Measuring reliable change in cognition using the Edinburgh Cognitive and Behavioural ALS Screen (ECAS)

Figures & data

Table 1. Mean scores and standard deviations for the ECAS-A, ECAS-B, and ECAS-C.

	ECAS-A	ECAS-B	ECAS-C
ALS Specific (max 100)	85.65 ± 8.56	86.39 ± 8.55	86.20 ± 9.49
Language (max 28)	27.14 ± 1.87	26.74 ± 2.09	26.76 ± 1.89
Fluency (max 24)	19.61 ± 2.71	19.82 ± 2.43	19.69 ± 4.22
Executive (max 48)	38.89 ± 5.59	39.82 ± 5.37	39.75 ± 5.68
ALS Non-Specific (max 36)	30.14 ± 3.73	29.89 ± 3.66	30.41 ± 3.11
Memory (max 24)	18.46 ± 3.43	18.51 ± 3.13	18.88 ± 2.89
Visuospatial (max 12)	11.68 ± 0.74	11.39 ± 1.00	11.53 ± 0.70
ECAS Total (max 136)	115.79 ± 11.26	116.28 ± 11.38	116.61 ± 11.87

Table 2. RCI thresholds for the ECAS-A, ECAS-B, and ECAS-C.

	JT method			Chelune method
	ALS Specific ΔX	ALS Non-Specific ΔX	ECAS Total ΔX	ALS Specific ΔX	ALS Non-Specific ΔX	ECAS Total ΔX
ECAS-A to ECAS-B	± 7.23	± 3.35	± 8.23	−6.50 < > 7.97	−3.60 < > 3.11	−7.74 < > 8.72
ECAS-B to ECAS-C	± 6.44	± 3.19	±7.18	−6.63 < > 6.25	−2.68 < > 3.70	−6.86 < > 7.51
ECAS-A to ECAS-C	± 6.02	± 2.97	± 6.64	−5.47 < > 6.56	−2.70 < > 3.24	−5.82 < > 7.46
Recommended for clinical use	≥ 8	≥ 4	≥ 9

ΔX is the change in score required to be considered significant. The Chelune method results in different upper and lower thresholds due to its subtraction of a constant.

Table 3. Simple linear regression equations for predicting ECAS-B and ECAS-C performance.

	R^2	SEE	C	β(ECAS-A)	X - ^X
Predicting ECAS-B from ECAS-A
ALS Specific	0.398	5.29	34.18	0.613	±8.70
ALS Non-Specific	0.297	2.25	15.31	0.490	±3.70
ECAS Total	0.448	6.23	41.62	0.648	±10.24
	R^2	SEE	C	β(ECAS-B)	X - ^X
Predicting ECAS-C from ECAS-B
ALS Specific	0.700	5.30	34.18	0.938	±8.72
ALS Non-Specific	0.597	2.00	11.60	0.628	±3.29
ECAS Total	0.741	6.12	14.78	0.872	±10.07

R² is the multiple R². SEE is the residual standard error, C is the intercept, β is the beta coefficient associated with the subscript ECAS, X - ^X is the residual (i.e. the difference between the model predicted score and the observed score). The X - ^X column indicates the number of points difference required between observed and estimated score to determine reliable difference – this is calculated as 1.645*(SEE).

Table 4. Multiple regression models to predict ECAS performance.

	R^2	SEE	C	β(ECAS-A)	β(Education)	X - ^X
Predicting ECAS-B from ECAS-A
ALS Specific	.451	5.01	28.82	.573	.587	± 8.24
ALS Non-Specific	.297	2.25	15.31	.490	–	± 3.70
ECAS Total	.483	5.97	37.94	.599	.628	± 9.82
	R^2	SEE	C	β(ECAS-B)	β(Education)	X - ^X
Predicting ECAS-C from ECAS-B
ALS Specific	.700	5.30	34.18	.938	–	± 8.72
ALS Non-Specific	.431	1.87	12.69	.472	.243	± 3.08
ECAS Total	.741	6.12	14.78	.872	–	± 10.07
	R^2	SEE	C	β(ECAS-A)	β(ECAS-B)	X - ^X
Predicting ECAS-C from ECAS-A AND ECAS-B
ALS Specific	.744	4.81	−1.28	.583	.432	± 7.91
ALS Non-Specific	.762	1.51	7.37	.473	.289	± 2.48
ECAS Total	.815	5.11	2.74	.563	.416	± 8.41
	R^2	SEE	C	β(ECAS-A)	β(ECAS-B)	β(Age)	X - ^X
ECAS-C performance from ECAS-A, ECAS-B, and age
ALS Non-Specific	.665	1.42	10.91	.473	.269	−.044	± 2.34

R² is the multiple R² when model contains one predictor and adjusted R² when model contains more than one predictor. SEE is the residual standard error, C is the intercept, β is the beta coefficient associated with the subscript, X - ^X is the residual (i.e. the difference between the model predicted score and the observed score). The X - ^X column indicates the number of points difference required between observed and estimated score to determine reliable difference – this is calculated as 1.645*(SEE).