Syndromic surveillance is useful for monitoring outbreaks of unknown etiologies or infectious disease pandemics when there is limited testing. However, traditional syndromic surveillance methods are not equipped to handle missingness in the data – especially when the missing completely at random (MCAR) assumption is violated. In this study, we analyze methods to improve syndromic surveillance in the context of missing data by taking advantage of the spatial and temporal structure of the data. We extend previous work using a seasonal temporal model to incorporate spatial and temporal effects in a frequentist framework commonly used in infectious disease modeling and also a Bayesian conditionally auto-regressive model. Inspired by COVID-19 symptom data collected via routine health systems in Liberia, we conduct simulations with different data generating processes, spatial correlation, temporal correlation, and missingness structures. For each simulation, we compare the ability of our models to detect COVID-19 outbreaks and to limit false positives. As the proportion of missingness in the baseline data increases, all of the methods’ point predictions stay relatively constant, while the uncertainty around these predictions increases, thus decreasing the ability to detect outbreaks. The Bayesian spatio-temporal model performs nearly as well as a baseline model in the absence of any spatial or temporal correlation, indicating its robustness. Researchers and public health experts will be able to use the insights from this paper to drive their modeling choices for syndromic surveillance.

Elevated blood pressure (BP) is a modifiable risk factor associated with cognitive impairment and cerebrovascular diseases. However, the causal effect of BP on white matter (WM) brain aging remains unclear. In this study, we conducted a two-sample Mendelian Randomization (MR) analysis to evaluate the causal effect of BP on WM brain aging in a cohort of individuals of European ancestry (N=219,968, 99,532 male, 120,436 female, mean age =56.55, including 1,6901 individuals with neuroimaging data) collected from UK Biobank. We first established a machine learning model to compute WM Brain Age Gap (BAG) from diffusion magnetic resonance imaging derived fractional anisotropy (FA) data and treated it as the outcome variable to measure the brain aging status. Females without hypertension were found younger in WM brain age than their male and hypertension counterparts of the same chronological age. The MR analyses showed an overall significant positive causal effect of diastolic blood pressure (DBP) on WM BAG, where every 10 mm Hg increase in DBP can lead to 0.371 years increase in brain age (CI: 0.034-0.709, p=0.0311). The stratified analysis by age and gender group found such significant causal effect of DBP on BAG to be most prominent among female women aged 50-59 (0.686 years/10mm Hg, CI: 0.054-1.318, p=0.0335) and aged 60-69 (0.962 years/10mm Hg, CI: 0.209-1.714, p=0.0122). Hypertension and genetic predisposition to higher BP can accelerate WM brain aging specifically targeting at late middle-aged women, providing insights on planning effective control of BP for women in this age group.

The widespread availability of COVID-19 testing has enabled the use of test-negative designs (TNDs) for evaluating vaccine efficacy in relation to the disease. These designs involve enlisting individuals who display COVID-19 symptoms and seek medical attention, subsequently testing them for the virus to distinguish cases from controls. TNDs help reduce selection bias stemming from healthcare-seeking behaviors. Nonetheless, the prevalent method for analyzing TND data is conditional logistic regression, which is essentially a special case of Cox regression and necessitates additional matching for potential confounders. In this study, we broaden the analysis method by utilizing the recurrent-event Cox regression model with time-dependent covariates to assess vaccine effectiveness. Furthermore, we present simulation studies for covariates of full and booster doses of COVID-19 vaccines, along with up to two infections for each individual. We also planned to conduct real data studies in Brazil to support our proposal.