Trunk Impairment Scale - Adaptation and Validation in French

he validity of the content is evaluated at four levels, described in detail in the methodology section of the translation and cross-cultural adaptation phase:

By the reading committee (for the translation, reconciliation, and back-translation phases) By a group of professional experts in the field (to approve or not the pre-final versions) By comparing the items of the original TIS with its back-translation, by Prof. Verheyden and his team By a cognitive interview with patients, to assess the proper understanding of each item.

The TIS is administered several times by the same examiner at different moments.

It is also considered as "the degree to which patients' scores remain the same during repeated measurements under several conditions" over time (test-retest), by the same person at different moments (intra-examiner), and at the same moment by different observers (inter-examiner). Two indices are used:

• Cohen's Kappa coefficient, which determines the level of agreement to compare results item by item. Kappa is used for dichotomous item results, and weighted Kappa for ordinal item results. According to Landis and Koch, agreement is considered strong above 0.6.
• The intraclass correlation coefficient (ICC) to compare the sub-scores and total scores of the scale. An ICC above 0.8 indicates excellent reliability, between 0.8 and 0.6 good reliability, between 0.6 and 0.4 moderate reliability, and below 0.4 low reliability.

The TIS is administered at the same moment by different observers.

It is also considered as "the degree to which patients' scores remain the same during repeated measurements under several conditions" over time (test-retest), by the same person at different moments (intra-examiner), and at the same moment by different observers (inter-examiner). Two indices are used:

• Cohen's Kappa coefficient, which determines the level of agreement to compare results item by item. Kappa is used for dichotomous item results, and weighted Kappa for ordinal item results. According to Landis and Koch, agreement is considered strong above 0.6.
• The intraclass correlation coefficient (ICC) to compare the sub-scores and total scores of the scale. An ICC above 0.8 indicates excellent reliability, between 0.8 and 0.6 good reliability, between 0.6 and 0.4 moderate reliability, and below 0.4 low reliability.

To assess the internal consistency of behavioral and clinical items. Internal consistency reflects the interrelationship between items and indicates whether each item measures the same underlying construct.

Cronbach's alpha coefficient will be calculated. Values > 0.7 indicate good internal consistency; values close to 1 may indicate redundancy.

To calculate the Standard Error of Measurement (SEM) for behavioral and clinical scores. SEM quantifies the variability of repeated measurements and reflects measurement error. It can be influenced by the examiner, the measurement tool, or the patient. SEM is used to compute the Minimum Detectable Change (MDC).

To determine the Minimum Detectable Change (MDC) for behavioral and clinical measures. MDC represents the smallest change beyond measurement error that reflects a true change in the patient's condition. It will be calculated using : MDC = SEM * 1.96 * √2

Convergent validity will be evaluated by comparing the TIS with the following scales:

• NIHSS (National Institutes of Health Stroke Scale) : This scale measures the intensity and severity of neurological signs caused by stroke, and monitors their evolution.

• Barthel Index : An index assessing functional independence across 10 categories by rating the amount of assistance required by the patient in each domain.

Demeurisse Motor Index : This scale assesses motor control of the upper and lower limbs after a stroke, only on the affected side.

• 10-Meter Walk Test:
• Sit-to-Stand Test

Known-groups discrimination will be performed by:

A subgroup analysis comparing patients who are able to stand independently, those who can walk, those who can walk quickly, those who perform transfers independently, and non-walkers. These categories may be established based on the results of the previously mentioned tests.

The concurrent validity of the TIS will be assessed by comparing its scores with established measures of postural control and trunk stability.

PASS (36 pts): Evaluates balance in supine, sitting, and standing positions, including transitions.

TCT (100 pts): Assesses four trunk movements: rolling toward affected/unaffected side, balancing sitting at edge of bed, and sitting up from lying.

EPA (4 pts): Evaluates seated postural balance with or without support, considering induced or self-generated imbalance.

Results reported as Pearson or Spearman correlations with TIS scores.

It demonstrates the difficulty of the tool in measuring the patient's abilities below (floor) or above (ceiling) a certain level. It shows the inadequacy of the tool for the situation of the patient being assessed. To highlight these effects, a large number of patients is required. The distribution of results becomes asymmetrical if one of these effects is observed.

The absence of responses to an item will be counted. The relevance of the item may be questioned if it accounts for too many missing responses.