To understand predictors and outcomes of ICU admission we conducted a retrospective study of consecutive patients undergoing allogeneic HCT over a 10 year period 1/2002 - 12/2012. Information about patient demographics, disease characteristics, disease status, co-morbidities, conditioning regimen, donor characteristics and ICU interventions were collected. Single and multivariable analysis was used to assess associations between various characteristics and outcomes. Survival probabilities were estimated using the Kaplan-Meier method.
118 patients underwent first transplant. Median age was 51 (range 20-72); 81% had leukemia. 39 (33%) patients were admitted to the ICU within the first 100 days; 22 within 14 days of transplant. Median survival of these patients was 54 days (95% CI 14-189) compared to 4.5 years for patients not admitted to ICU. In single variable analysis, TBI (p=0.04), URD (p=0.03), mismatched donor (p=0.009)*, and not in CR (p<0.001)* were associated with increased likelihood of ICU admission (*significance in multivariable analysis). HCT-CI score was not found to be associated with ICU admission.
Increased risk of death in ICU patients was associated with male sex (p<0.001), HCT-CI defined liver dysfunction (p=0.003) and vasopressor administration (p=0.020) in single variable analysis. Adjusting for sex (HR=4.63), vasopressor administration (HR 4.54, p=0.001) and number of ICU interventions within 48 hours of admission (HR 3.48 [>1 vs 0], p= 0.026) were associated with increased risk of death. APACHE score was not associated.
Compared to females, male patients had higher rates of myeloablative regimens (71% v 56%), TBI (48% v 33%), URD HCT (76% v 44%) and vasopressor use (33% v 24%) as well as lower rates of CR (19% v 28%) at time of transplant.
ICU admission was associated with an increased risk of death (HR = 3.70, 95% CI 2.26-6.06, p <0.001). ICU survivors tended to have worse long-term outcomes. Among 18 patients alive at day 90, 12 survived to one year (67%) compared to 61 out of 79 (77%) who never required ICU admission (p= 0.84). Further analysis to understand the causes of worse outcomes in males is ongoing. Poorer long-term outcomes in ICU survivors warrant validation in a larger cohort and further research to understand its mechanism and develop appropriate strategies. Our study is limited by type II and type I errors due to small sample size and multiple variables. Nonetheless, these results provide potentially important data for patient counseling and may help guide management of critical illness post-transplant.
Figure 1. Survival among patients alive at 90 days after transplant according to whether they were admitted to the ICU within 90 days after transplant.
|