Hepatitis A virus (HAV) is an RNA virus transmitted by fecal-oral contamination. HAV is usually asymptomatic in young children who shed the virus, leading to infection in susceptible adults who develop symptomatic hepatitis. This article reviews the epidemiology of HAV infection in Alaska and results of hepatitis A vaccine (HepA) studies conducted in Alaska, including evidence that routine early childhood HepA vaccination can eliminate transmission in an area that previously experienced repeat outbreaks. Historically, Alaska had periodic large HAV outbreaks. Alaska was the site of both Phase 3 and 4 clinical trials of the HepA vaccine, one of which demonstrated that only one dose of HepA vaccine administered to susceptible persons in communities experiencing an active outbreak could stop transmission in 4 to 8 weeks. Evidence indicates that the HepA vaccine offers protection for at least 25 years for most individuals, however, data on the durability of immunity beyond this period is lacking, particularly among people vaccinated as young children. We discuss future considerations for continued control of hepatitis A in Alaska.
Epidemiology has achieved substantial methodological refinement in recent decades, yet its social resonance has not always kept pace. This essay reflects on tendencies within influential sectors of the field toward methodological sophistication that, while yielding genuine intellectual advances, can unintentionally distance epidemiology from its civic and historical roots. By privileging what is analytically tractable, such developments may render broader contextual forces and socially patterned differences between individuals around population averages less visible. Drawing on traditions in social epidemiology, the essay advances a central argument: a substantial share of individual heterogeneity is intrinsically contextual. Differences between individuals are not pre-social deviations to be averaged away, but structured expressions of social, spatial, institutional, and historical contexts. From this perspective, the central challenge facing contemporary epidemiology is not primarily statistical but metaethical. It concerns how analytical choices shape interpretation, how values are embedded in measurement practices, and how these practices delimit the social purposes epidemiology is understood to serve. Crucially, epidemiology is not only a science of causal explanation, but also a discipline concerned with mapping, monitoring, and documenting how health and harm are distributed within populations over time. Even when major determinants of ill health are well established, epidemiology retains a core role in tracking how inequalities persist, change, or respond to policy. Rather than rejecting modern tools, the essay calls for a pluralistic and contextually grounded epidemiology that reconnects analytical rigor with social meaning. By treating individual heterogeneity as contextual rather than residual, epidemiology can reconcile population health and precision approaches and more fully realize its dual role as a scientific enterprise and a civic practice oriented toward equity.
Smoking, both active and passive, has been associated with an increased risk of numerous cancers worldwide, yet its impact on breast cancer remains a subject of debate. While many studies have explored the link between smoking and breast cancer risk, the findings have been inconsistent and often contradictory. To clarify the nature and strength of the association between smoking exposure (current, former, and passive) and breast cancer, we conducted an umbrella review of meta-analyses of observational studies (PROSPERO registration number: CRD42024610213). We conducted a systematic literature search in PubMed, Scopus, and Web of Science from inception to October 24, 2024, and manually screened reference lists. Systematic reviews and meta-analyses of observational studies (cohort, case-control, and/or cross-sectional) examining the association between smoking (current, former, and passive) and breast cancer were included. Predefined evidence classification criteria were applied to assess the credibility of associations, graded as convincing (class I), highly suggestive (class II), suggestive (class III), weak (class IV), or no evidence (class V). Data were extracted for effect estimates, heterogeneity (I²), 95 % prediction intervals, small study effects, and excess significance bias, using random-effects models and AMSTAR 2 to evaluate methodological quality. We identified 1095 articles, of which eight meta-analytic systematic reviews, encompassing 142 unique primary studies, were eligible for inclusion. These 142 articles provided data for six meta-analyses, covering former smoking, current smoking, and passive smoking (median cases = 1009; median population = 10,588). Convincing evidence (class I) was found for former smoking, with a summary relative risk (RR) of 1.13 (95 % CI 1.10-1.16). Highly suggestive evidence (class II) was observed for current smoking (RR 1.70, 95 % CI 1.66-1.74) and passive smoking (RR 1.53, 95 % CI 1.34-1.74). In our sensitivity analyses by study design, current smoking in cohort studies demonstrated a robust association with breast cancer risk (OR = 1.67, 95 % CI 1.62-1.71, Class II). Former smoking, analyzed in cohort studies, showed a modest but convincing association (OR = 1.10, 95 % CI 1.08-1.13, Class I), while in case-control studies, former smoking indicated a suggestive association with breast cancer risk (OR = 1.48, 95 % CI 1.19-1.82, Class III). Both former and current smoking, as well as passive exposure to smoke, were associated with an elevated risk of breast cancer. The association with former smoking may reflect long-lasting biological damage, while current and passive smoking likely act through similar carcinogenic pathways. Further studies are needed to establish causality and to account for potential confounding factors such as genetic predisposition and lifestyle. These findings highlight the importance of smoking cessation programs and stricter regulations on secondhand smoke to mitigate the global burden of breast cancer.
