Search

Cluster randomized trials (CRTs) often enroll large numbers of participants; yet due to resource constraints, only a subset of participants may be selected for outcome assessment, and those sampled may not be representative of all cluster members. Missing data also present a challenge: if sampled individuals with measured outcomes are dissimilar from those with missing outcomes, unadjusted estimates of arm-specific endpoints and the intervention effect may be biased. Further, CRTs often enroll and randomize few clusters, limiting statistical power and raising concerns about finite sample performance. Motivated by SEARCH-TB, a CRT aimed at reducing incident tuberculosis infection, we demonstrate interlocking methods to handle these challenges. First, we extend Two-Stage targeted minimum loss-based estimation to account for three sources of missingness: (i) subsampling; (ii) measurement of baseline status among those sampled; and (iii) measurement of final status among those in the incidence cohort (persons known to be at risk at baseline). Second, we critically evaluate the assumptions under which subunits of the cluster can be considered the conditionally independent unit, improving precision and statistical power but also causing the CRT to behave like an observational study. Our application to SEARCH-TB highlights the real-world impact of different assumptions on measurement and dependence; estimates relying on unrealistic assumptions suggested the intervention increased the incidence of TB infection by 18% (risk ratio [RR] = 1.18, 95% confidence interval [CI]: 0.85–1.63), while estimates accounting for the sampling scheme, missingness, and within community dependence found the intervention decreased the incident TB by 27% (RR = 0.73, 95% CI: 0.57–0.92). [ABSTRACT FROM AUTHOR]

Background Human immunodeficiency virus (HIV) treatment reduces tuberculosis (TB) disease and mortality; however, the population-level impact of universal HIV-test-and-treat interventions on TB infection and transmission remain unclear. Methods In a sub-study nested in the SEARCH trial, a community cluster-randomized trial (NCT01864603), we assessed whether a universal HIV-test-and-treat intervention reduced population-level incident TB infection in rural Uganda. Intervention communities received annual, population-level HIV testing and patient-centered linkage. Control communities received population-level HIV testing at baseline and endline. We compared estimated incident TB infection by arms, defined by tuberculin skin test conversion in a cohort of persons aged 5 and older, adjusting for participation and predictors of infection, and accounting for clustering. Results Of the 32 trial communities, 9 were included, comprising 90 801 participants (43 127 intervention and 47 674 control). One-year cumulative incidence of TB infection was 16% in the intervention and 22% in the control; SEARCH reduced the population-level risk of incident TB infection by 27% (adjusted risk ratio = 0.73; 95% confidence interval [CI]:.57–.92, P =.005). In pre-specified analyses, the effect was largest among children aged 5–11 years and males. Conclusions A universal HIV-test-and-treat intervention reduced incident TB infection, a marker of population-level TB transmission. Investments in community-level HIV interventions have broader population-level benefits, including TB reductions. [ABSTRACT FROM AUTHOR]

Benkeser et al. demonstrate how adjustment for baseline covariates in randomized trials can meaningfully improve precision for a variety of outcome types. Their findings build on a long history, starting in 1932 with R.A. Fisher and including more recent endorsements by the U.S. Food and Drug Administration and the European Medicines Agency. Here, we address an important practical consideration: how to select the adjustment approach—which variables and in which form—to maximize precision, while maintaining Type-I error control. Balzer et al. previously proposed Adaptive Pre-specification within TMLE to flexibly and automatically select, from a prespecified set, the approach that maximizes empirical efficiency in small trials (N  < 40). To avoid overfitting with few randomized units, selection was previously limited to working generalized linear models, adjusting for a single covariate. Now, we tailor Adaptive Pre-specification to trials with many randomized units. Using V -fold cross-validation and the estimated influence curve-squared as the loss function, we select from an expanded set of candidates, including modern machine learning methods adjusting for multiple covariates. As assessed in simulations exploring a variety of data-generating processes, our approach maintains Type-I error control (under the null) and offers substantial gains in precision—equivalent to 20%-43% reductions in sample size for the same statistical power. When applied to real data from ACTG Study 175, we also see meaningful efficiency improvements overall and within subgroups. [ABSTRACT FROM AUTHOR]

