Online Resource 2: Additional material section results

Cognitive screening in patients with intracranial tumors: Validation of the BCSE

Journal of Neuro-Oncology

Authors:

Juliane Becker 1, MA; Elisabeth Steinmann 1, MA; Maria Könemann 1, MA; Sonja Gabske 1, MA; Hubertus Maximilian Mehdorn 1, MD; Michael Synowitz 1, MD; Gesa Hartwigsen 2, PhD, MA; Simone Goebel 1, PhD, MA

1 Department of Neurosurgery, University Hospital Schleswig–Holstein, Kiel, Germany

2 Department of Psychology, University of Kiel, Kiel, Germany

Address correspondence to Juliane Becker, E-Mail:

Online Resource 2 figure 1: Flowchart of patient recruitment

AAT = Aachener Aphasie Test; KPS = Karnofsky Performance Scale

Prevalence of cognitive impairment in patients with intracranial tumors

According to the screening tools, 28.1% up to 69.0% of the patients were classified as cognitively impaired. For details see table 1. 51.9% (n = 28) of the patients were classified as impaired in at least one cognitive domain. Most frequently, cognitive impairment was found in only one domain. Patients showed most often deficits in memory and visuo-spatial functions.

Online Resource 2 table 1: Cognitive impairment in patients
Screening tools° / Md / range / n / %
BCSE (n=57) / 52 / 16-58 / 22 / 37.9
MoCA (n=58) / 23 / 11-30 / 40 / 69.0
MMSE (n=57) / 28 / 18-30 / 16 / 28.1
No. of impaired domains ^ (N=54) / n / % / cumulative %
0 / 26 / 48.1 / 100
1 / 15 / 27.8 / 51.9
2 / 9 / 16.7 / 24.0
3 / 3 / 5.6 / 7.4
4 / 1 / 1.9 / 1.9
z-scores / z < -1.5
Domain / M / SD / n / %
Language (n = 55) / 0.18 / 1.12 / 3 / 5.6
Attention and
visuo-motor speed (n = 53) / -0.39 / 1.31 / 6 / 11.3
Memory (n = 53) / -0.75 / 1.10 / 15 / 28.3
Executive functions (n = 50) / -0.62 / 1.24 / 9 / 18.0
Visuo-spatial functions (n = 54) / -1.19 / 1.82 / 15 / 27.8

BCSE = Brief Cognitive Status Exam; M = arithmetic mean; Md = Median; MMSE = Mini Mental State examination; MoCA = Montreal cognitive assessment; n = number; SD = standard deviation; °classification on basis of the manuals (MMSE cut-off: <26, MoCA cut-off <26, BCSE 15% percentile rank); ^ full protocol could not be applied to all patients: Priority was given to tests that were most relevant to the preparation and evaluation of the operative procedure (see [1]).

Sample characteristics healthy control group

Online Resource 2 table 2: Sample characteristics healthy control group (N= 22)
test statistics³
Mean age in years / M = 55.00 (SD= 15.37)
range = 23-83 / U = 589.0 ,
p = .602
Sex
Female
Male / n / % / ²(1) = 2.130 ,
p = .209
15
7 / 68.2
31.8
Education(highest level)
Elementary school
Junior high school
University entrance diploma
University degree
no graduation / 8
8
6
0
0 / 36.4
36.4
27.3
0
0 / ²(4) =2.231 ,
p = .738
cognitively impaired°
Md / range / n / %
BCSE (n = 22) / 56 / 42-58 / 4 / 18.2 / U=377.0; p=.003
MoCA (n = 22) / 28 / 26-30 / 0 / 0.0 / U=194.5; p=.000
MMSE (n = 21) / 30 / 27-30 / 0 / 0.0 / U=208.0; p=.000

BCSE = Brief Cognitive Status Exam; M = arithmetic mean; Md = Median; MMSE = Mini Mental State examination; MoCA = Montreal cognitive assessment; n = number; p = probability; SD = standard deviation; U = Test statistic for Mann-Whitney-U-Test;² = test statistic chi-square test; ³ in comparison with the patient group; °classification on basis of the manuals (MMSE cut-off: <26, MoCA cut-off <26, BCSE 15% percentile rank)

Results preliminary analysis

Analysis of feasibility:Additional results

The subjective strain differed significantly between MMSE (Md= 10, range =0-10) and MoCA (Md= 3, range= 0-8) (z=-3.838, p=.000). The subjective tolerability of MMSE had a median of 10 and a range of 0-10. The screening tools were judged as appropriateby the patients (MMSE 24.1%, n=14; MoCA 25.9%, n=15; BCSE 29.3% n=17).

Analysis of validity :Additional results

Correlation of MMSE and MoCA was slightly higher than for BCSE and both other tools (table 3). MMSE and MoCA sum scores correlated significantly with each of the five domains (r=.192 to .539). For details see table 4.

Online Resource 2 table 3: Correlational analysis of the screening tools (Kendall-Tau-b)
BCSE / MoCA
MMSE (n=78) / .475** p=.000 / .641** p=.000
MoCA (n=79) / .485** p=.000

BCSE = Brief Cognitive Status Exam; MMSE = Mini Mental State examination; MoCA = Montreal cognitive assessment; p = probability; * significant with p < .05 (two-tailed); ** significant with p < .01 (two-tailed)

Online Resource 2 table4: Correlational analysis of screening tools (sum score) and domains (mean z-score)(Kendall-Tau-b)
Domain / speech / attention and visuo-motor speed / memory / executive functions / visual-spatial functions
screening
(sum score)
BCSE / .388** / .302** / .378** / .351** / .070
(n=54) / (n=52) / (n=52) / (n=50) / p=.473 (n=53)
MoCA
.473** / .389** / .539** / .385** / .192*
(n=55) / (n=53) / (n=53) / (n=50) / (n=54)
MMSE
.447** / .395** / .499** / .333** / .213*
(n=54) / (n=52) / (n=52) / (n=50) / (n=53)

BCSE = Brief Cognitive Status Exam; MMSE = Mini Mental State examination; MoCA = Montreal cognitive assessment; n = number; *higher values indicating higher strain; °higher values indicating higher tolerability; * significant with p < .05 (two-tailed); ** significant with p < .01 (two-tailed)

Several risk factors showed significant relationship with the sum scores of the screening tools: higher age was correlated with lower sum scores in each screening tool (r=-.46 to -.36; p=.000), male patients showed lower sum scores for MoCA and BCSE (U=271.5 respective U=259.5; p=.019 respective .018), higher educational status was correlated with higher sum scores (r=.287 to .304; p=.004 to .006) and higher WHO-grade was correlated with lower sum scores (r=-.338 to -.197; p=.001 to .059). No relationship was found for lesion side (hemisphere) and premorbid IQ.

References

1.Satoer D, Vork J, Visch-Brink E et al. (2012) Cognitive functioning early after surgery of gliomas in eloquent areas. Journal of Neurosurgery 117:831–838

1