Table 3.

Progressively more comprehensive multivariable models for the associations of race/ethnicity with mortality (n = 6677)a

ModelHR for Mortality (95% CI Reference Category: Non-Hispanic White Patients [n = 3342])PbVariables Testedc
Hispanic Ethnicity (n = 673)Black (n = 2278)Asian (n = 231)Native American (n = 56)Other (n = 97)
1. Unadjusted0.72 (0.61 to 0.85)0.63 (0.57 to 0.71)0.68 (0.50 to 0.93)0.62 (0.35 to 1.11)0.74 (0.52 to 1.07)<0.001Race/ethnicity
2. Demographics0.88 (0.75 to 1.04)0.74 (0.66 to 0.83)0.73 (0.53 to 1.00)0.78 (0.43 to 1.42)0.97 (0.66 to 1.42)<0.001Model 1 plus gender, cause of ESRD, age, calendar year
3. Social characteristics0.88 (0.75 to 1.04)0.74 (0.66 to 0.83)0.73 (0.53 to 1.00)0.78 (0.43 to 1.42)0.97 (0.66 to 1.42)<0.001Model 2 plus marital status, living arrangements, level of education
4. Comorbidity0.93 (0.78 to 1.11)0.79 (0.70 to 0.89)0.86 (0.63 to 1.18)0.84 (0.46 to 1.56)1.10 (0.74 to 1.63)0.007Model 3 plus CVD, CHF, cancer, GI illness, neurologic illness, pulmonary illness, psychiatric illness, DNR status, functional status, vascular access
5. Nutritional indicators0.86 (0.72 to 1.03)0.87 (0.79 to 0.99)0.81 (0.57 to 1.17)0.94 (0.52 to 1.70)0.96 (0.62 to 1.47)0.27Model 4 plus albumin, creatinine, weight
6. Medical events0.84 (0.71 to 1.01)0.86 (0.76 to 0.98)0.81 (0.56 to 1.19)0.99 (0.53 to 1.83)0.94 (0.60 to 1.47)0.18Model 5 plus hospitalized days, outpatient procedures, vascular access procedures
7. Laboratory measures0.84 (0.70 to 1.01)0.96 (0.84 to 1.09)0.83 (0.57 to 1.21)0.96 (0.52 to 1.79)0.92 (0.58 to 1.45)0.52Model 6 plus bicarbonate, calcium-phosphorus product, ferritin, hemoglobin, potassium, TSAT, WBC
8. HD measures0.86 (0.72 to 1.03)0.97 (0.85 to 1.11)0.82 (0.56 to 1.20)0.95 (0.52 to 1.73)0.95 (0.60 to 1.50)0.66Model 7 plus HD treatment time (delivered), SBP, UF
  • a Models by Cox proportional hazards analysis, using ESRD vintage as time scale. Patients shown had complete covariate characterization for final multivariable model (model 8). CVD, atherosclerotic cardiovascular disease; CHF, congestive heart failure; CI, confidence interval; DNR, do not resuscitate; GI, gastrointestinal; HR, hazard ratio.

  • b For race/ethnicity, by the Wald test.

  • c Variables tested had P ≤ 0.20 for univariable association with mortality. Other than ESRD vintage (time scale for model) and BMI and Kt/Vurea (excessive missing observations), all variables in Table 2 were tested. Values tested were the most recent values recorded by the DOPPS for age, nutritional indicators, laboratory measures, and HD measures; the total number over the most recent 3 mo for medical events; and enrollment values for all other variables. Other than race/ethnicity, variables with P > 0.10 were sequentially excluded from the multivariable models. Variables excluded are indicated by italics. All other variables were included in the final multivariable model (model 8), and all but one of these (cause of ESRD, P = 0.02) had P ≤ 0.01 for the adjusted association with mortality. In ancillary models, BMI and Kt/Vurea did not meaningfully influence the associations of race/ethnicity with mortality.