In this issue, Naimi et al. (Am J Epidemiol. 2023;192(9):1536–1544) discuss a critical topic in public health and beyond: obtaining valid statistical inference when using machine learning in causal research. In doing so, the authors review recent prominent methodological work and recommend: 1) doubly robust estimators, such as targeted maximum likelihood estimation (TMLE); 2) ensemble methods, such as Super Learner, to combine predictions from a diverse library of algorithms; and 3) sample splitting to reduce bias and improve inference. We largely agree with these recommendations. In this commentary, we highlight the critical importance of the Super Learner library. Specifically, in both simulation settings considered by the authors, we demonstrate that reductions in bias and improvements in confidence-interval coverage can be achieved using TMLE without sample splitting and with a Super Learner library that excludes tree-based methods but includes regression splines. Whether extremely data-adaptive algorithms and sample splitting are needed depends on the specific problem and should be informed by simulations reflecting the specific application. More research is needed on practical recommendations for selecting among these options in common situations arising in epidemiology. [ABSTRACT FROM AUTHOR]

Mixed evidence exists of associations between mobility data and coronavirus disease 2019 (COVID-19) case rates. We aimed to evaluate the county-level impact of reducing mobility on new COVID-19 cases in summer/fall of 2020 in the United States and to demonstrate modified treatment policies to define causal effects with continuous exposures. Specifically, we investigated the impact of shifting the distribution of 10 mobility indexes on the number of newly reported cases per 100,000 residents 2 weeks ahead. Primary analyses used targeted minimum loss-based estimation with Super Learner to avoid parametric modeling assumptions during statistical estimation and flexibly adjust for a wide range of confounders, including recent case rates. We also implemented unadjusted analyses. For most weeks, unadjusted analyses suggested strong associations between mobility indexes and subsequent new case rates. However, after confounder adjustment, none of the indexes showed consistent associations under mobility reduction. Our analysis demonstrates the utility of this novel distribution-shift approach to defining and estimating causal effects with continuous exposures in epidemiology and public health. [ABSTRACT FROM AUTHOR]

Cluster randomized trials (CRTs) randomly assign an intervention to groups of individuals (e.g. clinics or communities) and measure outcomes on individuals in those groups. While offering many advantages, this experimental design introduces challenges that are only partially addressed by existing analytic approaches. First, outcomes are often missing for some individuals within clusters. Failing to appropriately adjust for differential outcome measurement can result in biased estimates and inference. Second, CRTs often randomize limited numbers of clusters, resulting in chance imbalances on baseline outcome predictors between arms. Failing to adaptively adjust for these imbalances and other predictive covariates can result in efficiency losses. To address these methodological gaps, we propose and evaluate a novel two-stage targeted minimum loss-based estimator to adjust for baseline covariates in a manner that optimizes precision, after controlling for baseline and postbaseline causes of missing outcomes. Finite sample simulations illustrate that our approach can nearly eliminate bias due to differential outcome measurement, while existing CRT estimators yield misleading results and inferences. Application to real data from the SEARCH community randomized trial demonstrates the gains in efficiency afforded through adaptive adjustment for baseline covariates, after controlling for missingness on individual-level outcomes. [ABSTRACT FROM AUTHOR]

Background Social network analysis can elucidate tuberculosis transmission dynamics outside the home and may inform novel network-based case-finding strategies. Methods We assessed the association between social network characteristics and prevalent tuberculosis infection among residents (aged ≥15 years) of 9 rural communities in Eastern Uganda. Social contacts named during a census were used to create community-specific nonhousehold social networks. We evaluated whether social network structure and characteristics of first-degree contacts (sex, human immunodeficiency virus [HIV] status, tuberculosis infection) were associated with revalent tuberculosis infection (positive tuberculin skin test [TST] result) after adjusting for individual-level risk factors (age, sex, HIV status, tuberculosis contact, wealth, occupation, and Bacillus Calmette–Guérin [BCG] vaccination) with targeted maximum likelihood estimation. Results Among 3 335 residents sampled for TST, 32% had a positive TST results and 4% reported a tuberculosis contact. The social network contained 15 328 first-degree contacts. Persons with the most network centrality (top 10%) (adjusted risk ratio, 1.3 [95% confidence interval, 1.1–1.1]) and the most (top 10%) male contacts (1.5 [1.3–1.9]) had a higher risk of prevalent tuberculosis, than those in the remaining 90%. People with ≥1 contact with HIV (adjusted risk ratio, 1.3 [95% confidence interval, 1.1–1.6]) and ≥2 contacts with tuberculosis infection were more likely to have tuberculosis themselves (2.6 [ 95% confidence interval: 2.2–2.9]). Conclusions Social networks with higher centrality, more men, contacts with HIV, and tuberculosis infection were positively associated with tuberculosis infection. Tuberculosis transmission within measurable social networks may explain prevalent tuberculosis not associated with a household contact. Further study on network-informed tuberculosis case finding interventions is warranted. [ABSTRACT FROM AUTHOR]

