Tree-based subgroup discovery using electronic health record data: heterogeneity of treatment effects for DTG-containing therapies.

The rich longitudinal individual level data available from electronic health records (EHRs) can be used to examine treatment effect heterogeneity. However, estimating treatment effects using EHR data poses several challenges, including time-varying confounding, repeated and temporally non-aligned measurements of covariates, treatment assignments and outcomes, and loss-to-follow-up due to dropout. Here, we develop the subgroup discovery for longitudinal data algorithm, a tree-based algorithm for discovering subgroups with heterogeneous treatment effects using longitudinal data by combining the generalized interaction tree algorithm, a general data-driven method for subgroup discovery, with longitudinal targeted maximum likelihood estimation. We apply the algorithm to EHR data to discover subgroups of people living with human immunodeficiency virus who are at higher risk of weight gain when receiving dolutegravir (DTG)-containing antiretroviral therapies (ARTs) versus when receiving non-DTG-containing ARTs.

Evaluation of Rhode Island's Early Geographic COVID-19 Vaccine Prioritization Policy.

Objectives. To determine whether geographic prioritization of limited COVID-19 vaccine supply was effective for reducing geographic disparities in case rates. Methods. Rhode Island allocated a portion of the initial COVID-19 vaccine supply to residents of Central Falls, a community already affected by structural policies and inadequate systems that perpetuate health inequities and experiencing disproportionately high COVID-19 morbidity and mortality. The policy was implemented with a culturally and linguistically appropriate community engagement plan and was intended to reduce observed disparities. Using a Bayesian causal analysis with population surveillance data, we evaluated the impact of this prioritization policy on recorded cases over the subsequent 16 weeks. Results. Early geographic prioritization of Central Falls accelerated vaccine uptake, averting an estimated 520 cases (95% confidence interval = 22, 1418) over 16 weeks and reducing cases by approximately 34% during this period (520 averted vs 1519 expected without early prioritization). Conclusions. Early geographic prioritization increased vaccine uptake and reduced cases in Central Falls, thereby reducing geographic disparities. Public Health Implications. Public health institutions should consider geographic prioritization of limited vaccine supply to reduce geographic disparities in case rates. (Am J Public Health. 2024;114(S7):S580-S589. https://doi.org/10.2105/AJPH.2024.307741).

HIV Incidence and Death Among Orphaned and Nonorphaned Children and Adolescents Living in Family-Based Settings in Western Kenya: A Prospective Cohort Analysis.

OBJECTIVE: To compare the incidence of HIV, death, and abuse among orphaned children to nonorphaned children living in households caring for orphaned children in Western Kenya. STUDY DESIGN: A random sample was taken of 300 households caring for at least one orphaned child in Uasin Gishu County, Kenya. All orphaned and nonorphaned children in each selected household were enrolled in a prospective cohort study between 2010 and 2013. A total of 1488 children (487 double orphans, 743 single orphans, and 258 nonorphans) were followed up annually until 2019. Survival analysis was used to estimate hazard ratios and 95% confidence intervals (CIs) of the association between the number of parents the child had lost (none, 1, or 2), and HIV incidence, death, combined HIV incidence or death, and incident abuse. RESULTS: Among 1488 children enrolled, 52% of participants were females, 23 were HIV positive, and the median age was 10.4 years. Over the course of the study, 16 orphaned children died and 11 acquired HIV. No deaths or incident HIV infections were observed among the nonorphaned children. Among children who were HIV negative at enrollment, loss of a parent was strongly associated with incident HIV (adjusted hazard ratio: 2.21 per parent lost, 95% CI: 1.03-4.73) and HIV or death (adjusted hazard ratio: 2.46 per parent lost, 95% CI: 1.37-4.42). There were no significant associations between orphan level and abuse. CONCLUSIONS: In similar households, orphaned children experience a higher risk of HIV and death than nonorphaned children. Both orphaned children and the families caring for them need additional support to prevent adverse health outcomes.

Characterization of HIV-Related Stigma in Myanmar.

Characterizing HIV-related stigma and its impacts are important for interventions toward their elimination. A cross-sectional study was conducted in 2016 to evaluate enacted and internalized stigma among adult people living with HIV (PLWH) across four cities in Myanmar using the India Stigma Index questionnaire. Multivariable regression analyses were performed to determine differences in measured enacted and internalized stigma outcomes. Among 1,006 participants, 89% reported any stigma indicator, 47% enacted stigma, and 87% internalized stigma. In regression analysis, city and duration of illness were associated with higher enacted stigma, and younger age was associated with higher internalized stigma. Those with HIV duration > 7.4 years had mean enacted stigma nearly 2 units higher than the overall mean. Internalized stigma increased with duration of illness and leveled off at 5 years. PLWH from smaller cities experienced lower stigma. In Myanmar, nearly 90% of PLWH experience stigma, results that reflect a unique transition point.

Validity of Self-Report for Ascertaining HIV Status Among Circular Migrants and Permanent Residents in South Africa: A Cross-Sectional, Population-Based Analysis.

While expanded HIV testing is needed in South Africa, increasing accurate self-report of HIV status is an essential parallel goal in this highly mobile population. If self-report can ascertain true HIV-positive status, persons with HIV (PWH) could be linked to life-saving care without the existing delays required by producing medical records or undergoing confirmatory testing, which are especially burdensome for the country's high prevalence of circular migrants. We used Wave 1 data from The Migration and Health Follow-Up Study, a representative adult cohort, including circular migrants and permanent residents, randomly sampled from the Agincourt Health and Demographic Surveillance System in a rural area of Mpumalanga Province. Within the analytic sample (n = 1,918), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of self-report were calculated with dried blood spot (DBS) HIV test results as the standard. Among in-person participants (n = 2,468), 88.8% consented to DBS-HIV testing. HIV prevalence was 25.3%. Sensitivity of self-report was 43.9% (95% CI: 39.5-48.5), PPV was 93.4% (95% CI: 89.5-96.0); specificity was 99.0% (95% CI: 98.3-99.4) and NPV was 83.9% (95% CI: 82.8-84.9). Self-report of an HIV-positive status was predictive of true status for both migrants and permanent residents in this high-prevalence setting. Persons who self-reported as living with HIV were almost always truly positive, supporting a change to clinical protocol to immediately connect persons who say they are HIV-positive to ART and counselling. However, 56% of PWH did not report as HIV-positive, highlighting the imperative to address barriers to disclosure.

A flexible approach for causal inference with multiple treatments and clustered survival outcomes.

When drawing causal inferences about the effects of multiple treatments on clustered survival outcomes using observational data, we need to address implications of the multilevel data structure, multiple treatments, censoring, and unmeasured confounding for causal analyses. Few off-the-shelf causal inference tools are available to simultaneously tackle these issues. We develop a flexible random-intercept accelerated failure time model, in which we use Bayesian additive regression trees to capture arbitrarily complex relationships between censored survival times and pre-treatment covariates and use the random intercepts to capture cluster-specific main effects. We develop an efficient Markov chain Monte Carlo algorithm to draw posterior inferences about the population survival effects of multiple treatments and examine the variability in cluster-level effects. We further propose an interpretable sensitivity analysis approach to evaluate the sensitivity of drawn causal inferences about treatment effect to the potential magnitude of departure from the causal assumption of no unmeasured confounding. Expansive simulations empirically validate and demonstrate good practical operating characteristics of our proposed methods. Applying the proposed methods to a dataset on older high-risk localized prostate cancer patients drawn from the National Cancer Database, we evaluate the comparative effects of three treatment approaches on patient survival, and assess the ramifications of potential unmeasured confounding. The methods developed in this work are readily available in the R $$ \mathsf{R}\kern.15em $$ package riAFTBART $$ \mathsf{riAFTBART} $$ .

