Finding a clinical assessment scale - Physical Medicine and Rehabilitation

Description:

The SCIM is a scale measuring functional independence dedicated to the paraplegic and tetraplegic subject following a spinal cord injury. It was developed by the Department of Loewenstein Rehabilitation Hospital in Israel from the finding that the FIM was unfortunately lacking in relevance and sensitivity to change in the spinal cord injury patient.

Three successive versions have been drawn up. Many translations have been the subject of a transcultural adaptation and validation. The latest version III was translated into French but has never been validated.

In its original version (version I) [1][2], the scale composed of 16 items covered 3 functional sections:
- Self-care (4 items): feeding, bathing, dressing and grooming (score 0- 20)
- Respiration and sphincter management (4 items): respiration, sphincter management – bladder, sphincter management – bowel, and use of toilet (score 0 - 40).
- Mobility (6 items): mobility in bed, transfers: bed-wheelchair, transfers: wheelchair-toilet, mobility indoors, mobility for moderate distances, mobility outdoors, stair management, and transfers: wheelchair-car (score 0-40).

The total score extended from 0 to 100The score levels differed from one item to the next and corresponded to a weighting linked to the importance of the item in the functional life of the spinal cord injury patients. The weighting of the items for this version drew on the national consensus of Israeli experts [5].

In the interim version (version II) [3], the scale introduced a distinction in the Self-Care field between bathing the upper and lower body, and between dressing the upper and lower body. It also modified the scoring levels of the sections Grooming, Sphincter management – bladder, Stair management, and Transfers: wheelchair-car, such that the total score still extended from 0 to 100.

In the latest version (version III) [5], the scale increased to 19 items spread over 4 functional sections:
- Personal care (6 items): feeding, bathing upper body, bathing lower body, dressing upper body, dressing lower body, and grooming (score 0 - 20)
- Respiration and sphincter management (4 items): respiration, sphincter management – bladder, sphincter management – bowel, and use of toilet (score 0 - 40)
- Mobility (room and toilet) (3 items): mobility in bed and action to prevent pressure sores, transfers: bed-wheelchair, and transfers: bed-toilet-tub (score 0 - 10)
- Mobility indoors and outdoors (6 items): mobility indoors, mobility for moderate distances (10-100 m), mobility outdoors (>100m), stair management, transfers: wheelchair-car, and transfers: ground-wheelchair (score 0 - 30)

The weighting of the items for this version drew on the national consensus of Israeli, European and American experts [5]. The total score still extends from 0 to 100.

The evaluation is carried out in theory on observation, completed by several members of the team in their field of competency. However, the authors validated its use by self-assessment [8] or by observation and/or interview [9]. Its use is simple and short (average of 30 minutes in observation mode and 15 minutes in interview mode). The scoring is associated with each item.

NB: only the observation grid is associated with this form.

> Access to the scale is free.

Psychometric properties:

General comment on reliability:

The SCIM version I showed the following metrological properties:
1) An inter-rater reliability (all types of examiners combined: occupational therapists, nurses, physiotherapists) that is excellent (r=0.98, p<0.001) but variable from one section to another and from one item to another: the Kappa coefficient varied between 0.66 and 0.98 and the reliability between 72 and 99% depending on the item [1].
2) A construct validity conveyed by the excellent correlation between the SCIM and the FIM (r=0.85, p<0.001) [1] and the WISCI (r=0.97) [12], or the good correlation between the SCIM and the Rivermead Mobility Index (r=0.75) or the Barthel index (r=0.70) [12]
3) Sensitivity to change, all neurological levels combined, proved higher than the FIM, which did not detect 26% of changes observed by the SCIM (p<0.001) [1][2]. The differential in the evolution of the total score for the SCIM is 10.6 points, vs the FIM estimated at 7.5 points (p<0.01). This sensitivity is expressed significantly for the Respiration and Sphincter Management items and for the items linked to mobility indoors and outdoors [2]. This finding is valid in the case of incomplete tetraplegia or paraplegia and for complete paraplegia.

SCIM version II
1) An inter-rater reliability:
- by inter-professional category (registered nurse versus occupational therapists, and registered nurse versus physiotherapists) that is more nuanced [3]. Out of the total score, the inter-rater reliability proved satisfactory (r=0.82 for registered nurse vs occupational therapists and r=0.94 for registered nurse vs physiotherapists). On the other hand, from one item to the next, the Kappa coefficient varied between 0.172 and 0.728 and the inter-rater reliability between 38 and 90%, meaning that the reliability in this version could be challenged for certain items within the framework of a comparison by professional category
- by interview vs observation examination mode [9]. In interview mode, the reliability from one section to the next varied (r varying between 0.765 and 0.940 p<0.0001) without significant difference observed with the observation mode.
2) A factor analysis confirmed the one-dimensional nature of each of the 4 sub-categories, validated moreover by the Rasch model [4].

SCIM version III
1) An inter-rater reliability from one item to the next estimated between 74.5 and 96.2%, a Kappa coefficient between 0.631 and 0.823 (p<0.001) [6], between 0.649 and 0.858 [11], and between 0.491 and 0.835 (p<0.001) [12].
2) An inter-rater reliability from one section to the next conveyed by an intraclass correlation coefficient varying between 0.94 and 0.97 [6], between 0.92 and 0.96 [11], or between 0.84 and 0.96 [12].
3) An excellent total inter-rater reliability (total score), comparable from one study to the next, 0.977[6], 0.956[11] and 0.960[12].
4) An acceptable internal consistency conveyed by a Cronbach's alpha score > 0.7 [6] [11] [12].
5) A construct validity conveyed by:
- a satisfactory correlation between the SCIM III score and that of the FIM (Pearson coefficient = 0.79 p<0.01 [6][11] and 0.84 p<0.001 [12])
- a satisfactory correlation between the SCIM score III in observation mode and that of the SCIM III in self-assessment mode (Pearson coefficient = 0.87 – confidence interval at 95%= 0.82-0.91) [8].
6) Sensitivity to change:
- greater than the FIM. McNemar's test conveying the small changes of at least 1 point between admission and discharge of the patient proved significant (p<0.001) for the following sections: respiration/sphincter management, mobility indoors and outdoors [6]. No significant change for the other sections
- significant regardless of the neurological level (save for the group of patients C1-C4) with a gain of a minimum of 3 points (level C5) and a maximum of 9 points (level C6).
7) A floor effect was observed – the item transfer: ground-wheelchair scored 0 for 53 of 86 patients [11] and for 13 of the 19 items in the case of C1-C4 injuries. All other levels present floor effects for different items [10].
8) A ceiling effect was present for all groups of spinal cord injury levels, relating to items from one level to the next [10].
9) A factor analysis certifies that the responses were part of a one-dimensional model and that the scale of intervals between the responses for each item is linear.

Reference update:

To notify us of a missing reference, please use: contact@scale-library.com

Related scales:

More information:

References: