Table 5.

Results of sensitivity analyses

AnalysisCKD Progression by eGFR Change Values
No Decline in Kidney
Function ( >0)Mild
(Between 0 and −1)Moderate
(Between −1 and −5)Severe (<−5)
No hospitalizations before cohort entry (n=196,322)
 Risk of hospitalizations89,873 (45.8)29,909 (15.2)56,920 (29.0)19,620 (10.0)
0.97 (0.94 to 1.01)1.001.19 (1.14 to 1.24)1.15 (1.09 to 1.22)
 Risk of readmissions20,419 (46.0)6474 (14.6)13,625 (30.7)3904 (8.8)
0.96 (0.90 to 1.04)1.001.19 (1.10 to 1.28)1.43 (1.30 to 1.58)
Including no. of hospitalizations before cohort entry as a covariate in the models (n=247,888)
 Risk of hospitalizations111,318 (44.9)36,132 (14.6)74,117(29.9)26,321 (10.6)
0.98 (0.95 to 1.01)1.001.18 (1.14 to 1.22)1.24 (1.19 to 1.30)
 Risk of readmissions33,783 (43.7)10,519 (13.6)24,781(32.1)8137 (10.5)
0.96 (0.91 to 1.02)1.001.16 (1.10 to 1.23)1.47 (1.37 to 1.58)
Including no. of eGFR measurements as a covariate in the models (n=247,888)
 Risk of hospitalizations111,318 (44.9)36,132 (14.6)74,117(29.9)26,321 (10.6)
1.03 (1.00 to 1.06)1.001.16(1.12 to 1.20)1.29 (1.24 to 1.35)
 Risk of readmissions33,783 (43.7)10,519 (13.6)24,781 (32.1)8,137 (10.5)
1.01 (0.95 to 1.07)1.001.16 (1.10 to 1.23)1.53 (1.43 to 1.64)
Restricted the cohort to patients with at least five eGFR measurements (n=172,770)
 Risk of hospitalizations72,502 (42.0)24,045 (13.9)59,719 (34.6)16,504 (9.5)
0.97 (0.94 to 1.01)1.001.13 (1.09 to 1.17)1.46 (1.40 to 4.52)
 Risk of readmissions27,235 (41.5)9013 (13.7)22,623 (34.5)6743 (10.3)
0.99 (0.93 to 1.05)1.001.20 (1.13 to 1.28)1.67 (1.55 to 1.80)
Including eGFR variability as a covariate in the models (n=226,636)
 Risk of hospitalizations100,737 (44.4)31,108 (13.7)71,951 (31.7)22,840 (10.1)
0.92 (0.90 to 0.95)1.001.10 (1.07 to 1.14)1.15 (1.10 to 1.20)
 Risk of readmissions32,363 (43.4)9968 (13.4)24,509 (32.9)7748 (10.4)
0.95 (0.90 to 1.01)1.001.15 (1.09 to 1.22)1.42 (1.32 to 1.52)
  • Data are presented as n (%) or hazard ratios (95% confidence intervals). eGFR change values are given in ml/min per 1.73 m2 per year. Models are adjusted for first eGFR value, age, race, sex, diabetes mellitus, hypertension, cardiovascular disease, hyperlipidemia, peripheral artery disease, cerebrovascular disease, chronic lung disease, hepatitis C, HIV, and dementia.