Study Question: Is sperm epigenetic aging (SEA) associated with probability of pregnancy among couples in the general population?Summary Answer: We observed a 17% lower cumulative probability at 12 months for couples with male partners in the older compared to the younger SEA categories.What Is Known Already: The strong relation between chronological age and DNA methylation profiles has enabled the estimation of biological age as epigenetic 'clock' metrics in most somatic tissue. Such clocks in male germ cells are less developed and lack clinical relevance in terms of their utility to predict reproductive outcomes.Study Design, Size, Duration: This was a population-based prospective cohort study of couples discontinuing contraception to become pregnant recruited from 16 US counties from 2005 to 2009 and followed for up to 12 months.Participants/materials, Setting, Methods: Sperm DNA methylation from 379 semen samples was assessed via a beadchip array. A state-of-the-art ensemble machine learning algorithm was employed to predict age from the sperm DNA methylation data. SEA was estimated from clocks derived from individual CpGs (SEACpG) and differentially methylated regions (SEADMR). Probability of pregnancy within 1 year was compared by SEA, and discrete-time proportional hazards models were used to evaluate the relations with time-to-pregnancy (TTP) with adjustment for covariates.Main Results and the Role Of Chance: Our SEACpG clock had the highest predictive performance with correlation between chronological and predicted age (r = 0.91). In adjusted discrete Cox models, SEACpG was negatively associated with TTP (fecundability odds ratios (FORs)=0.83; 95% CI: 0.76, 0.90; P = 1.2×10-5), indicating a longer TTP with advanced SEACpG. For subsequent birth outcomes, advanced SEACpG was associated with shorter gestational age (n = 192; -2.13 days; 95% CI: -3.67, -0.59; P = 0.007). Current smokers also displayed advanced SEACpG (P < 0.05). Finally, SEACpG showed a strong performance in an independent IVF cohort (n = 173; r = 0.83). SEADMR performance was comparable to SEACpG but with attenuated effect sizes.Limitations, Reasons For Caution: This prospective cohort study consisted primarily of Caucasian men and women, and thus analysis of large diverse cohorts is necessary to confirm the associations between SEA and couple pregnancy success in other races/ethnicities.Wider Implications Of the Findings: These data suggest that our sperm epigenetic clocks may have utility as a novel biomarker to predict TTP among couples in the general population and underscore the importance of the male partner for reproductive success.Study Funding/competing Interest(s): This work was funded in part by grants the National Institute of Environmental Health Sciences, National Institutes of Health (R01 ES028298; PI: J.R.P. and P30 ES020957); Robert J. Sokol, MD Endowed Chair of Molecular Obstetrics and Gynecology (J.R.P.); and the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland (Contracts N01-HD-3-3355, N01-HD-3-3356 and N01-HD-3-3358). S.L.M. was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health. The authors declare no competing interests.Trial Registration Number: N/A. [ABSTRACT FROM AUTHOR]

In this issue of the Journal , Mooney et al. (Am J Epidemiol. 2021;190(8):1476–1482) discuss machine learning as a tool for causal research in the style of Internet headlines. Here we comment by adapting famous literary quotations, including the one in our title (from "Sonnet 43" by Elizabeth Barrett Browning (Sonnets From the Portuguese , Adelaide Hanscom Leeson, 1850)). We emphasize that any use of machine learning to answer causal questions must be founded on a formal framework for both causal and statistical inference. We illustrate the pitfalls that can occur without such a foundation. We conclude with some practical recommendations for integrating machine learning into causal analyses in a principled way and highlight important areas of ongoing work. [ABSTRACT FROM AUTHOR]

