Skip to main content

Table 2 Unadjusted and adjusted hazard ratios (HR) associated with eGFR at initial dialysis

From: Timing of commencement of maintenance dialysis and mortality in young and older adults in Singapore

aRegression model

eGFR Category

(ml/min/1.73m2)

Number of overall patients

Number (incidence%) of death

HR (95% CI)

Æ—P value

HR

Æ—P value

Trend

Model 1

      

Late (<5)

1709

581 (34.0)

1.00

 

<0.001

Intermediate (5–10)

1359

635 (46.7)

1.54 (1.37–1.72)

<0.001

 

Early (≥10)

218

132 (60.6)

2.47 (2.04–2.99)

<0.001

 

Model 2

      

Late (<5)

1696

580 (34.2)

1.00

 

<0.001

Intermediate (5–10)

1347

628 (46.6)

1.41 (1.26–1.59)

<0.001

 

Early (≥10)

212

128 (60.4)

2.14 (1.77–2.60)

<0.001

 

Model 3

     

<0.001

Late (<5)

1395

359 (25.7)

1.00

  

Intermediate (5–10)

1018

383 (37.6)

1.30 (1.12–1.51)

<0.001

 

Early (≥10)

127

61 (48.0)

1.75 (1.31–2.32)

<0.001

 

bModel 4

      

Late (<5)

1183

297 (25.1)

1.00

 

<0.001

Intermediate (5–10)

867

315 (36.3)

1.23 (1.04–1.46)

0.015

 

Early (≥10)

98

46 (46.9)

1.91 (1.38–2.65)

<0.001

 
  1. Model 1: Univariate
  2. Model 2: eGFR adjusted for age gender, ethnicity and education
  3. Model 3: Model2+ smoking, diabetes, hypertension, cerebrovascular disease, ischemic heart disease, peripheral vascular disease, malignancy, Hepatitis B Ag, Anti-Hepatitis C, modality of dialysis and albumin
  4. Model 4: Model3+ haemoglobin, ferritin, transferrin saturation (TSAT), phosphate and intact parathyroid hormone (iPTH),
  5. aModels employed hierarchical Cox regression analyses. N in models 1 to 4 varied according to missing variables in each model.
  6. bModel4 explored mediating effects of laboratory parameters.
  7. Æ—P values were derived from Wald test.