How do we estimate survival? External validation of a tool for survival estimation in patients with metastatic bone disease—decision analysis and comparison of three international patient populations

Table 2 Comparison of categorical features for training (U.S.), first validation (Scandinavian) and second validation (Italian) datasets

Feature		Training set n = 189		Scandinavian set n = 815			Italian set n = 287
		No.	%	No.	%	% missing	No.	%	% missing	vs. training set p	vs. Scandinavian set p
Gender	male	85	45.0	369	45.3	0	120	42	0	0.50	0.31
Gender	female	104	55.0	446	54.7		167	58
Oncologic diagnosis grouping	1.0	52	27.3	173	21.3	0.4	63	23	2	0.42	0.007*
	2.0	34	18.2	74	9.2		44	16
	3.0	103	54.5	567	69.1		173	62
Organ metastases	yes	114	60.3	325	39.8	6.3	91	36	12	0.0001*	0.08
Organ metastases	no	75	39.7	441	53.9		161	64
Lymph node metastases	yes	36	18.8	169	20.8	61.6	96	40	16	0.0001*	0.0007*
Lymph node metastases	no	153	81.2	143	17.6		146	60
Skeletal metastases	solitary	55	29.0	123	15.2	3.4	139	49	1	0.0001*	0.0001*
Skeletal metastases	multiple	134	71.0	666	81.4		144	51
Pathologic fracture status	yes	84	44.2	614	75	0.9	143	52	5	0.08	0.0001*
Pathologic fracture status	no	105	55.8	196	24.1		131	48
ECOG performance status	0,1,2	93	49.2	558	68.3	0	123	54	20	0.39	0.0001*
ECOG performance status	3,4	96	50.8	257	31.7		106	46
Survival >3 months	yes	129	68.3	557	68.2	0	267	93	0	0.0001*	0.0001*
Survival >3 months	no	60	31.7	258	31.8		20	7
Survival >12 months	yes	79	41.8	241	29.8	0	181	63	0	0.0001*	0.0001*
Survival >12 months	no	110	58.2	574	70.2		106	37

Abbreviations: ECOG Eastern Cooperative Oncology Group, % missing, the proportion of unknown or missing data within the validation set
*Proportions are significantly different between training and validation sets by Chi-square method

ISSN: 1471-2407