Table 1.
Calculations for time to dialysis (eGFR of 10 ml/min/1.73 m2) in placebo and canagliflozin cohorts of the CREDENCE trial [25, 33]
| Calculations | Placebo cohort | Canagliflozin cohort |
|---|---|---|
| Base case | ||
| eGFR at week 3 = baseline eGFR + acute slope, ml/min/1.73 m2 | 56 + (− 0.55) = 55.45 | 56 + (− 3.72) = 52.28 |
| Time to dialysis = (10 − eGFR at week 3)/chronic slope | (10 − 55.45)/(− 4.59) = 9.90 years | (10 − 52.28)/(− 1.85) = 22.85 years |
| Delay in time to dialysis = canagliflozin time to dialysis − placebo time to dialysis | 22.85 − 9.90 = 12.95 years | |
| Lower bound (within 95% CIs, least steep slopes for placebo; steepest slopes for canagliflozin) | ||
| eGFR at week 3 = baseline eGFR + acute slope 95% CI, ml/min/1.73 m2 | 56 + (− 0.06) = 55.94 | 56 + (− 4.21) = 51.79 |
| Time to dialysis = (10 − eGFR at week 3)/chronic slope | (10 − 55.94)/(− 4.32) = 10.63 years | (10 − 51.79)/(− 2.10) = 19.90 years |
| Delay in time to dialysis = canagliflozin time to dialysis − placebo time to dialysis | 19.90 − 10.63 = 9.27 years | |
| Upper bound (within 95% CIs, steepest slopes for placebo; least steep slopes for canagliflozin) | ||
| eGFR at week 3 = baseline eGFR + acute slope 95% CI, ml/min/1.73 m2 | 56 + (− 1.04) = 54.96 | 56 + (− 3.23) = 52.77 |
| Time to dialysis = (10 − eGFR at week 3)/chronic slope | (10 − 54.96)/(− 4.86) = 9.25 | (10 − 52.77)/(− 1.60) = 26.73 |
| Delay in time to dialysis = canagliflozin time to dialysis − placebo time to dialysis | 26.73 − 9.25 = 17.48 years | |
CI confidence interval, eGFR estimated glomerular filtration rate