Multivariate adjusted analyses
After adjusting for transplant-associated variables, race was not found
to be associated with patient reports of QoL post-HSCT (Table 2).
Inclusion of socioeconomic variables in these analyses, such as income
(Table 3) and insurance type (Table 4) did not result in significant
associations between patient race and QoL scores.
Unlike patient race, patient diagnosis associated with physical
functioning scores, indicating that patients with non-malignant
conditions were likely to have greater physical functioning than
patients with malignant conditions (p = 0.039). This effect
persisted when analyses evaluated potential interactions of race and
estimated household income (Table 3, p = 0.04), but was not
significant in analyses highlighting the interaction of race and
insurance type (Table 4).
In supplementary analysis, White patients were used as the control group
(exponentiated coefficient = 1) to be able to probe more specific
two-group analyses. Hispanic, Black, and Native American patients did
not have QoL scores that differed from those of White patients across
all five types of functioning when controlling for transplant
characteristics (Supplemental Table S2) and socioeconomic factors
(Supplemental Table S3).