Toward evaluation of disseminated effects of medications for opioid use disorder within provider-based clusters using routinely-collected health data.

Routinely-collected health data can be employed to emulate a target trial when randomized trial data are not available. Patients within provider-based clusters likely exert and share influence on each other's treatment preferences and subsequent health outcomes and this is known as dissemination or spillover. Extending a framework to replicate an idealized two-stage randomized trial using routinely-collected health data, an evaluation of disseminated effects within provider-based clusters is possible. In this article, we propose a novel application of causal inference methods for dissemination to retrospective cohort studies in administrative claims data and evaluate the impact of the normality of the random effects distribution for the cluster-level propensity score on estimation of the causal parameters. An extensive simulation study was conducted to study the robustness of the methods under different distributions of the random effects. We applied these methods to evaluate baseline prescription for medications for opioid use disorder among a cohort of patients diagnosed with opioid use disorder and adjust for baseline confounders using information obtained from an administrative claims database. We discuss future research directions in this setting to better address unmeasured confounding in the presence of disseminated effects.

Continuity of Care is Associated with Higher Appointment Adherence Among HIV Patients in Low Clinician-to-Patient Ratio Facilities in Western Kenya.

We sought to determine the relationship between continuity of care and adherence to clinic appointments among patients receiving HIV care in high vs. low clinician-to-patient (C:P) ratios facilities in western Kenya. This retrospective analysis included 12,751 patients receiving HIV care from the Academic Model Providing Access to Healthcare (AMPATH) program, between February 2016-2019. We used logistic regression analysis with generalized estimating equations to estimate the relationship between continuity of care (two consecutive visits with the same provider) and adherence to clinic appointments (within 7 days of a scheduled appointment) over time. Adjusting for covariates, patients in low C:P ratio facilities who had continuity of care, were more likely to be adherent to their appointments compared to those without continuity (adjusted odds ratio = 1.50; 95% confidence interval, 1.33-1.69). Continuity in HIV care may be a factor in clinical adherence among patients in low C:P ratio facilities and should therefore be promoted.

Seroprevalence of SARS-CoV-2 Antibodies in Rhode Island From a Statewide Random Sample.

Objectives. To characterize statewide seroprevalence and point prevalence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in Rhode Island.Methods. We conducted a cross-sectional survey of randomly selected households across Rhode Island in May 2020. Antibody-based and polymerase chain reaction (PCR)-based tests for SARS-CoV-2 were offered. Hispanics/Latinos and African Americans/Blacks were oversampled to ensure adequate representation. Seroprevalence estimations accounted for test sensitivity and specificity and were compared according to age, race/ethnicity, gender, housing environment, and transportation mode.Results. Overall, 1043 individuals from 554 households were tested (1032 antibody tests, 988 PCR tests). The estimated seroprevalence of SARS-CoV-2 antibodies was 2.1% (95% credible interval [CI] = 0.6, 4.1). Seroprevalence was 7.5% (95% CI = 1.3, 17.5) among Hispanics/Latinos, 3.8% (95% CI = 0.0, 15.0) among African Americans/Blacks, and 0.8% (95% CI = 0.0, 2.4) among non-Hispanic Whites. Overall PCR-based prevalence was 1.5% (95% CI = 0.5, 3.1).Conclusions. Rhode Island had low seroprevalence relative to other settings, but seroprevalence was substantially higher among African Americans/Blacks and Hispanics/Latinos. Rhode Island sits along the highly populated northeast corridor, making our findings broadly relevant to this region of the country. Continued monitoring via population-based sampling is needed to quantify these impacts going forward.

Modeling the Impact of Social Determinants of Health on HIV.

