Skip to main content

Table 1 Characteristics of the studied populations

From: Would artificial neural networks implemented in clinical wards help nephrologists in predicting epoetin responsiveness?

  AIMS data (± SD) EOCdata (± SD)
N 340 92
Sex (% of males) 58.2 45.6
Age (y) 63.5 ± 14.6 69.7 ± 12.1
Weight (kg) 69.5 ± 15.3 69.9 ± 15.4
Diabetes mellitus (%) 27.4 34.7
Cardiomyopathy (%) 16.5 13.0a
ACE-I or ARB treatment (%) No data 89.1
Haemoglobin g/dL b 11.8 ± 1.4 11.6 ± 0.7
Creatinine μmol/L b 580 ± 156 552 ± 145
BUN mmol/L b No data 24.6 ± 6.0
Kt/V No data 1.33 ± 0.26
pH b No data 7.36 ± 0.04
Phosphate mmol/L b No data 2.00 ± 1.38
Ionized calcium mmol/L b No data 1.20 ± 0.07
Albumin g/L b No data 37.9 ± 2.9
Ferritin mg/mL 411 ± 297 482 ± 255
PTH pmol/L No data 25.8 ± 21.9
CRP mg/L b No data 15.7 ± 15.8
Iron intravenously mg/month No data 134 ± 50
Epoetin beta dose IU/Kg/week to reach a haemoglobin of 11.5 g/dL 149 ± 104 107 ± 63
Epoetin administration route % of subcutaneous 71 0
  1. a only patients with a left ventricular ejection fraction <50%; b pre-dialysis values
  2. Anthropometric data, comorbidity incidence, ACE-I or ARB treatment and mean results of the monthly biochemical parameters of the AIMS and EOC data pools.