A Simple Prognostic Score to Predict Short and Long-Term Survival Following ICU Admission in Critically Ill Adult Allogeneic Hematopoietic Stem Cell Transplant Recipients: Improved Survival Following Reduced Intensity Conditioning

A retrospective analysis of data collected from 164 consecutive adult Allo-HSCT recipients admitted to ICU (with a total of 214 ICU admissions), at a single centre, University College London Hospitals NHS Foundation Trust (UCLH) between June 1996 and December 2007 (11.5 year study period) was performed. Follow-up of surviving patients was recorded until 31 March 2011.

29% of all patients undergoing Allo-HSCT at our centre during the study period required one or more ICU admission. The ICU admission rate was significantly lower for patients undergoing reduced intensity conditioning (RIC) than myeloablative conditioning (17% vs. 38%, p<0.001). The median age of RIC-Allo recipients was significantly older than patients who received myeloablative conditioning (47 vs 36 years, P<0.0001). The most frequent reasons for ICU admission were sepsis (67%) and respiratory failure (55%), not mutually exclusive. Mechanical ventilation (MV) was required in 53% of admissions, inotropic support was required in 47%. Median acute physiology and chronic health score (APACHEII) was 23 (range 0–51). Overall ICU survival for all admissions was 48% (n=214). 35 patients (21%) had more 1 or more ICU admissions. Survival by patient (measured at discharge from final ICU admission) was 32%. ICU survival was significantly better after RIC Allo-HSCT than following myeloablative conditioning (OR 3.27, p=0.023). There was no difference in ICU outcome by stem cell source (sibling vs. unrelated). Patients who did not require ventilatory support (non-invasive ventilation and/or MV) had significantly better ICU outcome (OR 12.7, p<0.001). Multivariate analysis by Cox regression revealed raised urea, inotropic support and MV as independent determinants of death on ICU. Long term survival was significantly better for patients who underwent RIC Allo-HSCT (p=0.0055) (Figure 1). The subsequent long term survival for patients who survived ICU admission was excellent; 1, 2, and 5 year survival rates were 60%, 55%, and 50% respectively (median follow-up 3 years, 7 months).A simple prognostic scoring system has been derived which is predictive of both ICU survival and long term outcome (Table 1). Patients with 2 or more of the following factors: myeloablative conditioning, MV, elevated serum urea, score 1. Patients with <2 of these factors score 0. In our patient population, this binary index has been shown to be predictive of both death on ICU (OR 7.24, p=0.001) or poorer overall survival after ICU discharge (OR 3.04, p=0.01).

Figure 1

View largeDownload slide

Overall survival for patients admitted to ICU following allo-HSCT by conditioning type.

Figure 1

View largeDownload slide

Overall survival for patients admitted to ICU following allo-HSCT by conditioning type.

Close modal

In this study we report favourable ICU survival following allogeneic HSCT and for the first time report significantly better short and long-term outcome for patients who underwent RIC Allo-HSCT compared to those treated with myeolablative conditioning regimens despite a higher median age in the patients undergoing RIC transplants. For patients surviving ICU admission, subsequent long-term overall survival was excellent at 50% at 5 years, comparing favourably with other critically ill non-surgical patients admitted to ICU. A simple prognostic score has been generated which can be used to predict outcome in critically ill patients admitted to ICU following Allo-HSCT.

No relevant conflicts of interest to declare.