Background In generalized epidemic settings, strategies are needed to prioritize individuals at higher risk of human immunodeficiency virus (HIV) acquisition for prevention services. We used population-level HIV testing data from rural Kenya and Uganda to construct HIV risk scores and assessed their ability to identify seroconversions. Methods During 2013–2017, >75% of residents in 16 communities in the SEARCH study were tested annually for HIV. In this population, we evaluated 3 strategies for using demographic factors to predict the 1-year risk of HIV seroconversion: membership in ≥1 known "risk group" (eg, having a spouse living with HIV), a "model-based" risk score constructed with logistic regression, and a "machine learning" risk score constructed with the Super Learner algorithm. We hypothesized machine learning would identify high-risk individuals more efficiently (fewer persons targeted for a fixed sensitivity) and with higher sensitivity (for a fixed number targeted) than either other approach. Results A total of 75 558 persons contributed 166 723 person-years of follow-up; 519 seroconverted. Machine learning improved efficiency. To achieve a fixed sensitivity of 50%, the risk-group strategy targeted 42% of the population, the model-based strategy targeted 27%, and machine learning targeted 18%. Machine learning also improved sensitivity. With an upper limit of 45% targeted, the risk-group strategy correctly classified 58% of seroconversions, the model-based strategy 68%, and machine learning 78%. Conclusions Machine learning improved classification of individuals at risk of HIV acquisition compared with a model-based approach or reliance on known risk groups and could inform targeting of prevention strategies in generalized epidemic settings. Clinical Trials Registration NCT01864603. [ABSTRACT FROM AUTHOR]

Epidemiologic methods have advanced tremendously in the last several decades. As important as they are, even the most sophisticated approaches are unable to provide meaningful answers when the user lacks a clear study question. Yet, instructors have more and more resources on how to conduct studies and analyze data but few resources on how to ask clearly defined study questions that will guide those methods. Training programs have limited time for coursework, and if novel statistical estimation methods become the focus of instruction, programs that go this route may end up underemphasizing the process of asking good study questions, designing robust studies, considering potential biases in the collected data, and appropriately interpreting the results of the analysis. Given the demands for space in curricula, now is an appropriate time to reevaluate what we teach epidemiology doctoral students. We advocate that programs place a renewed focus on asking good study questions and following a comprehensive approach to study design and data analysis in which questions guide the choice of appropriate methods, helping us avoid methods for methods' sake and highlighting when application of a new method can provide the opportunity to answer questions that were intractable with traditional approaches. [ABSTRACT FROM AUTHOR]

Background Global guidelines recommend preexposure prophylaxis (PrEP) for individuals with substantial human immunodeficiency virus (HIV) risk. Data on PrEP uptake in sub-Saharan Africa outside of clinical trials are limited. We report on "early adopters" of PrEP in the Sustainable East Africa Research in Community Health (SEARCH) study in rural Uganda and Kenya. Methods After community mobilization and PrEP education, population-based HIV testing was conducted. HIV-uninfected adults were offered PrEP based on an empirically derived HIV risk score or self-identified HIV risk (if not identified by score). Using logistic regression, we analyzed predictors of early PrEP adoption (starting PrEP within 30 days vs delayed/no start) among adults identified for PrEP. Results Of 21212 HIV-uninfected adults in 5 communities, 4064 were identified for PrEP (2991 by empiric risk score, 1073 by self-identified risk). Seven hundred and thirty nine individuals started PrEP within 30 days (11% of those identified by risk score; 39% of self-identified); 77% on the same day. Among adults identified by risk score, predictors of early adoption included male sex (adjusted odds ratio 1.53; 95% confidence interval, 1.09–2.15), polygamy (1.92; 1.27–2.90), serodiscordant spouse (3.89; 1.18–12.76), self-perceived HIV risk (1.66; 1.28–2.14), and testing at health campaign versus home (5.24; 3.33–8.26). Among individuals who self-identified for PrEP, predictors of early adoption included older age (2.30; 1.29–4.08) and serodiscordance (2.61; 1.01–6.76). Conclusions Implementation of PrEP incorporating a population-based empiric risk score, self-identified risk, and rapid initiation, is feasible in rural East Africa. Strategies are needed to overcome barriers to PrEP uptake, particularly among women and youth. Clinical Trials Registration NCT01864603. [ABSTRACT FROM AUTHOR]

In a randomized controlled trial, we tested whether a structured, patient-centered phone call from a clinical officer after HIV testing improved linkage to/re-engagement in HIV care. Among 130 HIV-positive persons, those randomized to the phone call were significantly more likely to link to care by 7 and 30 days (P =.04). [ABSTRACT FROM AUTHOR]