Transplant Trial Watch

Risk prediction models for graft failure in kidney transplantation: a systematic review.

Kabore R, Haller MC, et al.

Nephrology Dialysis Transplantation 2017; 32(suppl 2): ii68-ii76.


Aims
To systematically review the risk prediction models that have been developed and validated in the past ten years for graft failure in kidney transplantation recipients.

Interventions
The databases Scopus and PubMed were searched between 1 January 2005 and 31 December 2015 for studies published in English, German or French that developed and validated a risk prediction model to predict graft failure in kidney transplant recipients, as well as studies that validated an existing model with or without model updating. One reviewer screened retrieved articles by title and abstract and two reviewers assessed the full text of each for eligibility of inclusion. Final discussion with a third reviewer occurred in any case of discordance between the two reviewers.

Participants
39 studies were included in the systematic review, of which 34 developed and validated a new risk prediction model and 5 validated an existing one.

Outcomes
Measured outcomes included donor and recipient age, donor type, cause of kidney failure, cold ischemia time, dialysis duration prior to kidney transplantation, body mass index of the donor and the recipient, dialysis or retransplantation and death.

Follow-up
1 year

CET Conclusions
This useful systematic review summarises published articles providing risk prediction models for renal transplant graft failure. The authors identify 39 articles describing or validating 34 different risk prediction models. Parameters used in the models varied greatly, including those available at the time of transplant and those only available post-transplant. Quality was variable. In particular, calibration was poorly reported (i.e. the agreement between observed and predicted risk), and many studies failed to account for competing risks such as death with a functioning graft. This summary provides a good starting point for those considering the development of future risk prediction models.

Quality notes
Quality assessment not appropriate

Trial registration
None

Funding source
Non-industry funded