Skip to main content

Table 1 Classification variables used in decision tree analysis

From: Determine the therapeutic role of radiotherapy in administrative data: a data mining approach

Classification variables

Description

Completeness*

Curative

Palliative

Fraction size (cGy)

Median: 200

Median: 300

99%

Time from the first treatment (no. of days)

Median: 19

Median: 9

100%

Body-region group

  

91%

Chest

123,378 (91%)

12144 (9%)

Organs and tissues in pelvis region

78,220 (97%)

2211 (3%)

Pelvis – single side

1,057 (44%)

1332 (56%)

Pelvis – both sides

27,902 (91%)

2685 (9%)

Brain

9,177 (52%)

8589 (48%)

Neck

16,378 (92%)

1336 (8%)

Head

14,256 (93%)

1020 (7%)

Bone – spine, limb, chest, head

3,817 (21%)

14353 (79%)

Abdomen

7,206 (73%)

2614 (27%)

Skin

983 (73%)

357 (27%)

Other regions

7,116 (96%)

262 (4%)

Disease site group

  

97%

Head/neck(140–144,146-149,160,161)

26,090 (97%)

924 (3%)

Other head/neck

3,735 (92%)

309 (8%)

Colon/intestines (152,153)

1,782 (58%)

1,271 (42%)

Rectum (154)

17,585 (91%)

1,709 (9%)

Liver (155)

140 (38%)

226 (62%)

Other GI (150,151,156-159)

6,736 (74%)

2,349 (26%)

Lung (162–165)

18,850 (55%)

1,5319 (45%)

Bone (170)

245 (68%)

114 (32%)

Soft tissue (171)

1,942 (86%)

303 (14%)

Melanoma (172)

1,390 (54%)

1,204 (46%)

skin (173)

134(85%)

23(15%)

Breast (174,175)

97,728 (93%)

6,897 (7%)

Ovary (183)

461 (46%)

541 (54%)

Other GYN (179–182,184)

1,2791 (90%)

1,391 (10%)

Prostate/Testis/Penis (185–187)

74,061 (95%)

3,896 (5%)

Bladder (188)

2,206 (67%)

1,075 (33%)

Kidney (189)

464 (24%)

1,504 (76%)

CNS (190–192)

7,524 (93%)

604 (7%)

Thyroid/Endo (193,194)

727 (70%)

318 (30%)

Unspecified group 1 (195,196)

1,271 (88%)

176 (12%)

Unspecified group 2 (197–199)

4,152 (66%)

2,153 (34%)

Lymphoid/leukemia(200,202,204-208)

7,295 (71%)

2,933 (29%)

Hodgkin’s disease (201)

1,697 (94%)

116 (6%)

Myeloma (203)

484 (24%)

1,548 (76%)

  1. *The completeness was calculated using all records of treatment given between 2005 and 2008; while the summary statistics were based on the records used in decision tree analysis, which are randomly selected from the records without missing data.