Twenty comparisons from 19 trials were included in the review (3,485 participants, based on figures in Table 1). There was no evidence of publication bias. Trial quality was low, with no trial assigned a Jadad score of over 3 out of 5 points. Exclusion of the two trials with a Jadad score of 1 did not alter the results for primary outcomes. Allocation concealment was reported as adequate, but intention-to-treat analysis was not presented in seven trials. Median follow-up was 12 months.
Kidney transplant function: Calcineurin inhibitor sparing improved glomerular filtration rate (WMD 4.4 mL/minute, 95% CI 2.9 to 5.9; 17 trials) and glomerular filtration rate change from baseline (WMD 4.6 mL/min, 95% CI 2.5 to 6.7; nine trials). The effect was lower, but remained significant in studies which reported intention-to-treat analyses compared to per-protocol analyses. A similar pattern to the glomerular filtration rate results was seen for the outcome of serum creatinine.
Graft survival (14 trials): There was no clear evidence of an effect of calcineurin inhibitor sparing on graft survival, either when excluding graft loss because of death or including death with a functioning graft.
Mortality (12 trials): There were no significant differences in mortality between the two groups.
Acute rejection rate (14 trials): There was significant heterogeneity between the five subgroups for the acute rejection outcome. No rejections occurred in patients with deteriorating function after calcineurin inhibitor sparing. In the elective calcineurin inhibitor elimination group, elimination was associated with an increase in biopsy-proven acute rejection (OR 2.23, 95%CI 1.57 to 3.17; six trials).
There were no consistent effects of calcineurin inhibitor sparing on other outcomes.