Table 3.

Effect of preparation method on residual CMV in leukocyte-reduced blood products

Observed frequencyProbability95% CIOdds ratioTest
dfχ2Probability
Probability of GE > 4/mL        
 Apheresis 7/32 .22 0.04-0.68 1.0 2 1.30 .522 
 Filtered Platelet 4/32 .13 0.02-0.56 0.5    
 Filtered RBC 7/32 .22 0.11-0.39 1.0    
Probability of GE > 50/mL        
 Apheresis 4/32 .13 0.01-0.58 0.6 2 4.45 .108 
 Filtered Platelet 1/32 .03 0.00-0.41 0.1    
 Filtered RBC 6/32 .19 0.09-0.36 1.0    
Probability of GE > 80/mL        
 Apheresis 1/32 .03 0.00-0.52 0.3 2 1.57 .456 
 Filtered Platelet 1/32 .03 0.00-0.52 0.3    
 Filtered RBC 3/32 .09 0.03-0.25 1.0    
Observed frequencyProbability95% CIOdds ratioTest
dfχ2Probability
Probability of GE > 4/mL        
 Apheresis 7/32 .22 0.04-0.68 1.0 2 1.30 .522 
 Filtered Platelet 4/32 .13 0.02-0.56 0.5    
 Filtered RBC 7/32 .22 0.11-0.39 1.0    
Probability of GE > 50/mL        
 Apheresis 4/32 .13 0.01-0.58 0.6 2 4.45 .108 
 Filtered Platelet 1/32 .03 0.00-0.41 0.1    
 Filtered RBC 6/32 .19 0.09-0.36 1.0    
Probability of GE > 80/mL        
 Apheresis 1/32 .03 0.00-0.52 0.3 2 1.57 .456 
 Filtered Platelet 1/32 .03 0.00-0.52 0.3    
 Filtered RBC 3/32 .09 0.03-0.25 1.0    

The detectable presence of CMV in the leukocyte-reduced product was regressed on treatment type using logistic regression to determine the effect of apheresis, filtered platelet, and filtered RBC methods on residual CMV. The results are shown as the probability of CMV above the detection thresholds 4 (primary analysis), 50 and 80 GE/mL (secondary analyses). The results of global tests of the equivalence of treatment type are shown with the chi-squared statistic and P value (Test). There was no significant difference in leukocyte-reduction (LR) methods.

CI indicates confidence interval; df, degrees of freedom. See Tables 1and 2 for other abbrevations.

View Large
Close Modal
This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal