Table 5. Inclusion frequencies of single variables and top 10 models out of the 5000 bootstrap-generated data sets.

	Inclusion frequency of single variables (%) a
	20.3	7.5	25.7	6.5	47.7	12.6	17.9	6.7	71.1	29.4	33.9	6.4	40.0	89.2	93.4
Top 10 Models b	ASE	Gen	Edu	MS	ECOG	SKR	ITI	TWY	CD	TFCA	PCS	MCS	ATRS	GSS	IDTQ	% c
1									√					√	√	5.9
2					√				√					√	√	3.0
3					√				√		√		√	√	√	2.5
4			√		√				√					√	√	2.2
5									√		√		√	√	√	2.0
6							√		√					√	√	1.9
7			√						√					√	√	1.8
8									√	√				√	√	1.8
9	√													√	√	1.5
10									√	√	√		√	√	√	1.4

Abbreviations: ASE=age at study entry; ATRS=additional treatment-related symptoms, i.e., overall mean symptom score (nausea, diarrhoea, oedema, skin problems, abdominal discomfort, musculoskeletal pain, headache, muscle cramps); CD=concomitant drug burden (ref. no); ECOG=Eastern Cooperative Oncology group performance status (ref. 0); edu=education (ref. high); gen=gender (ref. male); GSS=global social support; IDTQ=desire for more information on impact of disease and therapy on quality of life (ref. no); ITI=intolerance to imatinib (ref. no); MCS=mental health; MS=marital status (ref. married); PCS=physical health; SKR, Sokal risk (ref. low); TFCA=time from complete cytogenetic response to adherence evaluation; TWY=toxicities within 1 year at study entry (ref. no).

^aThis percentage refers to the number of times a single variable was selected as an independent factor in multivariate analysis out of the 5000 bootstrap-generated samples.

^bTop 10 models out of the 5000 bootstrap-generated samples. An empty box means the variable was not selected in the model and the following symbol: √ means the variable was selected in the model.

^cThis percentage refers to the number of times a given model was selected out of the 5000 bootstrap-generated samples.