Epidemiology has long been central to public health, guiding our understanding of the distribution and determinants of disease. As the field has evolved-from John Snow's cholera investigations to large-scale cohort studies and causal inference frameworks-it now faces a transformative juncture with the advent of artificial intelligence/machine learning (AI/ML). These technologies offer unprecedented opportunities to improve data measurement, inference, and population health insights, yet also pose methodological and ethical challenges. Anchored by the core epidemiologic domains of study population, measurement, and inference, we examine how epidemiologists can use AI/ML effectively. We consider the importance of careful population definition, informed sampling, and external validation to ensure generalizability and minimize bias when AI/ML is used. We also explore the need for rigorous assessment of data quality and model reliability, which strengthens the case for conceptual frameworks in guiding interpretation of scientific investigations. To realize AI/ML's potential, epidemiology must adapt its training, invest in infrastructure, and promote interdisciplinary collaboration. Doing so will ensure that epidemiologic science remains robust, reproducible, and relevant in a rapidly evolving informational landscape. This moment calls for a strategic integration of AI/ML into the fabric of epidemiologic practice and training to advance both science and public health.
The World Trade Center Health Program (WTCHP) plays a critical role in medical monitoring and treatment to those exposed to the terrorist attacks of September 11, 2001 (9/11). We investigated the association of WTCHP membership with mortality risk among 9/11 responders while controlling for comorbidities using inverse probability weighting. We prospectively analyzed 28,430 9/11 responders, followed from the time of their enrollment into the WTCHP or the WTC Health Registry, through 2020. NDI linkage provided death data. Non-cancer comorbidities were self-reported physician-diagnosis and cancer was identified through cancer registry linkage. We estimated the adjusted hazard ratio (aHR) with 95 % confidence interval (CI) for the association between WTCHP membership and all-cause and cause-specific mortality using Cox proportional hazards models and cause-specific hazard regression models, respectively. A total of 1657 deaths were identified over 444,425 person-years of follow-up. Compared to non-members, WTCHP members had a lower risk of all-cause mortality (aHR=0.87; 95 % CI=0.77-0.98) and smoking-related mortality (aHR=0.83; 0.69-0.99) after adjusting for demographics, WTC exposure, and weights of comorbidities. With the membership-sex interaction included, reduced risk of all-cause mortality remained statistically significant among males only (aHR=0.85; 0.75-0.96). Cancer- and heart-related mortality risk were not significantly different between WTCHP members and non-members. This study found that WTCHP membership may reduce risks of all-cause and smoking-related mortality among 9/11 responders, even after accounting for underlying medical conditions, underscoring the importance of comprehensive health monitoring and treatment services for disaster-relief workers.
To improve the identification of cerebral palsy cases in administrative health data. We included all children in a population-based cerebral palsy registry in Quebec, Canada, born from 1999 through 2002, and a sample of children without cerebral palsy. Population-based hospitalization and physician billing records through 2012 were obtained for all children. We used logistic regression to model the probability of cerebral palsy, using International Classification of Diseases codes for related diseases. We reported receiver operating characteristic (ROC) and precision-recall (PR) curves, and compared the accuracy to that of existing algorithms. We also reported the accuracy of cerebral palsy codes by age, data source, and gestational age at birth. The area under the ROC and PR curves of our model were 0.98 (95 % CI: 0.97-0.99) and 0.73 (95 % CI: 0.63-0.79), respectively. Cut-offs with a similar specificity to existing algorithms yielded sensitivities that were 1-14 %age-points higher. The sensitivity of cerebral palsy codes was higher (and the specificity was lower) with longer follow-up times since birth, when using both hospitalization and billing records, and among children born preterm. Our model improved identification of cerebral palsy cases in administrative data, but residual misclassification remained.
Research is needed to understand racial and ethnic differences in symptoms of depression. Unfortunately, most studies examine these differences using ethnically-stratified, mono-racial categories (e.g., non-Hispanic Black), producing inaccurate estimates due to heterogeneity across racial and ethnic identities. In this study, we compare different operationalizations of race and ethnicity in predicting symptoms of depression within a diverse cohort of young adults. We analyzed cross-sectional data from n = 2340 young adults (mean age: 21 yrs. (SD: 1.6), 59% female, 38% Hispanic, 29% low SES) via the Texas Adolescent Tobacco and Marketing Surveillance (TATAMS) study. Random forest models evaluated prediction and identified sociodemographic features for symptom classification. We modeled eight operationalizations of race and ethnicity, four applying mutually-exclusive categorizations (e.g., non-Hispanic White) and four allowing for overlapping categorizations. Models comprised a) race and ethnicity, alone, and b) included SES, age, sex, and geography. Models with race and ethnicity, alone, demonstrated poor prediction of symptoms of depression (Sn range: 0.33-0.48). Including other sociodemographic features, prediction remained poor for symptoms of depression (Sn range: 0.15 - 0.62). Prediction decreased upon separation of Hispanic ethnicity and 'Other' race (e.g. non-Hispanic Asian). SES was the most influential feature across all models. Race and ethnicity poorly predict symptoms of depression, particularly when using standard OMB categories (i.e., ethnically-stratified, mono-racial). Models allowing for overlapping racial and ethnic identities outperformed those using mutually-exclusive categorizations. Results suggest that health equity research should account for racial and ethnic heterogeneity and consider SES in addressing racial and ethnic differences in mental health among young adults at the population level.
Birth defects are a leading cause of infant mortality in the United States, but little is known about causes of many types of birth defects. Spatiotemporal disease mapping to identify high-prevalence areas is a potential strategy to narrow the search for potential environmental and other causes that aggregate over space and time. We described the spatial and temporal trends of the prevalence of birth defects in North Carolina during 2003-2015, using data on live births obtained from the North Carolina Birth Defects Monitoring Program. By employing a Bayesian space-time Poisson model, we estimated spatial and temporal trends of non-chromosomal and chromosomal birth defects. During 2003-2015, 52,524 (3.3 %) of 1598,807 live births had at least one recorded birth defect. The prevalence of non-chromosomal birth defects decreased from 3.8 % in 2003-2.9 % in 2015. Spatial modeling suggested a large geographic variation in non-chromosomal birth defects at census-tract level, with the highest prevalence in southeastern North Carolina. The strong spatial heterogeneity revealed in this work allowed us to identify geographic areas with higher prevalence of non-chromosomal birth defects in North Carolina. This variation will help inform future research focused on epidemiologic studies of birth defects to identify etiologic factors.
