Premium
Pretransplant risk stratification for early survival of renal allograft recipients
Author(s) -
Kikić Željko,
Herkner Harald,
Sengölge Gürkan,
Kozakowski Nicolas,
Bartel Gregor,
Plischke Max
Publication year - 2014
Publication title -
european journal of clinical investigation
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.164
H-Index - 107
eISSN - 1365-2362
pISSN - 0014-2972
DOI - 10.1111/eci.12203
Subject(s) - medicine , confidence interval , retrospective cohort study , receiver operating characteristic , cohort , transplantation , comorbidity , gastroenterology , kidney transplantation , cohort study , surgery
Background Baseline comorbidities influence patient outcomes in renal transplantation. Identification of high‐risk recipients for patient death and early allograft loss might lead to superior stratification. Material and Methods In this retrospective study, risk stratification models were developed in a cohort of 392 kidney transplant recipients and validated in an independent cohort to predict short‐term (2 year) outcomes. Results Peripheral arterial disease [ OR 7·7 (95% confidence interval ( CI ): 2·45–24·60); P < 0·001], use of oral anticoagulation [ OR 18·68 (95% CI : 3·77–92·46); P < 0·0001], smoking [ OR 5·15 (95% CI : 1·67–15·84); P = 0·004], recipient age > 60 years [ OR 7·28 (95% CI : 2·33–22·69; P = 0·001)], serum albumin < 40 g/L [ OR 5·08 (95% CI : 1·82–14·19); P = 0·002], serum calcium ≥ 2·42 mM [ OR 6·47 (95% CI : 1·37–30·58); P = 0·02] living donation [ OR 2·95, (95% CI : 0·31–28·29); P = 0·34)] and previous haemodialysis [ OR 3·33, (95% CI : 0·39–28·11); P = 0·27)] were included in the model. The validated model discriminated between low‐ (< 3 points) and high‐risk recipients (> 8·5 points) with mortality rates of 0% vs. 54%. The comparison of the model with the Charlson comorbidity index ( CCI ) yielded significantly better receiver operating characteristic ( ROC ) areas ( N ovel S core ROC : 0·87 vs. CCI : 0·72, P = 0·0012). Early allograft loss was associated with presensitization [ OR 3·02 (95% CI : 1·29–7·09); P = 0·011] and presence of hepatitis C antibodies [ OR 2·42 (95% CI : 1·09–5·34); P = 0·029]. A risk model ( ROC : 0·62) for allograft loss could not be developed. Conclusion Risk stratification based on the novel score might identify high‐risk recipients with disproportional risk of early patient death and lead to optimized strategies.