PrEPception: Expanding Assisted Reproductive Options for Serodiscordant Couples

Adherence will be measured in two ways: (1) Dried blood spots (DBS) to measure drug level two times, once after the woman has been taking the drug for one month and once when pregnancy is confirmed. We will perform unadjusted, bivariate analyses comparing participants below and equal to the drug level threshold for daily dosing using a one sample t-test for normally distributed continuous or ordinal variables on adherence. We will also perform a paired t-test for measuring the difference, if any, between the two adherence tests at the different time points for each individual. Adherence will also be measured through (2) participant self-report via a continuous variable of number of doses missed in the past month. We will perform unadjusted, bivariate analyses comparing the observed participant self-reported adherence levels to the gold standard of ≥95% self-reported adherence using a one-sample t test of means for normally distributed continuous or ordinal variables.

Condom use will be measured through participant self-report on number of times a condom was used during the past month compared to the number of times the participant had sexual intercourse with his/her partner or other individuals in the past month. We will perform unadjusted bivariate analyses comparing the observed participants in a one-sample student's t-test for normally distributed variables on condom use or a paired t-test for measuring the difference between two variables at different time points for the same subject (i.e. at baseline and at end-line evaluation). For a non-normal distribution, we will use the non-parametric equivalent such as Sign or Wilcoxon Signed rank test for a symmetric distribution.

We will quantify this outcome measure by examining the number of couples who complete a clinical protocol for PrEP compared to the total number of couples that initiate PrEP, while examining other variables that could make the two groups statistically different. We will perform unadjusted bivariate analyses comparing the number of couples that complete the clinical protocol and those who withdraw or become ineligible throughout the duration of the study. We will use a one-way ANOVA or Kruskal-Wallis non-parametric test for continuous variables with two or more categories such as those that complete the protocol versus those that withdraw from PrEP versus those that become ineligible. We will perform a two-sample test of means to compare the overall categories of those that complete the protocol versus those that do not complete the protocol and create cross-tabulations with chi-square and Fisher's exact tests.

For this outcome measure, we will perform unadjusted bivariate analyses in a paired t-test for measuring the difference in beliefs and sexual behavior at different time points for the same subject. For a non-normal distribution, we will use the non-parametric equivalent such as Sign or Wilcoxon Signed rank test for a symmetric distribution.

To assess renal toxicities, we will perform unadjusted bivariate analyses in a paired t-test for measuring the difference of creatinine clearance from the baseline compared to creatinine clearance at the conclusion of PrEP for the same subject. For a non-normal distribution, we will use the non-parametric equivalent such as Sign or Wilcoxon Signed rank test for a symmetric distribution. We will also report on basic drug side effects, while summarizing other self-reported effects of PrEP.

We will perform unadjusted bivariate analyses in a two-sample test of means to compare rates of teratogenicity of newborns with HIV-negative females using PrEP as compared to rates of birth defects within the general population in similar age group women.

For this outcome measure, we will report descriptively on whether there is a transmission or not within the observed participants taking PrEP for conception purposes.