Ten RCTs (625 patients) were included in the review. The risk of bias of the included trials was relatively high. None of the trials was double blinded. Allocation concealment was described in three trials. Three trials conducted an intention-to-treat analysis. Missing data handling was described adequately in one trial. Duration of follow-up ranged from six to 36 months.
Glomerular filtration rate (five RCTs, 340 patients): Glomerular filtration rate was significantly higher in the calcineurin inhibitor minimisation group compared with the routine calcineurin inhibitor regimen group (WMD 9.14mL/min, 95% CI 5.85 to 12.43) with no evidence of statistical heterogeneity. The effect was significant in the mycophenolate mofetil subgroup of four trials (WMD 9.10mL/min, 95% CI 5.64 to 12.55; 257 patients), but not in the one trial that used sirolimus.
Serum creatinine level (seven RCTs, 357 patients): Serum creatinine level was significantly lower in the calcineurin inhibitor minimisation group compared with the routine calcineurin inhibitor regimen group (WMD -0.21mg/dL, 95% CI -0.35 to -0.06) with some evidence of statistical heterogeneity (Ι²=39%). The effect was significant in the mycophenolate mofetil subgroup of six trials (WMD -0.27, 95% CI -0.39 to -0.14; 341 patients), but not in the one trial that which used sirolimus.
Creatinine clearance rate (five RCTs, 285 patients): Creatinine clearance rate was not significantly different in the calcineurin inhibitor minimisation group compared with the routine calcineurin inhibitor regimen group. Subgroup analyses found no significant effect in any drug class group (mycophenolate mofetil, sirolimus or everolimus), but did approach statistical significance in the two trials that used sirolimus (WMD 10.56mL/min, 95% CI -1.40 to 22.51).
Results from the observational studies were broadly similar to those from the RCTs for glomerular filtration rate and serum creatinine level, but showed a beneficial effect of calcineurin inhibitor minimisation on creatinine clearance rate which was not apparent in the RCT data.
Safety outcomes: There was no difference in acute rejection rate (10 studies) between the minimisation and the routine regimen groups. The results for incidence of infection and patient survival differed between the text and the forest plots, so were not clear.