Translation, cross-cultural adaptation and validation of English start back screening tool into Bangla for patients with low back pain

This study was intended to translate and culturally adapt the  STarT back screening tool to produce an equivalent  Bangla version. Total 58 patients with low back pain completed the newly developed Bangla version of STarT back screening tool and Bangla version of Ronald Morris Disability Questionnaire seven days apart. Reliability was assessed by internal consistency (Chronbach’s alpha for overall score 0.81 and for spychosocial subscale was 0.76)  and test-retest reliability (intraclass correlation coefficient for overall score was 0.78 and for spychosocial subscale was 0.71). Reliability of  Bangla version of STarT back screening tool was very good. Pearson’s correlation coefficient was carried out on the Bangla version of STarT back screening tool and Bangla version of RMDQ to assess construct validity (overall score was 0.88 and spychosocial subscale score was 0.83) which indicate a strong correlation between them. This study shows that the Bangla version of STarT back screening tool is a reliable, valid and culturally adapted responsive screening tool for the patients with low back pain. 


The STarT back screening tool
The STarT back screening tool is a validated simple, brief, and practical way to subgroup patients with nonspecific low back pain in primary care. The new STarT back screening tool identifies potentially modifiable prognostic indicators that may be appropriate targets for primary care interventions. 5 The STarT back screening tool's appeal is obvious. It is short, easy to complete (most responses yes/no format), and quick to score. It can be easily implemented into busy clinical practice by health care practitioners. The care pathway for each subgroup is well-defined. The low risk subgroup is appropriate for self-| Original | Article |

Abstract
This study was intended to translate and culturally adapt the STarT back screening tool to produce an equivalent Bangla version. Total 58 patients with low back pain completed the newly developed Bangla version of STarT back screening tool and Bangla version of Ronald Morris Disability Questionnaire seven days apart. Reliability was assessed by internal consistency (Chronbach's alpha for overall score 0.81 and for spychosocial subscale was 0.76) and test-retest reliability (intraclass correlation coefficient for overall score was 0.78 and for spychosocial subscale was 0.71). Reliability of Bangla version of STarT back screening tool was very good. Pearson's correlation coefficient was carried out on the Bangla version of STarT back screening tool and Bangla version of RMDQ to assess construct validity (overall score was 0.88 and spychosocial subscale score was 0.83) which indicate a strong correlation between them. This study shows that the Bangla version of STarT back screening tool is a reliable, valid and culturally adapted responsive screening tool for the patients with low back pain. The high risk subgroup should be referred to an appropriately skilled Rehabilitation physician to address psychological and physical obstacles to recovery. The keele STarT back screening tool is a 9item, self-reported questionnaire that includes treatment modifiable domains (spread of pain, disability, and psychological factors). It subgroups patients with non-specific low back pain into low, medium, and high risk of future disability with the purpose of matching each subgroup to a care pathway.

The adaptation process
The STarT back screening tool was adapted to the Bangladeshi population according to the guidelines of the Beaton et al forward-backward method. 19 Two bilingual translator independently translated the original English version of STarT back screening tool into Bangla. The independent translation matched absolutely. The two resulting Bangla draft versions were then synthesized into a single Bangla version by two translators. This version was then given to a third researcher, an independent native English speaker fluent in Bangla for back translation into English. The back translation matched the original version. To remove any comprehension barriers, the pre-final Bangla version was tested on 20 volunteer patients to guarantee that participants would define each item as originally intended. The final Bangla version was then evaluated to verify its reliability and validity.

Participants
Inclusion criteria: Native Bangladeshi patients who are diagnosed with low back pain by a physician, aged from 18 to 80 years, and whose mother language were Bangla.
Exclusion criteria: Patients with tumor, ankylosing spondylitis, spinal fracture, history of surgery, and with comorbid serious systemic diseases were excluded. Patients who are unable to fill the questionnaire due to visual, cognitive, literacy related problems are also excluded.

