Table 2

Adjusted all-cause hospitalization 1-year after baseline
Any Hospitalization (n=1,388) Total Number of Hospitalizations (n=665)3 Total Hospital Length of Stay (n=665)3
OR β β
Dialysis setting: Non-VA (ref: VA) .35*** (.24, .51) –.16* (.08) –.37* (.15)
Dual .99 (.70, 1.40) .01 (.07) –.05 (.15)
Age 1.00 (.99, 1.01) –.01* (.003) –.01 (.004)
Male .96 (.36, 2.57) –.08 (.25) .01 (.42)
Race: Non-White (ref: White) .90 (.68, 1.18) –.07 (.06) –.17 (.13)
 Unknown .20** (.07, .57) –.06 (.27) –.86* (.39)
Marital status: Unmarried (ref: married) 1.15 (.89, 1.49) .05 (.05) –.17 (.13)
 Unknown .64 (.26, 1.56) .15 (.20) –.15 (.32)
Copayment required .90 (.55, 1.48) –.11 (.11) .33 (.37)
Diagnostic Cost Group4 1.22*** (1.17, 1.27) .03** (.01) –.01 (.02)
History of dialysis prior to baseline: 6 months (ref: none) .62* (.39, .99) –.15 (.10) –.18 (.14)
 1 year .27*** (.16, .47) –.13 (.12) –.15 (.26)
 2 years .43*** (.33, .57) .01 (.06) .08 (.11)
History of hospitalization 1-year prior to baseline 2.13*** (1.65, 2.74) .28*** (.06) .28* (.13)
Distance to nearest VA dialysis, miles .999 (.997, 1.002) –.0005 (.0004) –.0005 (.001)
Region: VISN B (ref: VISN A) 1.59** (1.18, 2.13) .08 (.06) –.14 (.13)
C-statistic 0.79

Notes:

1. * p< .05 **p< .01 ***p<.001.

2. Statistics reported in parens reflect robust 95% Confidence intervals for odds ratio and standard errors for beta coefficients.

3. Total hospitalizations and hospital lengths of stay were modeled with the subset of patients who experienced any hospitalization (n= 655).

4. Results from a restricted model using DCG comorbidity adjustment are shown here. Models that included individual comorbidity indicators yielded similar results.

Wang et al.

Wang et al. BMC Health Services Research 2013 13:26   doi:10.1186/1472-6963-13-26

Open Data