Epidemiology
· 2025 Sep · PMID 40513048
·
Full text
Simulation studies are used to evaluate and compare the properties of statistical methods in controlled experimental settings. In most cases, performing a simulation study requires knowledge of the true value of the para...Simulation studies are used to evaluate and compare the properties of statistical methods in controlled experimental settings. In most cases, performing a simulation study requires knowledge of the true value of the parameter, or estimand, of interest. However, in many simulation designs, the true value of the estimand is difficult to compute analytically. Here, we illustrate the use of Monte Carlo integration to compute true estimand values in simple and more complex simulation designs. We provide general pseudocode that can be replicated in any software program of choice to demonstrate key principles in using Monte Carlo integration in two scenarios: a simple three-variable simulation where interest lies in the marginally adjusted odds ratio and a more complex causal mediation analysis where interest lies in the controlled direct effect in the presence of mediator-outcome confounders affected by the exposure. We discuss general strategies that can be used to minimize Monte Carlo error and to serve as checks on the simulation program to avoid coding errors. R programming code is provided illustrating the application of our pseudocode in these settings.
BACKGROUND: Migraine is a common risk factor for adverse perinatal outcomes, showing the importance of studying migraine in pregnancy. Despite the growing use of routinely collected administrative data in health research...BACKGROUND: Migraine is a common risk factor for adverse perinatal outcomes, showing the importance of studying migraine in pregnancy. Despite the growing use of routinely collected administrative data in health research, the validity of such data to detect migraine in pregnant populations is unestablished. We validated algorithms to identify a history of migraine among pregnant individuals using health administrative data and population-representative self-report data. METHODS: We included N = 8824 females in Ontario, Canada with a documented pregnancy with an estimated conception date from 1 September 2005 to 31 December 2021 who completed the Canadian Community Health Survey (CCHS) within 5 years before conception. We created algorithms using different combinations of diagnostic codes for headache disorders and migraine-specific drug claims with varying lookback periods before conception. We compared their performance to self-reported migraine diagnoses from the CCHS. Measures of validity were sensitivity, specificity, predictive values, and agreement. RESULTS: The prevalence of self-reported migraine from the CCHS was 18% (95% confidence interval [CI]: 16%, 19%). The prevalence using administrative data depended on the definition (range: 2%-25%). All algorithms had high specificity (81.7%-98.9%), while sensitivity varied (6.1%-53.2%). The algorithm requiring ≥2 physician visits or ≥1 hospitalizations or emergency department visits with diagnostic codes International Classification of Diseases, Ninth Revision: 346/International Classification of Diseases, Tenth Revision: G43, with a lifetime lookback, had high specificity (94.0%; 95% CI: 93.1%, 94.8%) and negative predictive value (86.3%; 95% CI: 85.0%, 87.6%) and modest sensitivity (30.4%; 95% CI: 27.3%, 33.6%) and positive predictive value (51.9%; 95% CI: 46.8%, 57.0%). Agreement was fair ( κ = 0.29; 95% CI: 0.25, 0.33). CONCLUSION: Longitudinally linked health administrative data are effective at identifying pregnant individuals with migraine, with high specificity and reasonable sensitivity.
Wolock CJ, Jacob S, Bennett JC
… +9 more, Elias-Warren A, O'Hanlon J, Kenny A, Jewell NP, Rotnitzky A, Cole SR, Weil AA, Chu HY, Carone M
Epidemiology
· 2025 Sep · PMID 40472281
·
Full text
BACKGROUND: For infectious diseases, characterizing symptom duration is of clinical and public health importance. Symptom duration may be assessed by surveying infected individuals and querying symptom status at the time...BACKGROUND: For infectious diseases, characterizing symptom duration is of clinical and public health importance. Symptom duration may be assessed by surveying infected individuals and querying symptom status at the time of survey response. For example, in a severe acute respiratory syndrome coronavirus 2 testing program at the University of Washington, participants were surveyed at least 28 days after testing positive and asked to report current symptom status. This study design yielded current status data: outcome measurements for each respondent consisted only of the time of survey response and a binary indicator of whether symptoms had resolved by that time. Such study design benefits from limited risk of recall bias, but analyzing the resulting data necessitates tailored statistical tools. METHODS: We review methods for current status data and describe a novel application of modern nonparametric techniques to this setting. The proposed approach is valid under weaker assumptions compared with existing methods, allows the use of flexible machine learning tools, and handles potential survey nonresponse. Our method relies on the assumption that the survey response time is conditionally independent of symptom resolution time within strata of measured covariates, and we propose an approach to assess the sensitivity of results to deviations from conditional independence. RESULTS: From the university study, we estimate that 19% of participants experienced ongoing symptoms 30 days after testing positive, decreasing to 7% at 90 days. We found the estimates to be more sensitive to violations of the conditional independence assumption at 30 days compared with 90 days. Female sex, fatigue during acute infection, and higher viral load were associated with slower symptom resolution. CONCLUSION: The proposed method and accompanying sensitivity analysis procedure provide tools for investigators faced with current status data.
