Negative Correlation Between the Hematopoietic Stem Cell Transplant Comorbidity Index (HCT-CI) and Total Hospital Stay in Patients Who Underwent Unrelated Donor Hematopoietic Stem Cell Transplantation (HSCT)

Abstract 4462

The hematopoietic stem cell transplant comorbidity index (HCT-CI) was developed as a risk assessment tool in patients undergoing allogeneic HCT. HCT-CI scores have been shown to be associated with non-relapse mortality (NRM) and overall survival (OS) in retrospective studies. However the correlation between HCT-CI scores and post-transplant resource utilization has not been investigated. Previously, we reported a correlation between the HCT-CI and increased hospitalization in the matched related donor setting (Szwed E, et al. Blood. 2010; 116: 4748a). In the unrelated donor population, we retrospectively analyzed 229 consecutive patients who had undergone matched unrelated donor stem cell transplant between January 2001 and December 2008. Donor stem cell sources included peripheral blood (n=90), bone marrow (n=67), and umbilical cord blood (n=72). Conditioning regimens were divided into myeloablative (MA, n = 159), reduced intensity (RIC, n = 28) and non-myeloablative (NMA, n = 42). HCT-CI scores were calculated by method of Sorror (Sorror ML, et al. Blood. 2005 106: 2912–19). Median follow-up was 6.2 years. Consistent with prior findings, HCT-CI correlated with NRM (r_s = 0.16, p = 0.015) and overall survival (median survival when score ≤ 3 = 761 days vs score > 3 = 212 days [p = 0.0005]). A negative correlation was found between HCT-CI score and total hospital days as well as days spent in hospital after discharge from initial transplant hospitalization, implying that the higher the score the less time spent in the hospital after transplant (Table 1). However, this correlation disappeared when adjusted for patient survival. We also noted that the majority of readmissions and hospitalizations occurred within the first 100 days from transplant. In subgroup analysis, we found a relationship between HCT-CI score and total hospital days (r_s = 0.37, p = 0.0032) and in number of hospital days after discharge for transplant (r_s = 0.35, p = 0.0047) for patients who underwent a myeloablative cord blood transplant, even after adjusting for patient survival. In summary, our analysis validated the correlation between HCT-CI score and NRM/overall survival, but we were unable to find a statistically significant correlation between HCT-CI and hospitalizations, except in the myeloablative cord blood setting. We found that most readmissions and hospital days occur early in the post-transplant period (<100 days). We hope further analyses will help unravel these apparent discrepancies in this particular transplant population.

Statistically significant values are bold