The study examined whether poor emotional control shown at a young age increases the risk of having stroke and/or ischemic heart disease (IHD) later in life among Swedish men. The risk of stroke and IHD was also compared between smokers and non-smokers. The study was based on data from an historical nationwide survey of Swedish men aged 18-20 years who underwent mandatory military conscription. Psychologists assessed emotional control based on semi-structured interviews with the conscripts. A total of 45,169 men were followed up for a first event of ischemic stroke and IHD approximately 40 to 70 years of age. Cox' regression estimated hazard ratios (HR) and 95% confidence intervals (95%CI). Poor emotional control was associated with increased stroke risk, which remained statistically significant after adjusting for covariates (HR: 1.22, 95%CI: 1.08-1.39). No significant association with IHD was found after adjusting for covariates. In the comparison between smokers and non-smokers (at military conscription), poor emotional control was significantly associated with later stroke risk only among the smokers (HR: 1.31, 95%CI: 1.12-1.54). Findings of this study do not provide evidence for an effect of poor emotional control on long-term risk of stroke and IHD independent of smoking.
When decision makers use evidence from a randomized trial to inform population-level decisions, the target population they envision rarely aligns with the population of individuals who enrolled in the trial. Here, we extend inferences from the VALIDATE-SWEDEHEART randomized trial (hereafter, the index trial), which compared the effects of bivalirudin and heparin during percutaneous coronary intervention on the risk of death, reinfarction, and bleeding, to two clinically relevant target populations: first, the trial-eligible population of individuals eligible for the index trial regardless of enrollment, and second, the treatment-candidate population of individuals who are considered candidates for bivalirudin and heparin under routine care, regardless of eligibility for the index trial. Using data from the index trial, we fit logistic regression models for the outcome at 180 days in each group based on assigned treatment. We then standardized risk estimates to the baseline covariate distribution of the trial-eligible and treatment-candidate target populations, which were characterized using data from Swedish healthcare registries. The estimated risk difference comparing bivalirudin vs. heparin was -1.1% (-3.1%, 0.9%) in the trial-eligible population and -1.0% (-3.0%, 1.0%) in the treatment-candidate population. The corresponding risk ratios were 0.92 (0.80, 1.07) and 0.93 (0.80, 1.07), respectively, aligning closely with estimates from the index trial. Absolute risks in each treatment group were, however, between 0.8 and 1.2 percentage points higher in comparison with the index trial. Estimated risk ratios for the broader trial-eligible and treatment-candidate populations generally align with the findings from the index trial. While trials provide essential evidence for healthcare, questions often arise about wider, clinically relevant populations beyond the population of trial participants. By leveraging data from trials and observational data sources, we can attempt to address questions in these wider target populations.
The World Health Organization's 2030 Sustainable Development Goals include reducing the risk of fetal death. Even in high-income countries such as the United States (US), fetal deaths remain under-counted, with reporting showing variable quality across place and time. In the US, a uniform national definition of fetal death does not exist. Scant work characterizes whether, and to what extent, definitional changes in fetal death in US states over time affect fetal death reporting as well as counting of live births among similarly very frail (i.e., periviable [born <26 weeks]) infants. We aimed to (i.) identify state-level changes in fetal death reporting guidelines or definitions for all 50 US states from 1995 to 2020 and (ii.) examine whether counts of fetal deaths, periviable births, and neonatal deaths among periviable births shifted in the years following such changes. We retrieved data for all 50 US states from 1996 to 2021 for this descriptive analysis (n = 642,551 fetal deaths, n = 420,000 periviable births, n = 195,663 neonatal deaths among periviable births). We reviewed the fetal death user guides for state changes in reporting guidelines and conducted an internet search to find other state changes in the definition of fetal death. Next, we modeled fixed-effects linear regressions to examine associations between changes in fetal death reporting guidelines and our three outcomes. Over the test period, 12 states changed their definition of fetal death. Regression results show increases in the counts of fetal deaths, periviable births, and neonatal deaths among periviable births following changes in reporting guidelines. These increases followed any change in reporting guidelines-whether perceived as a more inclusive or more restrictive change. Results hold across a range of alternative specifications. Our findings cohere with the notion that any state-level change in fetal death definitions corresponds with broader efforts to improve data collection and reporting protocols among not only fetal deaths but also periviable births. The fact that we observe such associations should encourage strategies to control for such "data breaks" for scientists and officials concerned with fetal death and/or periviable birth.
