Table 3.
Risk of KFRT comparing rosuvastatin use versus atorvastatin use, overall and across eGFR levels
| Group | Unweighted No. of Events/N | IPTW-IR (95% CI), per 1000 PYs | IPTW-IRD (95% CI), per 1000 PYs | P for Heterogeneitya | IPTW-HR (95% CI) | P for Heterogeneitya | ||
|---|---|---|---|---|---|---|---|---|
| Rosuvastatin | Atorvastatin | Rosuvastatin | Atorvastatin | |||||
| Overall eGFR (ml/min per 1.73 m2) |
464/152,101 | 2190/795,799 | 0.92 (0.82 to 1.03) | 0.80 (0.76 to 0.83) | 0.12 (0.02 to 0.23) | 1.15 (1.02 to 1.30) | ||
| ≥60 | 125/130,506 | 568/687,461 | 0.27 (0.22 to 0.34) | 0.24 (0.22 to 0.26) | 0.034 (−0.03 to 0.10) | 0.31 | 1.14 (0.90 to 1.44) | 0.71 |
| 30–59 | 171/20,427 | 827/102,392 | 2.54 (2.14 to 3.03) | 2.40 (2.24 to 2.57) | 0.14 (−0.33 to 0.61) | 1.06 (0.88 to 1.28) | ||
| <30 | 168/1168 | 795/5946 | 60.9 (50.9 to 73.5) | 52.1 (48.5 to 56.0) | 8.8 (−2.9 to 20.5) | 1.21 (0.99 to 1.47) | ||
IPTW-HRs were from stratified Cox proportional hazards regression models by cohort. IPTW, inverse-probability of treatment weight; IR, incidence rate; PYs, person-years; IRD, incidence rate difference.
a
P for heterogeneity in IRD across eGFR subgroups was estimated using fixed-effects meta-analysis and P for heterogeneity in HR was estimated using stratified Cox models with interaction term between rosuvastatin use and eGFR category.