the measurement error should be considered a possible ‘real’ change in the functional outcome and hence the test for measurement error becomes important part of the validation process. (13)
Our study included 97 patients with lower extremity sarcomas. Considering the previous guidelines concerning the validation of instruments have set a minimum of 100 patients as an excellent sample size, while 50 to 99 patients constitute a good sample size, this study included a very good sample size.
This study had several limitations. The limited number of patients and the oncologic type distribution of the patients may have influenced the results. There was a gap of 10 days interval between the two-measurement taken by the same observer and it could potentially cause a deviation in their score. Although it is necessary to wait for this period in validity studies, this could be one of the limitations of our study. Patients were identified and evaluated retrospectively using our institutional records based on diagnosis and the data related to imaging, pathology examination, or intraoperative findings were not available in detail for each patient. A major limitation of our study was that the functional scores like SF-36 and Toronto Extremity Salvage Score (TESS) were not evaluated. This was, in part, also impacted by ongoing global pandemic (COVID-19), which resulted in alteration of routine patient visit and follow up.