The importance of analyzing graft and patient survival by cause of failure: An example using pediatric small intestine transplant data

Background. Although graft and patient survival are vital in reporting overall results of clinical transplant studies, these outcomes do not account for distinct types of graft failure and death, which clearly exist in pediatric small intestine transplantation (Itx). The use of a cause-specific hazard (CSH) approach may provide more precise identification and thus greater insight as to why certain factors are prognostically important. Methods. Among 119 pediatric patients who received primary ltx at our center since 1994, Cox model stepwise regression analyses were performed to identify prognostic factors for the following CSH rates: intestinal graft failure (IGF)/death due to rejection, death due to infection not triggered by IGF, and intestinal graft loss/death due to other causes. Results. Two factors were associated with a significantly higher rate of developing IGF due to rejection (23 such failures): receiving an isolated intestine or liver-intestine transplant (P=0.00001) and receiving no induction agent (P=0.006). Conversely, age at transplant < 1 year was the single factor associated with a significantly higher death rate due to infection (P=0.0005) (21 such deaths). Two characteristics were associated with a significantly higher death rate due to other causes: being in the hospital pretransplant (P=0.007) and not receiving daclizumab induction therapy (P=0.02) (24 such deaths). Although these four factors (transplant type/age/hospital status/induction therapy) were, for the most part, associated with graft/patient survival, the CSH analysis more precisely identified their prognostic value and achieved greater statistical power. Conclusions. A CSH approach should be used in conjunction with overall Outcome analyses.