Catherine Bergego Scale (CBS)

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the Catherine Bergego Scale (CBS). Eight articles were reviewed.

Floor/Ceiling Effects

Azouvi et al. (2003) reported that 3.6% of patients with right hemisphere stroke achieved a total CBS score of 0 (no neglect), indicating an adequate floor/ceiling effect.

Reliability

Internal consistency:
Bergego et al. (1995) examined the internal consistency of the CBS with 18 patients with right hemisphere stroke and reported adequate to excellent correlations between the total score and all items (rho range = 0.48 – 0.87, p<0.05).

Azouvi et al. (1996) examined the internal consistency of the CBS with 50 patients with right hemisphere stroke, using Spearman rank correlation coefficients. An adequate correlation was found between the CBS total score and the “mouth cleaning” item (rho=0.58, p<0.0001). An excellent correlation was found between the total score and all other items (rho range from 0.69 – 0.88, p<0.0001). An adequate correlation was found between the therapist’s score and the patient’s self-evaluation (rho = 0.52, p<0.001). There was an excellent correlation between the therapist’s score and the anosognosia score (i.e. therapist’s total score – patient’s self-evaluation score) (rho = 0.75, p<0.0001).

Azouvi et al. (2002) examined the anosognosia component of the CBS in 69 patients with subacute right hemisphere stroke. Patients’ self-assessment score was significantly lower than the examiner’s score (p<0.0001). There was an excellent correlation between the therapist’s score (neglect severity) and the anosognosia score (r=0.82, p<0.0001).

Azouvi et al. (2003) examined the internal consistency of the CBS with 83 patients with right hemisphere stroke and reported adequate to excellent correlations between items (correlation coefficient range = 0.48 – 0.73, p<0.0001). Patients’ self-assessment scores were significantly lower than examiners’ total CBS scores (p<0.0001). An excellent correlation was found between the therapist’s score (neglect severity) and the anosognosia score (r=0.79, p<0.0001). Patients with moderate or severe neglect achieved high anosognosia scores, whereas patients with mild or no neglect achieved negative scores, indicating higher self-rating of severity than scores attributed by the therapist.

Luukkainen-Markkula et al. (2011) examined the internal consistency of the CBS using Spearman correlations. There were adequate to excellent correlations between the CBS total score and all item scores, except for the “eating” item (grooming: r=0.64, p<0.05; mouth cleaning: r=0.73, p<0.05; gaze orientation: r=0.80, p<0.01; knowledge of left limbs: r=0.61, p<0.05; auditory attention: r=0.89, p<0.01; collisions when moving: r=0.89, p<0.01; spatial orientation: r=0.89, p<0.01; dressing: r=0.51, p<0.05; finding personal belongings: r=-0.66, p<0.05). Adequate to excellent correlations were also reported between many CBS items (r range from 0.51 to 0.94, p<0.05 to p<0.01).

Azouvi et al. (2003) examined the reliability of the CBS. The item reliability index was 0.93, resulting in 3.5 strata of significantly different difficulty (p<0.05). Although most items vary in their degree of difficulty, 2 items (grooming, auditory attention), share the same level of difficulty. The person reliability index was 0.88, resulting in 2.7 statistically different levels of ability (p<0.05).

Azouvi et al. (2003) examined the internal structure of the CBS by Rasch analysis. Item 2 (dressing) obtained an outlier outfit value (mnsq = 0.58), indicating potential redundancy of this item due to too little variance. Overall mean fit scores were very close to 1.00, supporting unidimensionality. This is further supported by high positive-point biserial correlation coefficients (i.e. the extent to which the items correlate with the linear measure) between each item score, and cumulative scores obtained across the whole sample.

Intra-rater:
No studies have reported on the intra-rater reliability of the CBS.

Inter-rater :
Bergego et al. (1995) examined the inter-rater reliability of the CBS among 18 patients and found adequate to excellent inter-rater reliability for the 10 items (kappa coefficient range = 0.59 – 0.99). Correlation between the two examiners’ total scores, measured by Spearman rank correlation coefficient, was excellent (rho = 0.96, p<0.0001).

Test-retest:
No studies have reported on the test-retest reliability of the CBS.

Validity

Content:
No studies have reported on the content validity of the CBS.