This study aimed to develop a cluster-based measure of multiple co-occurring social determinants of health by applying unsupervised machine learning to a population-based cohort, offering a data-driven approach to organize complex social exposures. Unsupervised clustering was applied to a population-based cohort of Ontario respondents to six-cycles of the Canadian Community Health Survey (2001-2012) linked to the Canadian census and vital statistics data. Clusters were evaluated using internal metrics, visualization techniques, descriptive analysis and theoretical considerations to determine the optimal number of clusters. Sensitivity analyses were integrated across the iterative clustering process. Premature mortality rates were generated assess validity. Optimal clustering solutions included 4-clusters and 6-clusters. Both cluster solutions revealed distinct social typologies. The 6-cluster solution offered greater granularity and theoretical interpretability. The 4-cluster solution showed greater heterogeneity within certain marginalized groups. Premature mortality rates differed meaningfully across clusters, supporting the clustering approach in capturing risk associated with social exposure. Unsupervised machine learning methods identified meaningful population subgroups reflecting complex patterns of social exposures. This approach offers a flexible, data-driven method for characterizing social exposures that can be considered alongside theoretical frameworks and used for equity monitoring, intervention planning and policy development.
The atherogenic index of plasma (AIP), a novel biomarker reflecting atherosclerosis burden, has been associated with an increased risk of metabolic dysfunction-associated steatotic liver disease (MASLD). However, the impact of long-term AIP trajectory patterns on MASLD development remains unclear. This retrospective longitudinal study enrolled 13,211 adults who received serial health screenings between January 2017 and November 2024. AIP was derived using the formula: log (triglycerides/HDL-C). MASLD incidence among AIP subgroups was compared via Kaplan-Meier analysis. Restricted cubic splines evaluated potential nonlinear associations between AIP and MASLD risk. Latent class trajectory modeling was used to identify distinct AIP trajectory patterns over time. Over a median follow-up of 2173 days, 2744 cases of MASLD progression were documented. A 1-SD rise in AIP corresponded to a 184% increased MASLD risk. Quartile-based analyses yielded consistent findings. Trajectory modeling stratified participants into low-stable, medium-stable, and high-stable groups. Compared with the low-stable group, the medium-stable and high-stable groups exhibited significantly increased risks of MASLD, with hazard ratios (HRs) of 2.75 (95% CI: 2.43-3.11) and 4.76(95% CI: 4.14-5.47), respectively. Both elevated baseline AIP levels and sustained high-stable AIP trajectories were strongly associated with an increased risk of MASLD progression. Continuous monitoring of AIP may enable early risk stratification and inform targeted preventive strategies.
Intersectionality, rooted in Black feminist scholarship, offers a critical lens for examining how interlocking systems of oppression shape the health of individuals and populations at the intersection of multiple social identities. Despite growing interest in quantitative intersectionality research within epidemiology and public health, researchers still face substantial measurement- and analysis-related challenges. Most research to-date has relied on static, unidimensional, individual-level measures of social identities as proxies for exposure to systems of oppression. Recently, researchers have called for the development of lifecourse-informed, multidimensional, and structurally-focused measures of both social identities and systems of oppression. Analytic methods for quantitative intersectionality research, including Intersectional Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (I-MAIHDA), have expanded opportunities to examine the role of multiple contexts and structural-level exposures in shaping intersectional health inequities both cross-sectionally and across the lifecourse, yet use of causal inference methods to quantify the potential impacts of specific structural-level interventions remains scarce. Importantly, key epidemiologic considerations, including measurement, selection, and confounding bias, remain under-examined and under-theorized, which is especially concerning given the complexities of quantitative intersectionality research. Addressing these important measurement- and analysis-related challenges is essential for generating valid and actionable evidence to guide efforts to advance health equity.
