Skip to main content

Table 5 Demographic and clinical characteristics of patients grouped into 4 proposed clusters using K-means CA

From: Cluster analysis and its application to healthcare claims data: a study of end-stage renal disease patients who initiated hemodialysis

 

Cluster 1: Average to High

Cluster 2: Very High to High

Cluster 3: Average to Average

Cluster 4: Increasing Costs, High at Both Points

(n = 113)

(n = 89)

(n = 16,624)

(n = 1554)

Age (y), mean (SD)

57.6 (11.6)

55.5 (14.8)

63.9 (14.0)

56.2 (12.8)

Age (y), n (%)

 18-24

0 (0.0)

4 (4.5)

121 (0.7)

33 (2.1)

 25-34

2 (1.8)

7 (7.9)

355 (2.1)

54 (3.5)

 35-44

15 (13.3)

6 (6.7)

1026 (6.2)

156 (10.0)

 45-54

24 (21.2)

19 (21.3)

2401 (14.4)

375 (24.1)

 55-64

50 (44.2)

33 (37.1)

4652 (28.0)

609 (39.2)

 65+

22 (19.5)

20 (22.5)

8069 (48.5)

327 (21.0)

Sex, n (%)

 Male

66 (58.4)

48 (53.9)

9599 (57.7)

924 (59.5)

 Female

47 (41.6)

41 (46.1)

7025 (42.3)

630 (40.5)

Region in the United States, n (%)

 Northeast

12 (10.6)

12 (13.5)

1843 (11.1)

192 (12.4)

 North central

32 (28.3)

18 (20.2)

6084 (36.6)

444 (28.6)

 South

38 (33.6)

39 (43.8)

6354 (38.2)

625 (40.2)

 West

30 (26.5)

19 (21.3)

2235 (13.4)

286 (18.4)

 Unknown

1 (0.9)

1 (1.1)

108 (0.6)

7 (0.5)

Health insurance type, n (%)

 FFS

87 (77.0)

71 (79.8)

13,967 (84.0)

1237 (79.6)

 HMO and POS capitation

20 (17.7)

17 (19.1)

2304 (13.9)

270 (17.4)

 Missing

6 (5.3)

1 (1.1)

353 (2.1)

4 (3.0)

Comorbidity Score Indicesa

Pre-HD Period

Post-HD Period

Pre-HD Period

Post-HD Period

Pre-HD Period

Post-HD Period

Pre-HD Period

Post-HD Period

 ECI, mean (SD)

6.9 (3.4)

10.8 (3.5)

9.0 (4.1)

9.3 (3.7)

5.7 (2.5)

6.8 (2.8)

6.5 (3.0)

9.5 (3.2)

 CCI, mean (SD)

5.0 (2.7)

7.1 (2.4)

5.6 (3.2)

6.2 (2.9)

4.6 (2.2)

5.1 (2.3)

5.0 (2.5)

6.5 (2.7)

  1. FFS fee-for-service, HD, hemodialysis, HMO health maintenance organization, PPS, point-of-service, ECI elixhauser comorbidity index, CCI charlson comorbidity index, SD standard deviation. aIdentification was based on non-rule out diagnosis