Propensity score methods for confounding control in observational studies of therapeutics for COVID-19 infection.

The authors provide a brief overview of different propensity score methods that can be used in observational research studies that lack randomization. Under specific assumptions, these methods result in unbiased estimates of causal effects, but the different ways propensity score are used may require different assumptions and result in estimated treatment effects that can have meaningfully different interpretations. The authors review these issues and consider their implications for studies of therapeutics for COVID-19.

Uptake and Persistence of Safer Conception Strategies Among South African Women Planning for Pregnancy.

Safer conception strategies can minimize HIV acquisition during periconception periods among women living in HIV-endemic areas. We examined uptake and predictors of persistent use of the same safer conception strategy among a cohort of HIV-uninfected South African women ages 18-35 years planning for pregnancy with a partner living with HIV or of unknown HIV-serostatus. The safer conception strategies we evaluated included oral PrEP, condomless sex limited to peak fertility, and waiting for a better time to have a child (until, for example, the risks of HIV acquisition are reduced and/or the individual is prepared to care for a child); persistence was defined as using the same safer conception strategy from the first visit through 9 months follow-up. Modified Poisson regression models were used to examine predictors of persistent use of the same strategy. The average age of 227 women in our cohort was 24.6 (range: 18.0, 35.7) years. In this cohort, 121 (74.2%) women reported persisting in the same strategy through 9 months. Employment and HIV knowledge were associated with the persistent use of any strategy. Our results highlight the need to provide safer conception services to women exposed to HIV during periconception periods. Findings also offer some insights into factors that might influence persistent use. Further research is needed to better understand how to involve male partners and how their involvement might influence women's consistent use of safer conception strategies during periconception periods.

Staging and clean room: Constructs designed to facilitate transparency and reduce bias in comparative analyses of real-world data.

PURPOSE: We describe constructs designed to protect the integrity of the results from comparative analyses using real-world data (RWD): staging and clean room. METHODS: Staging involves performing sequential preliminary analyses and evaluating the population size available and potential bias before conducting comparative analyses. A clean room involves restricted access to data and preliminary results, policies governing exploratory analyses and protocol deviations, and audit trail. These constructs are intended to allow decisions about protocol deviations, such as changes to design or model specification, to be made without knowledge of how they might affect subsequent analyses. We describe an example for implementing staging with a clean room. RESULTS: Stage 1 may involve selecting a data source, developing and registering a protocol, establishing a clean room, and applying inclusion/exclusion criteria. Stage 2 may involve attempting to achieve covariate balance, often through propensity score models. Stage 3 may involve evaluating the presence of residual confounding using negative control outcomes. After each stage, check points may be implemented when a team of statisticians, epidemiologists and clinicians masked to how their decisions may affect study outcomes, reviews the results. This review team may be tasked with making recommendations for protocol deviations to address study precision or bias. They may recommend proceeding to the next stage, conducting additional analyses to address bias, or terminating the study. Stage 4 may involve conducting the comparative analyses. CONCLUSIONS: The staging and clean room constructs are intended to protect the integrity and enhance confidence in the results of analyses of RWD.

High PrEP uptake and objective longitudinal adherence among HIV-exposed women with personal or partner plans for pregnancy in rural Uganda: A cohort study.