There is growing evidence for the key role of social determinants of health (SDOH) in understanding morbidity and mortality outcomes globally. Factors such as stigma, racism, poverty or access to health and social services represent complex constructs that affect population health via intricate relationships to individual characteristics, behaviors and disease prevention and treatment outcomes. Modeling the role of SDOH is both critically important and inherently complex. Here we describe different modeling approaches and their use in assessing the impact of SDOH on HIV/AIDS. The discussion is thematically divided into mechanistic models and statistical models, while recognizing the overlap between them. To illustrate mechanistic approaches, we use examples of compartmental models and agent-based models; to illustrate statistical approaches, we use regression and statistical causal models. We describe model structure, data sources required, and the scope of possible inferences, highlighting similarities and differences in formulation, implementation, and interpretation of different modeling approaches. We also indicate further needed research on representing and quantifying the effect of SDOH in the context of models for HIV and other health outcomes in recognition of the critical role of SDOH in achieving the goal of ending the HIV epidemic and improving overall population health.

Low Seroprevalence of SARS-CoV-2 in Rhode Island blood donors during may 2020 as determined using multiple serological assay formats.

BACKGROUND: Epidemic projections and public health policies addressing Coronavirus disease (COVID)-19 have been implemented without data reporting on the seroconversion of the population since scalable antibody testing has only recently become available. METHODS: We measured the percentage of severe acute respiratory syndrome- Coronavirus-2 (SARS-CoV-2) seropositive individuals from 2008 blood donors drawn in the state of Rhode Island (RI). We utilized multiple antibody testing platforms, including lateral flow immunoassays (LFAs), enzyme-linked immunosorbent assays (ELISAs) and high throughput serological assays (HTSAs). To estimate seroprevalence, we utilized the Bayesian statistical method to adjust for sensitivity and specificity of the commercial tests used. RESULTS: We report than an estimated seropositive rate of RI blood donors of approximately 0.6% existed in April-May of 2020. Daily new case rates peaked in RI in late April 2020. We found HTSAs and LFAs were positively correlated with ELISA assays to detect antibodies specific to SARS-CoV-2 in blood donors. CONCLUSIONS: These data imply that seroconversion, and thus infection, is likely not widespread within this population. We conclude that IgG LFAs and HTSAs are suitable to conduct seroprevalence assays in random populations. More studies will be needed using validated serological tests to improve the precision and report the kinetic progression of seroprevalence estimates.

Sex differences in health status, healthcare utilization, and costs among individuals with elevated blood pressure: the LARK study from Western Kenya.

BACKGROUND: Elevated blood pressure is the leading risk factor for global mortality. While it is known that there exist differences between men and women with respect to socioeconomic status, self-reported health, and healthcare utilization, there are few published studies from Africa. This study therefore aims to characterize differences in self-reported health status, healthcare utilization, and costs between men and women with elevated blood pressure in Kenya. METHODS: Data from 1447 participants enrolled in the LARK Hypertension study in western Kenya were analyzed. Latent class analysis based on five dependent variables was performed to describe patterns of healthcare utilization and costs in the study population. Regression analysis was then performed to describe the relationship between different demographics and each outcome. RESULTS: Women in our study had higher rates of unemployment (28% vs 12%), were more likely to report lower monthly earnings (72% vs 51%), and had more outpatient visits (39% vs 28%) and pharmacy prescriptions (42% vs 30%). Women were also more likely to report lower quality-of-life and functional health status, including pain, mobility, self-care, and ability to perform usual activities. Three patterns of healthcare utilization were described: (1) individuals with low healthcare utilization, (2) individuals who utilized care and paid high out-of-pocket costs, and (3) individuals who utilized care but had lower out-of-pocket costs. Women and those with health insurance were more likely to be in the high-cost utilizer group. CONCLUSIONS: Men and women with elevated blood pressure in Kenya have different health care utilization behaviors, cost and economic burdens, and self-perceived health status. Awareness of these sex differences can help inform targeted interventions in these populations.

