Skip to main content
. 2018 Jul 28;21(1):69–85. doi: 10.1093/biostatistics/kxy032

Table 2.

Patient characteristics and hospital admissions

Characteristics Beginning of follow-up End of follow-up Hospital admissions Unadjusted relative intensity
Gender 488   488   525      
 M 216 (44.3) 216 (44.3) 228 (43.4) 0.98 (0.69, 1.40)
 F 272 (55.7) 272 (55.7) 297 (56.6) 1.00 Reference
Race 488   488   525      
 Black 174 (35.7) 174 (35.7) 265 (50.5) 1.85 (1.30, 2.63)
 White 314 (64.3) 314 (64.3) 260 (49.5) 1.00 Reference
Age 488   488   525      
 Mean, SD 56.0 15.1 60.9 15.0 60.3 15.7    
 Min, Max 21.0 89.0 26.0 94.0 24.0 92.0    
 21–44 104 (21.3) 66 (13.5) 87 (16.6) 1.00 Reference
 45–54 127 (26.0) 94 (19.3) 99 (18.9) 0.90 (0.50, 1.62)
 55—64 116 (23.8) 137 (28.1) 115 (21.9) 0.85 (0.48, 1.53)
 65–74 78 (16.0) 92 (18.9) 100 (19.0) 1.21 (0.65, 2.28)
 75+ 63 (12.9) 99 (20.3) 124 (23.6) 1.48 (0.83, 2.64)
ZIP area 488   488   525      
 Urban 378 (77.5) 378 (77.5) 454 (86.5) 1.90 (1.19, 3.05)
 Suburb 110 (22.5) 110 (22.5) 71 (13.5) 1.00 Reference
Insurance 488   488   525      
 Government 65+ 141 (28.9) 191 (39.1) 224 (42.7) 1.00 Reference
 Government Inline graphic65 134 (27.5) 138 (28.3) 208 (39.6) 1.08 (0.73, 1.59)
 Commercial Inline graphic65 192 (39.3) 143 (29.3) 71 (13.5) 0.31 (0.20, 0.48)
 Other Inline graphic65 21 ( 4.3) 16 ( 3.3) 22 ( 4.2) 0.77 (0.45, 1.32)
Insulin 488   488   525      
 Yes 191 (39.1) 206 (42.2) 391 (74.5) 4.56 (3.35, 6.21)
 No 297 (60.9) 282 (57.8) 134 (25.5) 1.00 Reference
Metformin 488   488   525      
 Yes 176 (36.1) 176 (36.1) 168 (32.0) 0.76 (0.53, 1.08)
 No 312 (63.9) 312 (63.9) 357 (68.0) 1.00 Reference
AInline graphic(%) 389   416   488      
 Mean, SD 7.1 1.7 7.1 1.6 6.9 1.7    
 Min, Max 4.4 18.5 4.5 15.6 4.1 16.4    
Inline graphic9 355 (91.3) 366 (88.0) 437 (89.5) 1.00 Reference
Inline graphic9 34 ( 8.7) 50 (12.0) 51 (10.5) 1.25 (0.83, 1.88)
 Missing 99   72   37      
Calcium(mg/dL) 415   435   520      
 Mean, SD 9.3 0.5 9.4 0.6 8.9 0.7    
 Min, Max 7.2 10.9 7.2 11.6 6.2 11.6    
Inline graphic9.1 151 (36.4) 125 (28.7) 334 (64.2) 7.35 (5.67, 9.51)
Inline graphic9.1 264 (64.6) 310 (71.3) 186 (35.8) 1.00 Reference
 Missing 73   53   5      
PVD 488   488   525      
 Yes 172 (35.2) 272 (55.7) 488 (93.0) 16.16 (11.19, 23.34)
 No 316 (64.8) 216 (44.3) 37 ( 7.0) 1.00 Reference
PUD 488   488   525      
 Yes 239 (49.0) 333 (68.2) 480 (91.4) 7.49 (5.09, 11.01)
 No 249 (51.0) 155 (31.8) 45 ( 8.6) 1.00 Reference

We identified a cohort of patients suffering diabetes via EHR to determine which covariates are related to the risk of hospital admission. In the first column, we provide a selection of the covariates. In the second (third) column, we summarize these covariates at the beginning (end) of follow-up where the unit of analysis is the subject. In the fourth column, we summarize these covariates for hospital admissions where the unit of analysis is an admission. In the fifth column, we provide unadjusted relative intensities (and 95% CI) computed via a counting process Cox model.