Robertson SE, Rysavy MA, Blakely ML
… +2 more, Steingrimsson JA, Dahabreh IJ
Epidemiology
· 2026 Mar · PMID 40472279
·
Full text
We discuss generalizability analyses under a partially nested trial design, where part of the trial is nested within a cohort of trial-eligible individuals, while the rest of the trial is not nested. This design arises,...We discuss generalizability analyses under a partially nested trial design, where part of the trial is nested within a cohort of trial-eligible individuals, while the rest of the trial is not nested. This design arises, for example, when only some centers participating in a trial are able to collect data on non-randomized individuals, or when data on non-randomized individuals cannot be collected for the full duration of the trial. Our work is motivated by the Necrotizing Enterocolitis Surgery Trial, which compared initial laparotomy versus peritoneal drain for infants with necrotizing enterocolitis or spontaneous intestinal perforation. During the first phase of the study, data were collected from randomized individuals as well as consenting non-randomized individuals; during the second phase of the study, however, data were only collected from randomized individuals, resulting in a partially nested trial design. We propose methods for generalizability analyses with partially nested trial designs. We describe identification conditions and propose estimators for causal estimands in the target population of all trial-eligible individuals, both randomized and non-randomized, in the part of the data where the trial is nested while using trial information spanning both parts. We evaluate the estimators in a simulation study and provide an illustration using the Necrotizing Enterocolitis Surgery Trial study.
Causal mediation analysis examines the mechanism by which exposure affects outcome via mediators. In contrast to single-mediator scenarios, the presence of multiple ordered mediators introduces complex pathways and corre...Causal mediation analysis examines the mechanism by which exposure affects outcome via mediators. In contrast to single-mediator scenarios, the presence of multiple ordered mediators introduces complex pathways and corresponding path-specific effects, which are difficult to interpret due to the cross-world counterfactual definition. Path-specific effects also require convoluted and unverifiable assumptions for identification. This article proposes a framework of separable path-specific effects as an extension of the separable effect method to the case of multiple ordered mediators. Compared to the traditional approach, separable path-specific effects can be interpreted as the causal effects of several separated components on the outcome, facilitating a more intuitive understanding of underlying mechanisms. We elucidate the relationship between separable and traditional path-specific effects by demonstrating their equivalence under the individual-level isolation assumptions and identifying both effects under the finest fully randomized causally interpretable structured tree graph (FFRCISTG) model, which inherently makes individual-level isolation assumptions. Moreover, weakening the individual-level isolation assumptions to their population-level counterparts, separable path-specific effects remain identifiable under the FFRCISTG model. Under this causal model, the assumptions for identifying separable path-specific effects can be verified in future experiments, thereby addressing the problem of relying on unverifiable cross-world assumptions in the traditional method. We also discuss how this framework can detect violations of assumptions such as the presence of intermediate confounders and the misspecification of causal order among mediators. In summary, compared with the traditional path-specific effects method, the proposed separable method provides a more verifiable and interpretable approach for causal multiple mediation analysis.
