Sexual minority (SM) people may be at increased risk of "unintended" pregnancies, but limited research has examined potential disparities. We examined the intention status of lifetime pregnancies by sexual orientation. We pooled data from the Nurses' Health Study 3 and Growing Up Today Study (n = 36 967 pregnancies 1978-2024). Using multinomial models, we estimated relative risk ratios (RRRs) for pregnancy intention (actively trying/wanted then or sooner, not trying but glad, wanted later but not then, and unwanted then and in the future) among SM groups (completely heterosexual with same-sex experience, mostly heterosexual, bisexual, and lesbian/gay) versus completely heterosexual participants with no same-sex experience. Model predicted probabilities show that 15%-33% of pregnancies within each sexual orientation group were wanted later/unwanted. Compared to pregnancies to completely heterosexual participants with no same-sex experience, all SM groups had more pregnancies that were wanted later (RRRs from 1.22 [completely heterosexual with same-sex experience] to 2.28 [bisexual]) and unwanted then and in the future (RRRs ranging from 1.74 [completely heterosexual with same-sex experience] to 5.23 [bisexual participants]), compared to pregnancies occurring when participants were actively trying/wanted the pregnancy then or sooner. Our findings suggest that pregnancies among SM groups disproportionately are unwanted or occur too early.

We study life-expectancy loss (LEL) from the COVID-19 pandemic in the U.S. from March 2020 through April 2023. We use individual-level death-certificate data and the cohort-based approach to life expectancy. We also examine how LEL varies with age, gender, race/ethnicity, county-level socio-economic status (county-SES). The U.S. recorded 983 000 COVID-19 deaths. The COVID-19 Mortality Rate (COVID-MR, the fraction of the population who died of COVID-19) was 0.30%. An upper bound estimate of years of life lost per person (YLL) is 14.6 years (thus population loss of 14.32 million life-years). Estimated LEL based on this YLL estimate was 0.043 years (16 days). However, LEL rose rapidly with age averaging 8 weeks for ages 75+. LEL was substantially higher for Native Americans, Blacks, and Hispanics than for Whites at all ages. LEL also increased as county-SES decreased. Our cohort-based LEL estimates are a small fraction of period-based estimates during the pandemic, which were based on mortality in a particular year and assumed the pandemic would continue indefinitely with no change in severity. Using current life expectancies, cohort-based LEL for the COVI|D-19 pandemic was a fraction of LEL for the 1918 influenza pandemic (14 weeks), but well above three recent, severe influenza seasons (average of 0.12 weeks).

Prevalence estimates for opioid use disorder (OUD) are essential for guiding investments in prevention and treatment. Previous estimates have been produced using modeling approaches including capture-recapture (CRC). Empirical evaluations of CRC are needed to assess population size impacts of methodological practices utilized. We linked administrative data sources including New York State (NYS) adults with OUD by indication between January 1-December 31, 2020 and applied Bayesian CRC methods to evaluate how methodological practices impact OUD prevalence: (1) using sensitive versus specific case definition characteristics to identify OUD, (2) including versus excluding inferred referrals from emergency medical services to healthcare settings, and (3) including versus excluding overdose mortality as a source list. We estimated OUD prevalence overall and by demographic characteristics and treatment coverage with medications for OUD (MOUD). Across methodological scenarios, OUD prevalence estimates ranged from 4.6%-9.9%. The largest estimate resulted from a model in which the more sensitive OUD case definition was applied, while the smallest resulted from a model using the more specific case definition. An estimated 4.2% of NYS adults had OUD in 2020, with 20% receiving MOUD. Findings demonstrate that CRC methodological decisions can meaningfully influence OUD prevalence estimates, with recommended methodological practices provided.

Three indicators of severe maternal morbidity (SMM) available in US birth certificate data are increasingly used in research, yet prior studies suggest they are poorly ascertained. It is unclear whether low validity of these data threatens causal inference in evaluations of policies or shocks that vary across states and over time, including abortion bans enacted after the Supreme Court's Dobbs decision. Using birth certificate and hospital discharge data from 12 states (2018-2021), we assessed data quality for three SMM indicators (hysterectomy, blood transfusion, and eclampsia) and a composite measure and compared state-year trends across six states that subsequently imposed abortion bans shortly after the Dobbs decision and six that did not, including tests of the parallel trends assumption. Compared with hospitalization data, birth certificates under-ascertained hysterectomy, blood transfusion, and the composite measure while over-ascertaining eclampsia. Composite SMM rates appeared higher in abortion-ban states using birth certificate data but were similar or lower in hospitalization data. Over time, birth certificate ascertainment improved in states with bans but remained flat elsewhere, potentially introducing data artifacts. Violations of the parallel trends assumption were detected in hospitalization but not birth certificate data. Reliance on birth certificate data for policy evaluation can produce erroneous inferences.

Estimating, understanding, and communicating uncertainty is fundamental to statistical epidemiology, where model-based estimates regularly inform real-world decisions. However, sources of uncertainty are rarely formalised, and existing classifications are often inconsistent. This lack of structure hampers interpretation, model comparison, and targeted data collection. Connecting ideas from machine learning, information theory, experimental design, and health economics, we present a first-principles decision-theoretic framework that defines uncertainty as the expected loss incurred by making an estimate using incomplete information, arguing that this is a practically relevant definition for epidemiology. We show how reasoning about future data leads to a notion of expected uncertainty reduction, which induces formal definitions of reducible and irreducible uncertainty. We illustrate our approach with a simple worked example and a case study of SARS-CoV-2 wastewater surveillance in Aotearoa New Zealand, estimating the uncertainty reduction if wastewater surveillance were expanded to the full population. We then connect our framework to relevant literature from adjacent fields, showing how it unifies and extends many of these ideas. Our article serves as a gateway for applying a wide range of approaches to epidemiological models. Altogether, our framework provides a foundation for more reliable, consistent, and policy-relevant uncertainty quantification in infectious disease epidemiology.

Determining whether vaccine efficacy wanes is important for individual and public decision making. Yet, quantification of waning is a subtle task. The classical approaches cannot be interpreted as measures of declining efficacy unless we impose unreasonable assumptions. Recently, formal causal estimands designed to quantify vaccine waning have been proposed. These estimands can be bounded under weaker assumptions, but the bounds are often too wide to make claims about the presence of waning. We propose a different approach: a formal test to assess whether a treatment effect is constant over time at the individual level. This test provides a considerable power gain over existing approaches and is valid under interpretable assumptions in vaccine trials. We illustrate the increase in power through real and simulated examples, using three different approaches to compute the test statistics. Two of these approaches are based solely on summary data, accessible from existing clinical trials. Beyond our test, we also give new results that bound the waning effect. We use our methods to reanalyze data from a randomized controlled trial of the BNT162b2 COVID-19 vaccine. While prior analysis did not establish waning, our test rejects the null hypothesis of no waning.