Fixed choice design and augmented fixed choice design for network data with missing observations.

The statistical analysis of social networks is increasingly used to understand social processes and patterns. The association between social relationships and individual behaviors is of particular interest to sociologists, psychologists, and public health researchers. Several recent network studies make use of the fixed choice design (FCD), which induces missing edges in the network data. Because of the complex dependence structure inherent in networks, missing data can pose very difficult problems for valid statistical inference. In this article, we introduce novel methods for accounting for the FCD censoring and introduce a new survey design, which we call the augmented fixed choice design (AFCD). The AFCD adds considerable information to analyses without unduly burdening the survey respondent, resulting in improvements over the FCD, and other existing estimators. We demonstrate this new method through simulation studies and an analysis of alcohol use in a network of undergraduate students living in a residence hall.

Effects of social health insurance on access and utilization of obstetric health services: results from HIV+ pregnant women in Kenya.

BACKGROUND: Reducing maternal morbidity and mortality remains a top global health agenda especially in high HIV/AIDS endemic locations where there is increased likelihood of mother to child transmission (MTCT) of HIV. Social health insurance (SHI) has emerged as a viable option to improve population access to health services, while improving outcomes for disenfranchised populations, particularly HIV+ women. However, the effect of SHI on healthcare access for HIV+ persons in limited resource settings is yet to undergo rigorous empirical evaluation. This study analyzes the effect of health insurance on obstetric healthcare access including institutional delivery and skilled birth attendants for HIV+ pregnant women in Kenya. METHODS: We analyzed cross-sectional data from HIV+ pregnant women (ages 15-49 years) who had a delivery (full term, preterm, miscarriage) between 2008 and 2013 with their insurance enrollment status available in the electronic medical records database of a HIV healthcare system in Kenya. We estimated linear and logistic regression models and implemented matching and inverse probability weighting (IPW) to improve balance on observable individual characteristics. Additionally, we estimated heterogeneous effects stratified by HIV disease severity (CD4 < 350 as "Severe HIV disease", and CD4 > 350 otherwise). FINDINGS: Health Insurance enrollment is associated with improved obstetric health services utilization among HIV+ pregnant women in Kenya. Specifically, HIV+ pregnant women covered by NHIF have greater access to institutional delivery (12.5-percentage points difference) and skilled birth attendants (19-percentage points difference) compared to uninsured. Notably, the effect of NHIF on obstetric health service use is much greater for those who are sicker (CD4 < 350) - 20 percentage points difference. CONCLUSION: This study confirms conceptual and practical considerations around health insurance and healthcare access for HIV+ persons. Further, it helps to inform relevant policy development for health insurance and HIV financing and delivery in Kenya and in similar countries in sub-Saharan Africa in the universal health coverage (UHC) era.

Causal comparative effectiveness analysis of dynamic continuous-time treatment initiation rules with sparsely measured outcomes and death.

Evidence supporting the current World Health Organization recommendations of early antiretroviral therapy (ART) initiation for adolescents is inconclusive. We leverage a large observational data and compare, in terms of mortality and CD4 cell count, the dynamic treatment initiation rules for human immunodeficiency virus-infected adolescents. Our approaches extend the marginal structural model for estimating outcome distributions under dynamic treatment regimes, developed in Robins et al. (2008), to allow the causal comparisons of both specific regimes and regimes along a continuum. Furthermore, we propose strategies to address three challenges posed by the complex data set: continuous-time measurement of the treatment initiation process; sparse measurement of longitudinal outcomes of interest, leading to incomplete data; and censoring due to dropout and death. We derive a weighting strategy for continuous-time treatment initiation, use imputation to deal with missingness caused by sparse measurements and dropout, and define a composite outcome that incorporates both death and CD4 count as a basis for comparing treatment regimes. Our analysis suggests that immediate ART initiation leads to lower mortality and higher median values of the composite outcome, relative to other initiation rules.