BACKGROUND: In Uganda, fertility rates and adult HIV prevalence are high, and many women conceive with partners living with HIV. Pre-exposure prophylaxis (PrEP) reduces HIV acquisition for women and, therefore, infants. We developed the Healthy Families-PrEP intervention to support PrEP use as part of HIV prevention during periconception and pregnancy periods. We conducted a longitudinal cohort study to evaluate oral PrEP use among women participating in the intervention. METHODS AND FINDINGS: We enrolled HIV-negative women with plans for pregnancy with a partner living, or thought to be living, with HIV (2017 to 2020) to evaluate PrEP use among women participating in the Healthy Families-PrEP intervention. Quarterly study visits through 9 months included HIV and pregnancy testing and HIV prevention counseling. PrEP was provided in electronic pillboxes, providing the primary adherence measure ("high" adherence when pillbox was opened ≥80% of days). Enrollment questionnaires assessed factors associated with PrEP use. Plasma tenofovir (TFV) and intraerythrocytic TFV-diphosphate (TFV-DP) concentrations were determined quarterly for women who acquired HIV and a randomly selected subset of those who did not; concentrations TFV ≥40 ng/mL and TFV-DP ≥600 fmol/punch were categorized as "high." Women who became pregnant were initially exited from the cohort by design; from March 2019, women with incident pregnancy remained in the study with quarterly follow-up until pregnancy outcome. Primary outcomes included (1) PrEP uptake (proportion who initiated PrEP); and (2) PrEP adherence (proportion of days with pillbox openings during the first 3 months following PrEP initiation). We used univariable and multivariable-adjusted linear regression to evaluate baseline predictors selected based on our conceptual framework of mean adherence over 3 months. We also assessed mean monthly adherence over 9 months of follow-up and during pregnancy. We enrolled 131 women with mean age 28.7 years (95% CI: 27.8 to 29.5). Ninety-seven (74%) reported a partner with HIV and 79 (60%) reported condomless sex. Most women (N = 118; 90%) initiated PrEP. Mean electronic adherence during the 3 months following initiation was 87% (95% CI: 83%, 90%). No covariates were associated with 3-month pill-taking behavior. Concentrations of plasma TFV and TFV-DP were high among 66% and 47%, 56% and 41%, and 45% and 45% at months 3, 6, and 9, respectively. We observed 53 pregnancies among 131 women (1-year cumulative incidence 53% [95% CI: 43%, 62%]) and 1 HIV-seroconversion in a non-pregnant woman. Mean pillcap adherence for PrEP users with pregnancy follow-up (N = 17) was 98% (95% CI: 97%, 99%). Study design limitations include lack of a control group. CONCLUSIONS: Women in Uganda with PrEP indications and planning for pregnancy chose to use PrEP. By electronic pillcap, most were able to sustain high adherence to daily oral PrEP prior to and during pregnancy. Differences in adherence measures highlight challenges with adherence assessment; serial measures of TFV-DP in whole blood suggest 41% to 47% of women took sufficient periconception PrEP to prevent HIV. These data suggest that women planning for and with pregnancy should be prioritized for PrEP implementation, particularly in settings with high fertility rates and generalized HIV epidemics. Future iterations of this work should compare the outcomes to current standard of care. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03832530 https://clinicaltrials.gov/ct2/show/NCT03832530?term=lynn+matthews&cond=hiv&cntry=UG&draw=2&rank=1.

Predictors of HIV seroconversion in Botswana: machine learning analysis in a representative, population-based HIV incidence cohort.

OBJECTIVE: To identify predictors of HIV acquisition in Botswana. DESIGN: We applied machine learning approaches to identify HIV risk predictors using existing data from a large, well-characterized HIV incidence cohort. METHODS: We applied machine learning (randomForestSRC) to analyze data from a large population-based HIV incidence cohort enrolled in a cluster-randomized HIV prevention trial in 30 communities across Botswana. We sought to identify the most important risk factors for HIV acquisition, starting with 110 potential predictors. RESULTS: During a median 29-month follow-up of 8,551 HIV-negative adults, 147 (1.7%) acquired HIV. Our machine learning analysis found that for females, the most important variables for predicting HIV acquisition were the use of injectable hormonal contraception, frequency of sex in the prior 3 months with the most recent partner and residing in a community with HIV prevalence of 29% or higher. For the small proportion (0.3%) of females who had all three risk factors, their estimated probability of acquiring HIV during 29 months of follow-up was 34% (approximate annual incidence of 14%). For males, non-long-term relationships with the most recent partner and community HIV prevalence of 34% or higher were the most important HIV risk predictors. The 6% of males who had both risk factors had a 5.1% probability of acquiring HIV during the follow-up period (approximate annual incidence of 2.1%). CONCLUSIONS: Machine learning approaches allowed us to analyze a large number of variables to efficiently identify key factors strongly predictive of HIV risk. These factors could help target HIV prevention interventions in Botswana. CLINICAL TRIALS REGISTRATION: NCT01965470.

