Introduction: Several countries have implemented their own kidney allocation policies to ensure the equity, transparency and patient benefit in the waiting list. In Catalonia, a new allocation model has been implemented during 2017, which uses a score developed to prioritize the recipients. Four analyses were designed to evaluate that score:
1. Concordance retrospective analysis: would that score have chosen the same recipient as the clinician did in the past?
2. Impact retrospective analysis: how would the recipient profile have changed if the new allocation model were applied in the past?
3. Prospective analysis: Has the concordance changed after providing the score only for information? Which are the causes for not choosing the recipients with higher scores?
4. Evaluation analysis: Has the concordance and patient profile changed after the new allocation model?
Methods: Data from the registries of the Catalan Transplant Organization were used. The effective kidney offers from January 2014 to June 2016 and the daily recipients active on the deceased donor renal waiting list were considered for retrospective concordance (n=955) and impact analysis (n=1.046). The effective kidney offers from December 2016 to March 2017 were used for the prospective analysis (n=98). The recipients from the beginning of the new model (12th June 2017) to 31st October 2017 were used to evaluate it (n=208)*.
Results: RIn the concordance analysis, 282 (29.5%) kidney transplants (KT) were performed to a patient within the top 25th percentile values of the score (concordance). In the impact analysis, comparing the 1.046 KT performed with the 1.046 theoretical KT that would be performed using the new allocation model, we observe an increase of mean time on dialysis (from 36.3 to 57.2 months) and cpra I+II mean (from 51.6% to 64.9%), a decrease of the mean age from 59.2 to 57.9 years and a reduction of first KT from 89.2% to 76.2%. In the 98 KT studied in the prospective analysis, concordance increased up to 65.3%. The main causes for not choosing the recipients with higher values were the disagreement with the candidate (50.3%), immunologic causes (13.2%) and mistakes with the status of the recipient (21.9%). After the implementation of the new model the concordance remained at 65.8%. The mean time on dialysis was 53.5 months, the mean cpra I+II% was 33.8 and the mean age was 57,4 years.
Conclusion: Although the concordance between the clinician and the developed score was initially low, it increased significantly after providing the waiting list sorted by that score. Changes in the recipient profile observed in the impact analysis were the expected and desired ones. Finally, after the implementation we observe the expected increase of time on dialysis and younger recipients meanwhile the mean of cpra I+II % was lower than expected.

* The results will be updated using recipients up to March 2018 in the presentation at 27th International Congress of The Transplantation Society.