Reliability
Internal consistency: Bangla version of STarT back screening tool was tested in 58 patients to yield internal consistency by Cronbach's alpha. A Cronbach's alpha of 1.0 indicates a perfect correlation among all items, whether ≥0.7 indicates satisfactory internal consistency.

Test-retest reliability:
For test-retest reliability, the Bangla version of STarT back screening tool was served to consecutive subjects of 58 low back pain patients after being enrolled along with their informed consent where the participants were requested to fill questionnaire again after 7 days of the first enrollment without any medications adjustment in between. Intraclass correlation coefficient was used to assess the test-retest reliability. The intraclass correlation coefficient was expected to be found ≥0.70; if so considered as acceptable for testretest reliability.

Validity
Construct validity: Construct validity was assessed by correlating the Bangla version of Roland Morris Disability Questionnaire (RMDQ) 20 with Bangla version of STarT back screening tool and examined by Pearson correlation coefficient. RMDQ is a widely used outcome measure which showed excellent responsiveness in Bangladeshi population. Correlation coefficient were rated as follows: Strong correlation >0.6, moderate ≤0.6 and >0.3 and poor ≤0.3.

Analysis
Cronbach's alpha was used to evaluate internal consistency, on the other hand Intraclass correlation coefficient was used to assess the test-retest reliability. Construct validity was assessed by correlating the RMDQ with Bangla version of STarT back screening tool and examined by Pearson correlation coefficient. All statistical analyses were performed using SPSS version 24.0. A p value less

Results
Among our study population just more than 20 were female and 38 were male. The mean age of which was 37.4 ± 12.4 years, for the most part of them were middle aged and older. Regarding literacy level, 6 were illiterate, 12, 14 and 19 went up to primary, secondary and higher secondary level respectively and 7 were graduate. Regarding profession most of them (21) were housewives, 7 were businessmen, 9 were garment worker, 5 were teacher, 6 were rikshaw puller, 4 were retired person, 4 were unemployed, and 2 were guard.

Translation process
We didn't come across any major linguistic or cultural barrier during forward and back translations. There was some minor inconsistencies. 6 of our patients were rikshaw puller and they were illiterate. For them some words had no equivalent in Bangla and were translated to the closest possible meaning, such as 'short distance' in questions 3, 'worrying' in question 6, 'terrible' in question 7 , 'enjoy' in question 8 and 'bothersome' in question 9. Certain phrases such as 'spread down' in question 1 and 'physically active' in question 5 posed difficulties for both forward translators.
But after finalize the Bangla version of STarT back screening tool almost all of our participants had no intricacy to recognize the translated items, and all of them responded to all components of the Bangla version of STarT back screening tool .

Internal consistency (Cronbach's alpha):
Internal consistency with Cronbach's alpha for Bangla version of STarT back screening tool instrument is 0.81 for overall score and 0.76 for psychosocial subbscale. Alpha should be >0.7, which is standard for all scores. 21

Test-retest reliability
Intra-class correlation co-efficient of Bangla version of STarT back screening tool: Calculated overall tool score is 0.78 and Psychosocial subscale score is 0.71 which signified substantial reliability. 22

Validity
Construct validity: The construct validity of Bangla version of STarT back screening tool is presented in Table I, which demonstrated that the overall tool score is 0.88 and psychosocial score is 0.83 against reference standard RMDQ. All of the items were statistically significant (p<0.001).

Discussion
The STarT back screening tool is one of the critically demanding and internationally renowned tools for screening low back pain patients and is called for its reliability, validity, development in different cultures, and responsiveness to clinical development. 23 The Bangla version of STarT back screening tool takes approximately 10 min to be completed, so it doesn't foist any extra strain on patients and researchers.
Reliability and validity is the core point of quality assessment. A reliable instrument yields the Xerox result on diverse occasions.

Ethical Issue
This study was approved by the institutional review board of Bangabandhu Sheikh Mujib Medical University, (BSMMU/2019/8433) Dhaka, Bangladesh and all participants signed an informed consent to participate in the study.