Oral preexposure prophylaxis uptake, adherence, and persistence during periconception periods among women in South Africa.

OBJECTIVE: We developed the Healthy Families-PrEP intervention to support HIV-prevention during periconception and pregnancy. We evaluated preexposure prophylaxis (PrEP) use with three objective measures. DESIGN: This single-arm intervention study enrolled women in KwaZulu-Natal, South Africa, who were HIV-uninfected, not pregnant, in a relationship with a partner with HIV or unknown-serostatus, and with pregnancy plans. PrEP was offered as part of a comprehensive HIV prevention intervention. Participants were followed for 12 months. METHODS: We evaluated periconception PrEP uptake and adherence using quarterly plasma tenofovir concentrations. We modeled factors associated with PrEP uptake and high plasma tenofovir (past day dosing). Patterns of use were analyzed using electronic pillcap data. Dried blood spots to measure intracellular tenofovir product (past 2 months dosing) were analyzed for a subset of women. RESULTS: Three hundred thirty women with median age 24 (IQR: 22-27) years enrolled. Partner HIV-serostatus was unknown by 96% ( N  = 316); 60% (195) initiated PrEP. High plasma tenofovir concentrations were seen in 35, 25, 22, and 20% of samples at 3, 6, 9, and 12 months, respectively. Similar adherence was measured by pillcap and dried blood spots. In adjusted models, lower income, alcohol use, and higher HIV stigma were associated with high plasma tenofovir. Eleven HIV-seroconversions were observed (incidence rate: 4.04/100 person-years [95% confidence interval: 2.24-7.30]). None had detectable plasma tenofovir. CONCLUSION: The Healthy Families-PrEP intervention supported women in PrEP use. We observed high interest in periconception PrEP and over one-third adhered to PrEP in the first quarter; one-fifth were adherent over a year. High HIV incidence highlights the importance of strategies to reduce HIV incidence among periconception women. CLINICAL TRIAL NUMBER: NCT03194308.

Adherence to daily, oral TDF/FTC PrEP during periconception among HIV-exposed South African women.

Background: Daily, oral tenofovir disoproxil fumarate/emtricitabine (TDF/FTC) as pre-exposure prophylaxis (PrEP) reduces HIV acquisition for African women. Adherence is key to efficacy and patterns of adherence can be highly variable in real-world settings. Using group-based trajectory modeling (GBTM), we sought to identify distinct patterns of periconception PrEP adherence and evaluate potential baseline predictors of such adherence trajectories. Methods: We conducted a single-arm longitudinal study for women aged 18-35 years living in Durban, South Africa with personal or partner plans for pregnancy with a partner with HIV or of unknown serostatus. Participants were offered safer conception counseling, including daily oral PrEP; women who initiated PrEP were given a bottle with an electronic pillcap that recorded when device opens. Weekly adherence to daily PrEP was modeled using GBTM with a censored normal outcome distribution as a function of weeks since PrEP initiation. The number and functional form of the adherence trajectory groups were primarily selected based on Bayesian information criteria (BIC) and confirmed by mean estimated probabilities of group membership. A multivariable version of the selected model assessed baseline predictors of membership in adherence trajectory groups. Results: Overall mean (95% CI) adherence to PrEP was 63% (60%, 67%). We identified four groups of women with distinct patterns of adherence: (1) high (i.e., ≥6 doses per week) steady adherence throughout follow-up (22% of PrEP initiators); (2) moderate (i.e., 4-5 doses per week), but steady adherence (31%); (3) initially high, but consistently declining adherence (21%); and (4) initially moderate adherence, followed by a rapid decline and subsequent rebound (26%). In multivariable-adjusted analyses, older age was associated with membership in the high, steady adherence group as compared to the group identified with an adherence trajectory of initially high, then decline, and finally a rebound. Conclusions: GBTM is useful for exploring potential heterogeneity in longitudinal patterns of medication adherence. Although a large proportion of women in this study achieved high levels of adherence by electronic pillcap initially, far fewer women maintained these levels consistently. Knowledge of different adherence trajectories could be used to develop targeted strategies for optimizing HIV prevention during periconception.