Table 4.
Clinical score predicting the probability of obtaining a diagnosis
| Variable | β-coefficient |
|---|---|
| Intercept | −0.622792 |
| Age (in years) | |
| If current CKD stage I or II | −0.052247 |
| If current CKD stage III, IV, or V | −0.041056 |
| First-degree history of kidney disease | 1.346556 |
| Current diagnosis on the basis of clinical presentation | |
| Vascular disease | 0 |
| Glomerular disease | 0.398665 |
| Tubulo-interstitial disease | 0.900757 |
| Undiagnosed/isolated hematuria or proteinuria/other disease | 0.389117 |
| Developmental disorders/nonclassic cystic disease | 1.753048 |
For a given patient, score value is equal to the sum of each predictor (β-coefficients x variable) + intercept. Estimated probability is then equal to 1/(1+e-score).
For instance, for a 23-year-old patient with a CKD stage III, with no first-degree history of kidney disease, presenting with a tubulo-interstitial disease, the score value is: −0.622792 + 23 x (−0.041056) + 0.900757 = −0.666323. The probability of obtaining a diagnosis is then equal to 1/(1+e−(−0.666323)) = 33.9%.