Van Wyk H, Brouwer AF, Lee GO
… +5 more, Márquez S, Andrade P, Ionides EL, Coloma J, Eisenberg JNS
Epidemiology
· 2025 Sep · PMID 40459176
·
Full text
BACKGROUND: Pathogen transmission of an outbreak generally begins well before it is identified by a surveillance system, particularly for infectious diseases in which a high proportion of cases are subclinical, as is the...BACKGROUND: Pathogen transmission of an outbreak generally begins well before it is identified by a surveillance system, particularly for infectious diseases in which a high proportion of cases are subclinical, as is the case for arboviruses. We aimed to ascertain the most likely date of the primary case (the first infection, whether detected or not) in an outbreak. METHODS: Using data from a 2019 dengue outbreak in a rural, riverine town in Northwestern Ecuador, we investigated potential undetected dengue virus transmission before the outbreak detected in mid-May. The outbreak was preceded by four reported cases on 9 February, 13 February, 28 March, and 2 May. Using a hidden Markov model, we estimate the most likely date of the primary case for different assumed case reporting fractions. RESULTS: For all reporting fractions, the most likely primary case occurred near the 2 February candidate index cases, ranging from 7 February to 12 February, over 2 months before the main outbreak. Individual simulations showed that earlier and later primary cases were also possible. Our results suggest that the dengue virus was circulating in the community for around 3 months before the outbreak. CONCLUSIONS: Surveillance systems that can detect low-level transmission in the early stages of an outbreak can provide time to intervene before the exponential phase of the outbreak, with the potential to substantially reduce transmission and disease burden.
McGee EE, Hernán MA, Giovannucci E
… +4 more, Mucci LA, Chiu YH, Eliassen AH, Dickerman BA
Epidemiology
· 2025 Sep · PMID 40459173
·
Full text
BACKGROUND: Many organizations recommend lifestyle modifications for cancer survivors. Effect estimates for these interventions are often based on observational data and are challenging to interpret due to vaguely define...BACKGROUND: Many organizations recommend lifestyle modifications for cancer survivors. Effect estimates for these interventions are often based on observational data and are challenging to interpret due to vaguely defined causal questions, design-induced biases, and lack of comparability between individuals. METHODS: We outlined a three-step procedure to address these challenges: target trial specification, emulation, and modification to explore lack of comparability due to unmeasured confounding or positivity violations. We illustrated this procedure by specifying the protocols of two target trials that estimate the effects of adhering to seven physical activity and dietary recommendations and abstaining from alcohol on 20-year mortality among adults with breast or prostate cancer. We emulated these target trials using data from the Nurses' Health Study, Nurses' Health Study II, and Health Professionals Follow-up Study. RESULTS: In the main analysis, we included 9,107 adults (5,840 with breast cancer, 3,267 with prostate cancer) and 1,791 deaths occurred. After we modified the target trials, mortality risk differences (95% confidence intervals) comparing the physical activity and dietary intervention versus no intervention ranged from -4.8% (-7.5%, -2.3%) to -13.0% (-15.8%, -9.8%) for breast cancer and from -3.0% (-7.4%, 0.9%) to -12.8% (-17.6%, -7.6%) for prostate cancer. Risk differences comparing no alcohol consumption versus no intervention ranged from 1.3% (0.1%, 2.4%) to 3.6% (2.5%, 4.9%) for breast cancer and from -1.7% (-4.3%, 1.0%) to 6.4% (4.0%, 9.0%) for prostate cancer. CONCLUSIONS: We described a three-step procedure that improves the interpretability of observational estimates of the effects of lifestyle interventions and showed how estimates varied under different modifications.
Wei Y, Castro E, Yin K
… +7 more, Shtein A, Vu BN, Danesh Yazdi M, Li L, Liu Y, Peralta AA, Schwartz JD
Epidemiology
· 2025 Sep · PMID 40433992
·
Full text
BACKGROUND: Wildfire activity in the United States has increased substantially in recent decades. Smoke fine particulate matter (PM 2.5 ), a primary wildfire emission, can remain in the air for months after a wildfire be...BACKGROUND: Wildfire activity in the United States has increased substantially in recent decades. Smoke fine particulate matter (PM 2.5 ), a primary wildfire emission, can remain in the air for months after a wildfire begins, yet large-scale evidence of its health effects remains limited. METHODS: We obtained hospitalization records for the residents of 15 states between 2006 and 2016 from the State Inpatient Databases. We used existing daily smoke PM 2.5 estimations at 10-km 2 grid cells across the contiguous United States and aggregated them to ZIP codes to match the spatial resolution of hospitalization records. We extended the traditional case-crossover design, a self-controlled design originally developed for studying acute effects, to examine associations between 3-month average exposure to smoke PM 2.5 and hospitalization risks for a comprehensive range of cardiovascular (ischemic heart disease, cerebrovascular disease, heart failure, arrhythmia, hypertension, and other cardiovascular diseases) and respiratory diseases (acute respiratory infections, pneumonia, chronic obstructive pulmonary disease, asthma, and other respiratory diseases). RESULTS: We found that 3-month exposure to smoke PM 2.5 was associated or marginally associated with increased hospitalization risks for most cardiorespiratory diseases. Hypertension showed the greatest susceptibility, with the highest hospitalization risk associated with 0.1 µg/m 3 increase in 3-month smoke PM 2.5 exposure (relative risk: 1.0051; 95% confidence interval = 1.0035, 1.0067). Results for single-month lagged exposures suggested that estimated effects persisted up to 3 months after exposure. Subgroup analyses estimated larger effects in neighborhoods with higher deprivation level or more vegetation, as well as among ever-smokers. CONCLUSIONS: Our findings provided unique insights into medium-term cardiorespiratory effects of smoke PM 2.5 , which can persist for months, even after a wildfire has ended.
