Abstract

We tried to find an appropriate method to determine a hierarchy of risk factors for post kidney transplant diabetes mellitus (PKTDM), taking in account two criteria in order to evaluate the risk. We applied a multi-criteria TOPSIS method on a data set for 177 patients who underwent kidney transplantation at the Centre Hospitalier Universitaire de Nancy. We compared the results obtained by using the TOPSIS method with the hierarchy gave by relative risk (Spearman correlation coefficient r=0.91, p=0.000001). In the top of the TOPSIS method hierarchy was Solumedrol treatment. The others risk factors for PKTDM: diabetes familial antecedents on the second place, CMV infection in the third place. The TOPSIS method ranked the significant risk factors: solumedrol treatment (p<0.001), graft rejection (p<0.001), CMV infection (p=0.014), diabetes familial antecedents (p=0.016), diuretics and beta-blockers treatments (p=0.02), male sex (p=0.02) in the first’s places. The protection factor for PKTDM: the maintenance treatment (RR=0.75 [0.62, 0.9], p=0.0016) was ranked in the last place 24th in the hierarchy and the induction treatment (RR=0.77, [0.57, 1.04], p=0.04) on the 23rd place. In conclusion, the TOPSIS method seems to be an adequate method for ranking the risk factors when there are multiple criteria.

Keywords

TOPSIS method, Post kidney transplant diabetes mellitus, Risk factors ranking.