Criterion:
Predictive:
No studies have reported on the predictive ability of the CBS.

Concurrent:
Bergego et al. (1995) compared the CBS total score with other conventional neglect tests in 18 patients with right hemisphere stroke, using Spearman rank correlation coefficients. Excellent correlations were reported between the CBS and the Bells test, Ogden’s scene drawing task, writing task, reading task and Albert’s Test (rho = 0.72, 0.72, 0.72, 0.70 and 0.67 respectively, p<0.01). There was no statistically significant correlation with a flower drawing task.

Azouvi et al. (1996) compared the CBS total score with other conventional neglect tests among 50 patients with right hemisphere stroke, using Spearman rank correlation coefficients. Excellent correlations were found between the CBS and the Bells test (rho = 0.74, p<0.0001), Albert’s Test (line cancellation) (rho = 0.73, p<0.0001) and a reading task (rho = 0.61, p<0.0001). Adequate correlations were found with the Ogden’s scene drawing task (rho = 0.56, p<0.001) and a daisy drawing task (rho = 0.50, p<0.001). Adequate correlations were found between the CBS anosognosia score and conventional tests of neglect (rho range = 0.45 – .58, p<0.01). A weak correlation was found between the CBS anosognosia score and Bisiach et al.’s (1986) scale of patients’ awareness of neurological deficit (rho = 0.31, p<0.05).

Azouvi et al. (2002) compared the CBS with a battery of conventional pen and paper neglect tasks among 69 patients with right hemisphere stroke. Conventional neglect tasks included the Bells test, a figure-copying task, clock drawing, line bisection tasks (5cm, 20cm), overlapping figures test, reading task and a writing task. There were adequate to excellent correlations between the CBS and all conventional neglect tasks (r range = 0.49 – 0.77, p<0.0001), except for the short (5cm) line bisection task. The strongest correlation was seen with the Bells test (total number of omissions). Further stepwise multiple regression analysis revealed that four conventional neglect tasks – the total number of omissions in the Bells test, starting point in the Bells test, figure copying task and the clock drawing task – significantly predict behavioral neglect. Moderate to excellent correlations were found between the CBS anosognosia score and all conventional neglect tasks (r range = 0.47 – 0.70, p<0.0001), except the short (5cm) line bisection task. Comparison with Bisiach et al.’s (1986) scale of visual and motor anosognosia found a moderate correlation with visual anosognosia (r=0.37, p<0.05), but a poor correlation with motor anosognosia (r=0.29, p<0.05).

Azouvi et al. (2003) compared the CBS total score with three conventional neglect tasks in 83 patients with right hemisphere stroke. Excellent correlations were found with the Bells test (number of omissions) (r=0.76, p<0.0001) and a figure-copying task (r=0.70, p<0.0001), and an adequate correlation was found with a short reading task (number of omissions) (r=0.54, p<0.0001). Comparison of the CBS anosognosia scores with conventional neglect tasks revealed adequate to excellent correlations (r range = 0.43 – 0.72, p<0.01).

Azouvi et al. (2006) reported on unpublished data from 54 patients with left hemisphere stroke. There was an adequate correlation between the CBS score and the Bells test – total omissions (r=0.41, p<0.01) and right minus left omissions (r=0.34, p<0.01). There was no significant correlation with a line bisection test.

Luukkainen-Markkula et al. (2011) compared the CBS with the conventional subtests of the Behavioral Inattention Test (BIT C) among 17 patients with right hemisphere stroke and hemispatial neglect. Only the BIT C line bisection subtest showed statistically significant correlations with the CBS, demonstrating excellent correlations with the grooming (r=-0.64, p≤0.01) and gaze orientation (-0.61, p≤0.01) subtests, and adequate correlations with the auditory attention (r=-0.56, p≤0.05) and spatial orientation (r=-0.54, p≤0.05) subtests and the CBS total score (r=-0.54, p≤0.05). Conversely, the CBS eating item was the only item to demonstrate statistically significant correlation with the BIT C, revealing excellent correlations with the line crossing (r=-0.95, p≤0.01), letter cancellation (r=0.83, p≤0.01) and star cancellation (r=-0.83, p≤0.01) subtests and the BIT C total score (r=-0.83, p≤0.01).