Liu EF, Jung S, Rudolph KE
… +5 more, Mujahid MS, Dow WH, Goin DE, Morello-Frosch R, Ahern J
Epidemiology
· 2025 Sep · PMID 40433985
·
Full text
BACKGROUND: In this article, we test the hypothesis that SARS-CoV-2 infection and the COVID-19 pandemic period had stronger adverse implications for perinatal outcomes among marginalized racial and ethnic groups in Calif...BACKGROUND: In this article, we test the hypothesis that SARS-CoV-2 infection and the COVID-19 pandemic period had stronger adverse implications for perinatal outcomes among marginalized racial and ethnic groups in California. METHODS: We used California birth certificates and hospital data from 2019 to 2021 to estimate marginal risk differences for SARS-CoV-2 infection and the COVID-19 pandemic period in relation to perinatal outcomes for Asian, Black, Hispanic, Multiracial, and White pregnant people using targeted maximum likelihood estimation. RESULTS: Among 849,401 deliveries, there were racial and ethnic disparities in the burden of SARS-CoV-2 infection and perinatal outcomes and in the magnitudes of risk associated with SARS-CoV-2 infection and the COVID-19 pandemic. Hispanic pregnant people had the highest incidence of SARS-CoV-2 infection. Asian and Black pregnant people had the greatest marginal risk differences for multiple outcomes, particularly outcomes already disproportionately experienced by these groups. CONCLUSIONS: Risks from SARS-CoV-2 infection and the COVID-19 pandemic period on perinatal outcomes were disproportionately experienced by marginalized racial and ethnic groups. Differential burdens of infection and larger risks experienced with pandemic exposures were associated with worse perinatal outcomes for Asian, Black, and Hispanic pregnant people in California compared with those for White pregnant people.
Epidemiology
· 2025 Sep · PMID 40433983
·
Full text
BACKGROUND: Longevity has stagnated during the last decade in the United States, but this stagnation has not been homogeneous. We aimed to explore the spatial variation of life expectancy by sex across commuting zones in...BACKGROUND: Longevity has stagnated during the last decade in the United States, but this stagnation has not been homogeneous. We aimed to explore the spatial variation of life expectancy by sex across commuting zones in the contiguous United States from 1990 to 2019. METHODS: We computed sex-specific life expectancy at birth for US commuting zones across six 5-year periods (1990-1994 to 2015-2019) and examined the spatial variability of life expectancy and clustering of baseline and changes in life expectancy during the study period. RESULTS: Overall life expectancy increased over time for both males and females and recently stagnated, while variability has increased for females. Regardless of sex, commuting zones with low baseline life expectancy that worsened over time were concentrated in the Appalachian region and Deep South. Areas with high baseline life expectancy and improved the most over time were scattered throughout the Midwest, Northwest, and West. CONCLUSION: The recent stagnation in life expectancy reflects wide spatial heterogeneity in changes in longevity. Growing spatial differences in longevity render males and females in the South, specifically the Appalachia and along the Mississippi River, to consistently live disproportionate short lives. Further studies should explore the contribution of different causes of death and the potential contextual drivers of these patterns.
