TABLE 3.
Variablea | Parameter estimate | SE | Pr > chi squareb | Odds ratio |
---|---|---|---|---|
African-American ethnic group | −0.46 | 0.25 | 0.06 | 0.63 |
Mother received AZT | −0.49 | 0.33 | 0.13 | 0.61 |
Infant received AZT | 0.82 | 0.47 | 0.08 | 2.28 |
Log of infant’s CD4 count | −0.51 | 0.24 | 0.04 | 0.60 |
Premature birth | 0.37 | 0.28 | 0.18 | 1.45 |
Birth after February 1994 | −1.25 | 0.46 | 0.01 | 0.29 |
Homozygous CCR5-59356T | 1.35 | 0.59 | 0.02 | 3.84 |
Log of mother’s CD4 count | −0.55 | 0.14 | 0.00 | 0.57 |
Logistic regression analysis of the combined WITS and ARIEL cohorts (total of 656; 94 HIV-1 infected) based 1 df for each variable. Variables that were thought to influence transmission were included. The African-American ethnic group was included as a variable because the allele frequency for CCR5-59356T is far greater in that group.
Probability that under the null hypothesis of no correlation with transmission, the observed data are attributable to chance alone.