Despite significant correlations between the CBS and traditional visual neglect tests, individual patients can demonstrate dissociations between visual and behavioral neglect (Azouvi et al., 1995, 2002, 2006; Luukkainen-Markkula et al., 2011).

Construct:
Azouvi et al. (2003) conducted a conventional factor analysis on raw scores, revealing a single underlying factor that explained 65.8% of total variance. The factor matrix showed that all 10 items obtained a high loading on this factor (range = 0.77 – 0.84). Further, principal component analysis on standardized residuals after the linear method was extracted showed that no strong factors remained hidden in the residuals between observed and expected scores.

Known Group Validity:
Azouvi et al. (1996) compared behavioral neglect in patients with neglect on conventional tests and patients with no neglect, using Mann Whitney tests. There was a significant difference in total CBS scores between the two groups (p<0.0001). Comparison of anosognosia revealed a significant difference between groups (p<0.001), whereby patients with no neglect tended to overestimate their difficulties (compared to the difficulties reported by the therapist) and patients with neglect tended to underestimate their neglect difficulties.

Azouvi et al. (1996) reported no significant difference in CBS anosognosia scores between patients with depression and patients without depression.

Luukkainen-Markkula et al. (2011) compared behavioral neglect in patients with right hemisphere stroke and hemispatial neglect with visual field deficits (n=8) and patients with intact visual fields (n=9), using Mann Whitney tests. Patients with visual field deficits demonstrated significantly more severe behavioral neglect (i.e. higher CBS total score) than patients with intact visual fields (p=0.03).

Convergent/Discriminant Validity :
Azouvi et al. (1996) examined the ability of the CBS to measure aspects of daily functioning related to neglect by comparing CBS and Barthel Index scores of 50 patients with right hemisphere stroke, using Spearman’s rank correlation coefficient. An excellent correlation was found between CBS total score and the Barthel Index (rho = -0.63, p<0.0001).

Azouvi et al. (2006) reported on unpublished data from 54 patients with left hemisphere stroke. There was an adequate correlation between the CBS score and the Functional Independence Measure (r=-0.48, p<0.01), and the Postural Assessment for Stroke Scale (r=-0.55, p<0.001).

Responsiveness

Samuel et al. (2000) reported that the CBS is responsive to clinical change following visuo-spatio-motor cueing intervention in patients with stroke and unilateral spatial neglect.

Sensitivity:
Azouvi et al (1995) examined the sensitivity of the CBS by determining the ability of each CBS item to detect the presence of neglect (i.e. a score of 1 or more) in a group of 50 patients with right hemisphere stroke. The most sensitive items at detecting neglect were items 2 (dressing), 4 (knowledge of left limbs) and 8 (collisions while moving), which all demonstrated neglect in more than 50% of patients. Sensitivity of other items in descending order was as follows: item 1 (grooming), 3 (eating), 10 (personal belongings), 5 (gaze orientation), 7 (auditory attention), 9 (spatial orientation) and 4 (mouth cleaning). The CBS was more sensitive to neglect than conventional neglect tasks (Bells test, reading task, line cancellation test), which were found to detect neglect in 42 – 49% of patients.

Azouvi et al. (2002) examined the sensitivity of the CBS among 69 patients with right hemisphere neglect. Similar to their earlier study (Azouvi et al., 1995), the most sensitive items were item 4 (knowledge of left limbs), item 8 (collisions) and item 2 (dressing). The authors compared the sensitivity of the CBS with a battery of conventional neglect tasks that comprised the Bells test, a figure copying task, clock drawing, line bisection task, overlapping figures test, reading task and a writing task. Stepwise multiple regression analysis revealed that four conventional neglect tasks – the total number of omissions in the Bells test, starting point in the Bells test, figure copying task and the clock drawing task – significantly predict behavioral neglect. These tasks revealed neglect in 71.84% of patients, but did not indicate neglect in 16.38% of patients who demonstrated neglect on the CBS. The highest incidence of neglect in conventional tests was 50%, whereas neglect was seen on at least 1 of the 10 CBS items in 76% of patients. This indicates that the CBS was more sensitive to neglect, although the difference in sensitivity between the CBS and the battery of conventional neglect tasks was not statistically significant.

Azouvi et al. (2003) compared the sensitivity of the CBS to three conventional neglect tasks in 83 patients with right hemisphere stroke. Incidence of neglect on individual conventional tasks was 44.3% on a figure copying task, 53.8% on the Bells Test and 64.2% on a reading task; 65.4% of participants showed neglect on at least one conventional neglect task and 32.7% of participants showed neglect on all three tasks. By comparison, 96.4% of participants demonstrated neglect based on total CBS score. Sensitivity of individual CBS items ranged from 49.5% (auditory attention, spatial orientation) to 79.5% (dressing).

Azouvi et al. (2006) reported on unpublished data from 54 patients with left hemisphere stroke, which showed that 77.3% of patients showed neglect on one item of the CBS scale. Only 5.4% of patients showed clinically significant neglect, compared to 36% of patients with right hemisphere neglect reported in an earlier study by Azouvi et al. (2002). The three items “neglect of right limbs’, “dressing” and “mouth cleaning after eating” showed higher incidence of neglect among patients with left hemisphere stroke, whereas the item “collisions while moving” obtained lower neglect scores. This is in contrast to earlier studies with patients with right hemisphere neglect (Azouvi et al., 1996, 2002).

References
  • Azouvi, P., Bartolomeo, P., Beis, J-M., Perennou, D., Pradat-Diehl, P., & Rousseaux, M. (2006). A battery of tests for the quantitative assessment of unilateral neglect. Restorative Neurology and Neuroscience, 24, 273-85.
  • Azouvi, P., Marchal, F., Samuel, C., Morin, L., Renard, C., Louis-Dreyfus, A., Jokie, C., Wiart, L., Pradat-Diehl, P., Deloche, G., & Bergego, C. (1996). Functional consequences and awareness of unilateral neglect: Study of an evaluation scale. Neuropsychological Rehabilitation, 6(2), 133-150.
  • Azouvi, P., Olivier, S., de Montety, G., Samuel, C., Louis-Dreyfus, A., & Tesio, L. (2003). Behavioral assessment of unilateral neglect: Study of the psychometric properties of the Catherine Bergego Scale. Archives of Physical Medicine and Rehabilitation, 84, 51-7.
  • Azouvi, P., Samuel, C., Louis-Dreyfus, A., Bernati, T., Bartolomeo, P., Beis, J-M., Chokron, S., Leclercq, M., Marchal, F., Martin, Y., de Montety, G., Olivier, S., Perennou, D., Pradat-Diehl, P., Prairial, C., Rode, G., Siéroff, E., Wiart, L., Rousseaux, M. for the French Collaborative Study Group on Assessment of Unilateral Neglect (GEREN/GRECO). (2002). Sensitivity of clinical and behavioral tests of spatial neglect after right hemisphere stroke. Journal of Neurology, Neurosurgery and Psychiatry, 73, 160-6.
  • Bergego, C., Azouvi, P., Samuel, C., Marchal, F., Louis-Dreyfus, A., Jokie, C., Morin, L., Renard, C., Pradat-Diehl, P., & Deloche, G. (1995). Validation d’une échelle d’évaluation fonctionnelle de l’héminegligence dans la vie quotidienne: l’échelle CB. Annales de Readaptation et de Medecine Physique, 38, 183-9.
  • Luukkainen-Markkula, R., Tarkka, I.M., Pitkänen, K., Sivenius, J., & Hämäläinen, H. (2011). Comparison of the Behavioral Inattention Test and the Catherine Bergego Scale in assessment of hemispatial neglect. Neuropsychological Rehabilitation, 21(1), 103-116.
  • Menon, A. & Korner-Bitensky, N. (2004). Evaluating unilateral spatial neglect post stroke: Working your way through the maze of assessment choices. Topics in Stroke Rehabilitation, 11(3), 41-66.
  • Plummer, P., Morris, M. E., & Dunai, J. (2003). Assessment of unilateral neglect. Physical Therapy, 83, 732-40.
  • Samuel, C., Louis-Dreyfus, A., Kaschel, R., Makiela, E., Troubat, M., Anselmi, N., Cannizzo, V., & Azouvi, P. (2000). Rehabilitation of very severe unilateral neglect by visuo-spatio-motor cueing: Two single case studies. Neuropsychological Rehabilitation, 10(4), 385-99.