Table 2.

Cox proportional hazards models for time to nephrology referral and visit

Nephrology Referral
Unadjusted CKD Stage 3aAdjusted CKD Stage 3bAdjusted, Death as Competing Risk, CKD Stage 3c
Non-Hispanic white
Non-Hispanic black2.97 (2.93 to 3.02)2.01 (1.97 to 2.04)1.96 (1.93 to 2.00)
Hispanic white1.54 (1.49 to 1.58)1.36 (1.33 to 1.40)1.37 (1.33 to 1.41)
Hispanic black2.24 (2.03 to 2.48)1.96 (1.77 to 2.17)1.97 (1.77 to 2.20)
Unadjusted CKD Stage 4aAdjusted, CKD Stage 4bAdjusted, Death as Competing Risk, CKD Stage 4c
Non-Hispanic white
Non-Hispanic black1.55 (1.51 to 1.59)1.33 (1.30 to 1.36)1.30 (1.28 to 1.33)
Hispanic white1.24 (1.19 to 1.29)1.18 (1.15 to 1.22)1.15 (1.12 to 1.19)
Hispanic black1.35 (1.15 to 1.58)1.26 (1.13 to 1.41)1.23 (1.09 to 1.39)
Nephrology Visit
Unadjusted CKD Stage 3aAdjusted CKD Stage 3bAdjusted, Death as Competing Risk, CKD Stage 3c
Non-Hispanic white
Non-Hispanic black3.13 (3.09 to 3.18)2.13 (2.09 to 2.16)2.09 (2.06 to 2.13)
Hispanic white1.75 (1.70 to 1.80)1.56 (1.51 to 1.60)1.56 (1.51 to 1.61)
Hispanic black2.86 (2.59 to 3.15)2.52 (2.28 to 2.78)2.50 (2.25 to 2.79)
Unadjusted CKD Stage 4aAdjusted, CKD Stage 4bAdjusted, Death as Competing Risk, CKD Stage 4c
Non-Hispanic white
Non-Hispanic black1.68 (1.64 to 1.72)1.46 (1.43 to 1.49)1.44 (1.40 to 1.47)
Hispanic white1.44 (1.38 to 1.50)1.39 (1.34 to 1.44)1.34 (1.30 to 1.39)
Hispanic black1.77 (1.53 to 2.06)1.69 (1.52 to 1.87)1.66 (1.48 to 1.86)
  • All results are reported as hazard ratio (95% confidence interval). All P values were <0.001. —, reference group.

  • a Unadjusted; death-censored.

  • b Adjusted for model 1: age, sex, entry eGFR, median BP, median body mass index, hypertension diabetes, coronary artery disease, cerebrovascular disease, congestive heart failure, depression, hemiplegia, dementia, liver disease, chronic lung disease, malignancy, alcohol abuse, drug abuse, hepatitis C, smoking status, marital status, service connection, socioeconomic status, distance to Veterans Affairs medical center, clustered by medical center, using multiple imputation for missing data; death-censored.

  • c Adjusted for model 1, addressing death as a competing risk.