Background: Given the elective nature of hematopoietic stem cell transplantation (HSCT), the increased mortality and the potential loss of decision making capacity, patients undergoing HSCT are encouraged to have advanced care planning (ACP). This however could result in undue anxiety for the patient and their families since going through the process of having ACP explicitly raises the possibility of death. We compared the outcomes of patients with or without ACP who received HSCT for hematologic malignancies.

Study Design: Patients age ≥19 yrs undergoing first allogeneic or autologous HSCT for a hematological malignancy between 2001 and 2003 were included in this study. Psychosocial assessments, including discussions about ACP, defined as having living will, power of attorney for health care, and life support instructions, conducted prior to transplant, were reviewed. Patients were classified according to presence or absence of ACP at HSCT. Multivariate Cox regression analysis was used to compare the risk of 100-day mortality between those with and without ACP while controlling for patient-, disease-, and transplant-related variables. Other outcomes evaluated include: days of hospitalization and in-hospital mortality.

Results: Of the 380 eligible patients, psychosocial assessments were available for 343 patients (90%). Of these, 146 either had ACP (n=138, 40%) or completed it (n=8, 2%) during the pre-transplant process, while 197 did not have it but were open to the idea of having it in the future (n=161, 47%) or did not plan to have ACP (n=36, 11%). Older patients (p <0.001), male sex (p <0.01), and patients with lymphoma versus leukemia (p<0.005) were more likely to have ACP. No differences in level of education, role of patient in the family, smoking or alcohol usage, usage of spiritual faith to cope with illness, stage of disease at transplant, type of transplant, and interval from diagnosis to transplant were noted between those with and without ACP. There was no difference in the risk of 100-day mortality (RR: 0.80; 95% CI: 0.36 – 1.81) between patients with or without ACP after adjusting for disease stage and transplant type.

Conclusions: The absence of differences in the outcomes of patients with or without ACP suggests that engagement in ACP is not reflective of HSCT risk. Factors other than outcome of HSCT need to be considered while counseling patients regarding ACP. The fact that only 40% of patients engaged in ACP prior to HSCT and 11% were not interested in ACP suggests that other factors determine interest in ACP and that a policy of discussing ACP with all HSCT patients will not result in universal acceptance of ACP.

Outcomes following HSCT based on ACP

OUTCOME WITHOUT ACP (n = 197) WITH ACP (n = 146) p VALUE 
Probability of 100-day mortality (95% CI) 8 (4–12) 7 (4–12) 0.79 
In-hospital deaths 5 (2%) 3 (2%) 0.94 
Median hospital stay in days (range) 13 (8–59) 13 (3–158) 0.7 
OUTCOME WITHOUT ACP (n = 197) WITH ACP (n = 146) p VALUE 
Probability of 100-day mortality (95% CI) 8 (4–12) 7 (4–12) 0.79 
In-hospital deaths 5 (2%) 3 (2%) 0.94 
Median hospital stay in days (range) 13 (8–59) 13 (3–158) 0.7 

Corresponding author