Risk Factors For Readmission Following Allogeneic Hematopoietic Stem Cell Transplantation and Impact On Overall Survival

Recent initiatives to improve patient safety and reduce healthcare costs have focused on preventing hospital readmissions. Historically, patients treated with allogeneic hematopoietic stem cell transplantation (HSCT) have high rates of hospital readmission. The purpose of this study was to identify the incidence and associated risk factors for readmissions in allogeneic HSCT patients and to evaluate the impact of readmissions on overall survival.

A retrospective review of patients receiving a myeloablative (MAC) or reduced intensity conditioning (RIC) HSCT at Dana Farber/Brigham and Women’s Hospital between January 1, 2005 and December 31, 2010 was performed. At our institution, RIC transplant patients are typically discharged on day +1. The 30-day readmission rate, a standard benchmark used by the Centers for Medicare & Medicaid Services, and the day 100, a traditional assessment point in transplantation, readmission rates were examined. Reasons for readmission as well as sociodemographic, disease, and HSCT-related variables were evaluated. Risk factors for readmission and the impact of readmission on overall survival were assessed by multivariate regression analysis.

A total of 1097 HSCT patients were reviewed. In the MAC group, 130 of 495 (26.3%) patients were readmitted within 30 days of discharge and 194 (39.2%) patients were readmitted by day 100 following transplantation. 74.2% of the MAC patients had one readmission by day 100. In the RIC group, 105 of 602 (17.4%) patients were readmitted within 30 days of discharge and 185 (30.7%) patients were readmitted by day 100 following transplantation. 69% of the RIC patients had one readmission by day 100. In both groups, the most frequent reasons for readmission were infection (27.6% in MAC group, 26% in RIC group), fever without a source (19.1% in MAC group, 19% in RIC group), and graft versus host disease (17.9% in MAC group, 15.1% in RIC group).

In the MAC group, a multivariate logistic regression model of the probability of being readmitted suggested that the principal risk factors for readmission by day 100 were infection during the index transplant admission (OR 1.9, p=0.0006) and Latino ethnicity (OR 4.6, p=0.013). In the RIC group, active disease at the time of HSCT (OR 2.1, p=0.0001), infection during the index admission (OR 4.8, p<0.0001), a mismatched donor (OR 2.1, p=0.030) and non-private (32.1% Medicaid, 66.4% Medicare, 1.5% other) insurance (OR 1.6, p=0.029) were significant risk factors for readmission by day 100.

In a landmark analysis of patients who survived beyond the studied time points, the 5-year overall survival (OS) for those readmitted within 30 days of discharge from the index HSCT in the MAC group was 42% compared with 56% among patients not readmitted (p=0.0026). Similarly, OS in the RIC group was 26% compared with 50% (p<0.0001). The 5-year OS for those readmitted by day 100 following HSCT in the MAC group was 52% compared with 61% among patients not readmitted (p=0.058) and in the RIC group was 26% compared with 57% (p=<0.0001). After adjusting for age, donor type, and the disease risk index (DRI), a multivariate analysis confirmed that readmission within 30 days of discharge or by day 100 was associated with decreased OS (table 1).

Infection and fever without a source were the most common causes of readmission after HSCT. In the RIC group, disease, transplant, and sociodemographic factors were associated with readmission. Being readmitted within 30 days of discharge from transplant was a significant risk factor for a lower 5-year overall survival rate in both the RIC and MAC groups. A better understanding of the risk factors for readmission in the HSCT population will allow for more transitional care and clinical resources to be focused on the highest risk patients. Strategies to decrease readmissions may improve the overall survival of patients undergoing allogeneic HSCT. More research is needed to better learn how to balance early discharge with preventable readmissions.

No relevant conflicts of interest to declare.