Skip to main content
. 2009 Oct 19;28(4):648–654. doi: 10.1200/JCO.2009.22.2950

Table 4.

Univariate and Multivariate Logistic Regression Analysis of Risk of Reinduction Failure

Variable No. Univariate Analyses
Multivariate Analyses
OR* 95% CI* OR 95% CI
Treatment attempt
    Second 191 1.0 1.0
    Third 108 7.3 4.2 to 12.6 4.5 2.4 to 8.4
    Fourth 56 15.9 7.8 to 32.5 5.8 2.4 to 14.1
    Fifth through ninth 65 41.5 17.9 to 96.2 14.3 4.9 to 41.5
    P < .001 < .001
    P trend < .001 < .001
Duration of previous remission
    CR not achieved 105 4.1 2.3 to 7.2 2.3 1.02 to 5.0
    CR achieved, < 18 months duration 147 1.0 1.0
    CR achieved, 18 to 36 months duration 82 0.31 0.17 to 0.56 0.52 0.27 to 1.03
    CR achieved, ≥ 36 months duration 67 0.082 0.031 to 0.21 0.20 0.07, to 0.55
    Missing 19
    P < .001 < .001
    P trend < .001 < .001
NCI risk criteria at diagnosis
    Non-infants, standard risk 176 1.0 1.0
    Non-infants, high risk 180 1.5 0.98 to 2.3 1.4 0.82 to 2.5
    Non-infants, unknown 53 2.0 1.1 to 3.7 0.74 0.32 to 1.7
    Infants 11 0.93 0.26 to 3.3 2.1 0.47 to 9.8
    P .011 .31
Extramedullary involvement in relapse
    No 331 1.0 1.0
    Yes 82 0.53 0.32 to 0.88 0.68 0.36 to 1.3
    Missing 7
    P .011 .23
Immunophenotype
    Pre B cell 348 1.0 1.0
    T cell 72 2.0 1.2 to 3.3 1.2 0.60 to 2.4
    P .012 .61

Abbreviations: OR, odds ratio of achieving complete remission v < complete remission; CR, complete remission; NCI, National Cancer Institute.

*

ORs and 95% CIs from univariate logistic models.

ORs and 95% CIs from multivariate logistic models after adjusting for the other variables in the Table.

Test for trend.