Objective
Cesarean delivery before labor and before ruptured membranes is efficacious in the prevention of mother-to-child transmission (MTCT) of HIV. However, cesarean delivery is associated with maternal and infant morbidity, and viral load (VL; copies per milliliter) results at delivery often are not available. A VL at 34-36 weeks of gestation may predict the VL around delivery and, thus, indicate the need for cesarean delivery. Our objective was to evaluate the correlation between a VL at 34-36 weeks of gestation and in the immediate postpartum period (imPP) in a cohort of HIV-infected pregnant women using highly active antiretroviral therapy (HAART) at a referral center in Rio de Janeiro, Brazil (Hospital Federal dos Servidores do Estado; HFSE) from January 2007 to September 2013.
Study Design
Data were extracted from the HFSE database. According to the HFSE institutional review board, informed consent for this analysis was not required. The exclusion criteria are shown in the Figure . The correlation between 2 VLs was evaluated by the Pearson correlation coefficient ( r ). The Cohen kappa (κ) of the antepartum VL was calculated in relation to the imPP VL, with cutoff values of ≤50, ≤400, and ≤1000 copies/mL. Covariates that were statistically significant ( P ≤ .15) in univariate analysis were included in multivariable analysis. Only variables with a probability value of ≤ .05 remained in the multivariate model. Statistical analysis was performed with SPSS WIN software (version 16.0; SPSS Inc, Chicago, IL).
Study Design
Data were extracted from the HFSE database. According to the HFSE institutional review board, informed consent for this analysis was not required. The exclusion criteria are shown in the Figure . The correlation between 2 VLs was evaluated by the Pearson correlation coefficient ( r ). The Cohen kappa (κ) of the antepartum VL was calculated in relation to the imPP VL, with cutoff values of ≤50, ≤400, and ≤1000 copies/mL. Covariates that were statistically significant ( P ≤ .15) in univariate analysis were included in multivariable analysis. Only variables with a probability value of ≤ .05 remained in the multivariate model. Statistical analysis was performed with SPSS WIN software (version 16.0; SPSS Inc, Chicago, IL).