Epidemiology
· 2025 Sep · PMID 40433968
·
Full text
Reproductive complications tend to recur. The risk of gestational diabetes is much higher in the second pregnancy if it occurred in the first. Such recurrence risks are regarded as reflecting heterogeneity among couples...Reproductive complications tend to recur. The risk of gestational diabetes is much higher in the second pregnancy if it occurred in the first. Such recurrence risks are regarded as reflecting heterogeneity among couples in their inherent risk. Pregnancy complications not only predict their own recurrence but have been shown to be associated with different later health problems like hypertension and heart disease. Epidemiologically considering reproductive history as a risk factor has been challenging, however, because women vary in their number of pregnancies and there's no obvious way to account for both prior occurrences and prior nonoccurrences. We propose a simple empirical Bayes approach, the Beta Approach for Risk Summarization (BARS). We apply BARS to retrospective data reported at enrollment in a large cohort, the Sister Study, to estimate propensity to gestational diabetes, and use that to predict subsequent occurrences of gestational diabetes based on successively updated pregnancy histories. We assess the calibration of our predictive model for gestational diabetes and demonstrate that it works well. We then apply the method to prospective data from the Sister Study, revisiting an earlier paper that linked gestational diabetes to the risk of breast cancer, but now using BARS and additional person time.
Tuohetasen S, Qu Y, Hopke PK
… +16 more, Zhang K, Liu Y, Lin S, Gu H, Wang X, Lau SSS, Lin X, Gao X, Wu Y, Zhou X, Lin Z, Zhang M, Sun Y, Liu X, Chen J, Zhang W
BACKGROUND: Although maternal exposure to artificial light at night has shown negative associations with pregnancy outcomes, its impact on the risk of congenital heart disease remains unclear. This study examined the ass...BACKGROUND: Although maternal exposure to artificial light at night has shown negative associations with pregnancy outcomes, its impact on the risk of congenital heart disease remains unclear. This study examined the association between maternal exposure to artificial light at night during pregnancy and occurrence of congenital heart disease in offspring, considering potential interactions with sociodemographics. METHODS: We included newborns diagnosed prenatally with congential heart disease and healthy volunteers from 21 cities in southern China. Using satellite data, we estimated annual exposure to artificial light at night at maternal residential addresses during pregnancy. We evaluated associations using marginal structural logistic models and assessed multiplicative and additive interaction between sociodemographics and light exposure. RESULTS: Each 1-unit increase in light at night during pregnancy was associated with an elevated risk of total congenital heart disease (odds ratio [OR]: 1.2, 95% confidence interval [CI]: 1.2, 1.3), and of almost all specific disease subtypes, in offspring. Using quartiles of light at night confirmed a monotonic dose-response relationship between exposure and disease. The association was more pronounced in severe disease. Some sociodemographic characteristics modified associations between light at night and congenital heart disease, with detrimental associations more pronounced among offspring of mothers with lower education (OR: 1.3, 95% CI: 1.2, 1.3), lower income (OR: 1.2, 95% CI: 1.1, 1.3), or being usual residents (OR: 1.3, 95% CI: 1.2, 1.4), based on the continuous model. CONCLUSIONS: Maternal exposure to artificial light at night during pregnancy was substantially associated with an elevated risk of congenital heart disease in offspring. This association was more pronounced among some sociodemographic groups.
Trial engagement effects are effects of trial participation on the outcome that are not mediated by treatment assignment. Most work on extending (generalizing or transporting) causal inferences from a randomized trial to...Trial engagement effects are effects of trial participation on the outcome that are not mediated by treatment assignment. Most work on extending (generalizing or transporting) causal inferences from a randomized trial to a target population has, explicitly or implicitly, assumed that trial engagement effects are absent, allowing evidence about the effects of the treatments examined in trials to be applied to nonexperimental settings. Here, we define novel causal estimands and present identification results for generalizability and transportability analyses in the presence of trial engagement effects. Our approach allows for trial engagement effects under assumptions of no causal interaction between trial participation and treatment assignment on the absolute or relative scales. We show that under these assumptions, even in the presence of trial engagement effects, the trial data can be combined with covariate data from the target population to identify average treatment effects in the context of usual care as implemented in the target population (i.e., outside the experimental setting). The identifying observed data functionals under these no-interaction assumptions are the same as those obtained under the stronger identifiability conditions that have been invoked in prior work. Therefore, our results suggest a new interpretation for previously proposed generalizability and transportability estimators. This interpretation may be useful in analyses under causal structures where background knowledge suggests that trial engagement effects are present but interactions between trial participation and treatment are negligible.