Table 4. Growth factor parameter estimates for three-class conditional models for FACT-G(Ch) subscales.

		Intercept			Linear Slope
	Mean	SE	p-value	Mean	SE	p-value
Physical subscale# (Possible range 0–28) (Normative data: US general population mean 22.7 [44]; German general population mean 24.9 [45])
High-deteriorating	23.926	0.613	<.001	−1.375	0.146	<.001
Recovery	13.843	1.317	<.001	0.896	0.204	<.001
High-Stable	24.561	0.268	<.001	−0.037	0.048	0.445
Emotional subscale# (Possible range 0–20) (Normative data: US general population mean 19.9 [44]; German general population mean 19.5 [45])
High-stable	14.775	0.195	<.001	0.205	0.030	<.001
High-deteriorating	13.027	1.405	<.001	−0.543	0.315	0.084
Recovery	6.384	2.246	.004	0.835	0.362	0.021
Social/family subscale# (Possible range 0–28) (Normative data: US general population mean 19.1 [44]; German general population mean 20.2 [45])
High-stable	22.071	0.500	<.001	0.159	0.085	.060
High-deteriorating	20.676	0.857	<.001	−0.368	0.118	.002
Recovery	15.543	1.633	<.001	0.995	0.141	<.001
Low-stable	14.051	1.093	<.001	0.308	0.210	0.143
Functional subscale# (Possible range 0–28) (Normative data: US general population mean 18.5 [44]; German general population mean 21.4 [45])
High-stable	17.993	0.525	<.001	0.093	0.076	0.217
Low-decline	12.675	0.502	<.001	−0.266	0.131	0.043

SE, standard error;

^#Higher scores indicate better quality of life.