Clinical Study
Comparison of a Bayesian Network with a Logistic Regression Model to Forecast IgA Nephropathy
Table 1
Characteristics of 149 patients with analyzable renal biopsy specimens between January 2002 and December 2009.
| Age (years) | Mean ± SEM | | | Age < 40 | 36% | | 40 ≤ age < 60 | 34% | |
Age ≥ 60 | 30% | Male | | 64% | History of hypertension | | 44% | Microhematuria | | 45% | Gross hematuria | | 17% | Family history of hematuria | | 3% | History of diabetes | | 11% | Mean eGFR (MDRD mL/min/1.73 m²) | Mean ± SEM | | Stage of renal failure | eGFR ≤ 29 mL/min/1.73 m² | 15% | | 30 ≤ eGFR ≤ 59 mL/min/1.73 m² | 31% | | 60 mL/min/1.73 m² ≤ eGFR | 54% | Serum Ig A | Increased > 3.6 g/L | 18% | | Normal ≤ 3.6 g/L | 34% | | Not performed | 48% | Proteinuria (g/24 h) | | 3.44 + 0.43 | | Proteinuria < 0.3 g/24 h | 12% | | 0.3 g/24 h ≤ proteinuria < 1 g/24 h | 20% | | 1 g/24 h ≤ proteinuria < 3 g/24 h | 34% | | Proteinuria ≥ 3 g/24 h | 34% |
|
|