Drug use disorders (DUDs) are emerging global public health challenges. Here we investigated the global and regional estimates of the prevalence and burden of DUDs, including amphetamine, cannabis, cocaine and opioid use disorders, from 1990 to 2023 for 204 countries and territories by using the Global Burden of Disease Study 2023. Overall, trends in global age-standardized disability-adjusted life-years of DUDs increased from 169.3 (95% uncertainty interval (95% UI), 134.4-203.9) per 100,000 people in 1990 to 212.0 (95% UI, 179.2-245.6) in 2023. In 2023, both prevalence and burden of DUDs were higher in high-income countries, particularly in the USA. The most prevalent DUDs in 2023 were cannabis use disorder (age-standardized prevalence, 270.8 (95% UI, 201.7-350.0) per 100,000 people) and opioid use disorder (205.9 (95% UI, 178.7-235.0)). Particularly, opioid use disorder showed a nearly twofold increase in prevalence and burden between 1990 and 2023. In 2023, compared with countries where cannabis use was illegal, countries permitting both recreational and medical cannabis use had higher prevalence rates for all types of DUDs. Proactive and effective policies are essential to mitigate the increasing global burden of DUDs.
To synthesis the available research on the association between hypertensive disorders during pregnancy (HDP) and adverse cognitive outcomes in children. PubMed, MEDLINE, Embase, Scopus, CINAHL, and PsycInfo were searched from inception to February 2026 to identify relevant studies. Effect estimates and 95% confidence intervals (CIs) were extracted and pooled using inverse variance-weighted random-effects meta-analysis. Thirty-nine observational studies published between 1982 and 2026, covering over 12.5 million mother-offspring pairs, were included in the final analysis. The findings indicate that HDP were associated with a 39% increased risk of Intellectual Disability (ID) (OR = 1.39, 95% CI = 1.30-1.47) and a 51% increased risk of a low Mental Developmental Index (MDI) (OR = 1.51, 95% CI = 1.09-2.09) in children. However, we observed no significant difference in mean Intelligence Quotient (IQ) scores between children of mothers with and without HDP (MD = -0.09, 95% CI = -0.45-0.26). The findings highlight that exposure to HDP is associated with an increased risk of ID and low MDI in offspring. Implementing targeted early screening and intervention programs for children of mothers with HDP is essential to address these developmental challenges.
To evaluate the effectiveness of telephone counselling, including quitlines, for tobacco cessation and to compare outcomes between randomized controlled trials (RCTs) and observational studies across diverse populations and settings. Systematic review and meta-analysis METHODS: We systematically searched PubMed, EMBASE, Scopus, Cochrane Library, and ScienceDirect for English-language studies (2002-2025). Eligible studies included RCTs, non-randomized trials, and observational studies of individuals using any form of tobacco. The primary outcome was continuous abstinence; secondary outcomes included 7-day and 30-day point prevalence abstinence. Risk of bias was assessed using the Cochrane RoB-2 and Newcastle-Ottawa Scale; certainty of evidence was rated with GRADE. PROSPERO registration: CRD42023418243. Thirty-six studies (RCTs=25; observational=11) were included. Across RCTs, telephone counselling increased continuous abstinence by 43% (RR = 1.43; 95% CI: 1.11-1.84; I² = 48%). 7-day PPA improved by 45% (RR = 1.45; 95% CI: 1.13-1.86; I² = 74%), while 30-day PPA (4 RCTs) also showed benefit (RR = 1.37; 95% CI: 1.14-1.65). Subgroup analyses demonstrated consistent effects across biochemical and self-reported verification. Observational studies reported 7-day PPA ranging from 2.6%-67.4%, 30-day PPA around 20.5%, and mean quit rates of 35.5%, with higher cessation among those completing ≥ 3 counselling sessions. Meta-regression indicated that differences between RCT and observational estimates were driven by follow-up duration, not study design. Telephone counselling significantly increases quit rates across populations, with evidence from both trial and real-world settings. Scaling up quitline services, particularly in low- and middle-income countries, should be prioritized to strengthen global tobacco control efforts.
