Posttransfusion Alloimmunization Following Corticosteroid Treatment

Article Figures & Data

Table 1.

Baseline characteristics by corticosteroid exposure status

Cohort Characteristic	Overall (N = 708), n (%)	No Corticosteroids (N = 558), n (%)	Corticosteroids (N = 150), n (%)	P Value^{^a}
Age, mean (±SD), y	51.73 (±12.60)	51.78 (±12.61)	51.55 (±12.62)	.806
Sex
Female	396 (55.93)	317 (44.77)	79 (11.16)	.364
Male	312 (44.07)	241 (34.04)	71 (10.03)	.364
Race
White	521 (73.59)	406 (57.34)	115 (16.24)	.667
Black	165 (23.31)	135 (19.07)	30 (4.24)
Asian	20 (2.82)	15 (2.12)	5 (0.71)
American Indian or Alaskan	2 (0.28)	2 (0.28)	0 (0.0)
Ethnicity
Not Hispanic or Latino	544 (76.84)	425 (60.03)	119 (16.81)	.414
Hispanic or Latino	164 (23.16)	133 (18.79)	31 (4.38)	.414

↵a Statistics tests included Wilcoxon rank-sum for age, chi-square for sex and ethnicity, and Fisher exact for race.

Table 2.

First-time alloimmunization by corticosteroid exposure

Exposure Group	Negative Alloimmunization, n (%)	Positive Alloimmunization, n (%)	Fisher Exact P Value	Odds Ratio (95% CI)
No corticosteroids (n = 558)	538 (75.99)	20 (2.82)	.098	0.181 (0.004–1.149)
Corticosteroids (n = 150)	149 (21.05)	1 (0.14)	.098	0.181 (0.004–1.149)

Table 3.

First-time alloimmunization by corticosteroid classification

Corticosteroid Classification	Negative Alloimmunization, n	Positive Alloimmunization, n	Event Rate, %	Fisher Exact P Value
None (n = 558)	538	20	3.58	.351
Short acting (n = 2)	2	0	0.00
Intermediate acting (n = 37)	36	1	2.70
Long acting (n = 78)	78	0	0.00
Combination (n = 33)	33	0	0.00
Total (n = 708)	687	21	2.97

Table 4.

Firth’s penalized logistic regression for positive first-time alloimmunization

Exploratory Model	Predictors	Corticosteroid Exposure, Odds Ratio	P Value	95% CI
Model 1	Treatment only	0.264	.1182	0.029–1.041
Model 2	Treatment and age	0.264	.1144	0.029–1.041
Model 3	Treatment, age, and ethnicity	0.266	.1136	0.029–1.051