To assess the association between parental age differences (PAD) and the risk of spontaneous abortion (SAB). A large-scale population-based retrospective cohort study was conducted on women aged 20-49 years who participated in the National Free Pre-pregnancy Check-ups Project and became pregnant during 2010-2018. PAD was defined as paternal age minus maternal age. To overcome the perfect collinearity between PAD and parental ages, we employed a two-stage instrumental variable approach. In the first stage, county-level average paternal and maternal ages were used as instruments to predict individual parental ages. In the second stage, inverse probability weighting based on generalized linear mixed models (GLMM) with county-level random intercepts was employed to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for SAB associated with PAD, adjusting for the predicted parental ages and a comprehensive set of covariates. Generalized additive mixed models (GAMM) with penalized splines were used to explore nonlinear exposure-response relationship. The overall incidence of SAB was 2.60%. Compared with couples with no age difference (PAD=0 years), the ORs for SAB were 1.93 (95% CI: 1.68-2.21) for a PAD of < -10 years, 1.27 (1.20-1.33) for -10 to -6 years, 1.07 (1.05-1.08) for -5 to -1 years, 1.00 (1.00-1.01) for 1 to 5 years, 1.05 (1.03-1.06) for 6 to 10 years, and 1.12 (1.09-1.16) for > 10 years. A V-shaped nonlinear relationship was observed between PAD and the risk of SAB (Pnon-linear <0.001), with a steeper increase in risk when the female partner was older than the male partner. Larger PAD was associated with an increased risk of SAB, especially in couples where the mother was older than the father.
Using the Target Trial Emulation Framework, we evaluated the impact of initiating dolutegravir versus efavirenz on 12- and 24-month weight, body mass index (BMI), blood pressure (BP), and incident hypertension among treatment-naïve individuals in Johannesburg, South Africa from 2019 to 2022. We used linear models to estimate the mean difference in weight, BMI, BP and a log-binomial model to estimate the causal risk difference of incident hypertension, adjusting for patient characteristics via inverse probability weighting. Among 2930 people initiating treatment from 2019 to 2022, 1847 initiated dolutegravir and 1083 initiated efavirenz. At 12-months, mean difference comparing dolutegravir to efavirenz in weight was 2.9 kg (95% Confidence Interval (CI): -0.3, 5.5), BMI was 0.8 kg/m2 (95% CI: -0.3, 1.9), diastolic BP was 1.6 mmHg (95% CI: -0.7, 3.9) and systolic BP was 3.9 mmHg (95% CI: 1.2, 6.6). Risk of incident hypertension rose by 35% (95% CI: 0.04, 0.5). At 24-months, mean weight difference was 1.9 kg (95% CI: -1.3, 5.1), BMI was 0.6 kg/m2 (95% CI: -0.6, 1.9), diastolic BP was -0.4 mmHg (95% CI: -1.8, 5.1) and systolic BP was 1.7 mmHg (95% CI: -1.8, 5.1). Risk of incident hypertension rose by 22% (95% CI: -0.1, 0.4). Dolutegravir was associated with greater increases in weight, systolic BP, and incident hypertension over 24-months, particularly in the first 12-months. Future research is needed to determine whether this reflects a direct effect of dolutegravir or the weight-suppressing effects of efavirenz.
A critical function of public health is to monitor diseases that impede quality of life and burden affected communities. The Diabetes in Children, Adolescents and Young Adults (DiCAYA) Network aims to advance disease monitoring for diabetes using multi-site electronic health record (EHR) data. This work involved validating and refining case definitions for accurate identification of type 1 and type 2 diabetes cases to estimate incidence and prevalence of diabetes in children, adolescents, and young adults through age 44 years. In this essay, we describe the challenges experienced by the Network and lessons learned. Challenges included accessing EHR data, harmonizing EHR data from heterogeneous health systems to a common data model, and developing methods to account for bias introduced by the non-representativeness of health care utilization data. Lessons learned included approaches for data quality assessment, bias correction, and scalability. As the US continues to evolve its public health data systems and its approach to chronic disease monitoring, the DiCAYA Network offers guidance on factors for success as well as pitfalls to avoid.