In one way, the pandemic magnifies socioeconomic inequalities, such as poverty and unemployment in workers' countries of origin, which heighten the risk of victimisation of workers through labour trafficking [1] . Conversely, labour trafficking serves as a social origin for COVID-19 through a unique set of social disadvantages-i.e., reluctance to seek medical care due to fear of legal prosecution-that make infected workers potential disease vectors which threaten the health of the host country's population. The resurgence of COVID-19 in Thailand substantiates this notion that labour trafficking is a critical but neglected public health issue [2] .Once lauded for achieving a 103-day period of zero domestic transmission after its first surge of COVID-19, Thailand unveiled a new cluster of 2,629 COVID-19 cases (65.7%) among an estimated 4,0 0 0 international migrant workers at the country's largest shrimp market between December 20-27, 2020 [3] . Within days, the cluster resulted in the spread of disease to 44 out of 77 provinces. This incident raises the question: What is the mechanism underlying this impactful pathway of disease transmission that harms the public at large? The answer lies at Thailand's border.Economic hardships abroad and border restrictions in Thailand during the pandemic have contributed to a spike in labour trafficking of international migrant workers. Once victims of trafficking, workers are limited to low-paying labour-intensive jobs which make poor quality housing unavoidable. Furthermore, workers' status due to illegal entry into the country inhibits their ability to access healthcare [4] . Regardless of these specific social contexts, pre-existing disease control measures tend to emphasize biomedi- * Corresponding author.

Background Dental fluorosis can be a disease of social inequity in access to safe drinking water. This dental public health issue becomes prominent in socially disadvantaged agrarian communities in fluoride endemic areas where the standard irrigation system is unavailable and groundwater containing natural fluoride is the major drinking water source. This study aimed to determine the prevalence and severity of dental fluorosis in children and to evaluate its association with fluoride in groundwater in the aforementioned setting in Thailand. Methods A cross-sectional survey of 289 children in Nakhon Pathom Province was conducted in 2015. Children with very mild to severe fluorosis were regarded as ‘cases’ while their counterparts were ‘controls’ for a subsequent case–control study. Records of fluoride concentrations in groundwater used for household supply corresponding to resident and number of years by age of each child during 2008–2015 were retrieved. Other exposure variables were measured using a questionnaire. Prevalence ratio (PR), a measure indicating the relative effect of different levels of fluoride on dental fluorosis, was obtained from Poisson regression with robust standard error. Result There were 157 children with very mild to moderate dental fluorosis (54.3% prevalence). The univariable analysis revealed that the prevalence of dental fluorosis among children with fluoride concentrations in water sources of 0.7–1.49 (index category 1) and ≥ 1.5 ppm (index category 2) was 1.62 (95% CI; 0.78, 3.34) and 2.75 (95% CI; 1.42, 5.31) times the prevalence among those with fluoride < 0.7 ppm (referent category). After adjusting for all covariates, the adjusted prevalence ratios in both index categories were 1.64 (95% CI; 0.24, 11.24) and 2.85 (95% CI; 0.44, 18.52) which were close to their corresponding crude estimates. Since the magnitude of confounding, measured by (PRcrude–PRadjusted)/PRadjusted, were less than 10% for both index categories; this indicated the limited confounding effect of all covariates. Conclusions In fluoride endemic areas, groundwater containing natural fluoride utilized for household consumption resulted in high dental fluorosis prevalence, particularly in the groundwater with fluoride concentrations of ≥ 1.5 ppm.

This study empirically illustrates the mechanism by which epidemiological effect measures and statistical evidence can be misleading in the presence of Simpson�s paradox and identify possible alternative methods of analysis to manage the paradox. Three scenarios of observational study designs, including cross-sectional, cohort, and case-control approaches, are simulated. In each scenario, data are generated, and various methods of epidemiological and statistical analyses are undertaken to obtain empirical results that illustrate Simpson�s paradox and mislead conclusions. Rational methods of analysis are also performed to illustrate how to avoid pitfalls and obtain valid results. In the presence of Simpson�s paradox, results from analyses in overall data contradict the findings from all subgroups of the same data. This paradox occurs when distributions of confounding characteristics are unequal in the groups being compared. Data analysis methods which do not take confounding factor into account, including epidemiological 2�2 table analysis, independent samples t-test, Wilcoxon rank-sum test, chi-square test, and univariable regression analysis, cannot manage the problem of Simpson�s paradox and mislead research conclusions. Mantel-Haenszel procedure and multivariable regression methods are examples of rational analysis methods leading to valid results. Therefore, Simpson�s paradox arises as a consequence of extreme unequal distributions of a specific inherent characteristic in groups being compared. Analytical methods which take control of confounding effect must be applied to manage the paradox and obtain valid research evidence regarding the causal association.

Background: In epidemiologic investigations of disease outbreaks, multivariable regression techniques with adjustment for confounding can be applied to assess the association between exposure and outcome. Traditionally, logistic regression has been used in analyses of case-control studies to determine the odds ratio (OR) as the effect measure. For rare outcomes (incidence of 5% to 10%), an adjusted OR can be used to approximate the risk ratio (RR). However, concern has been raised about using logistic regression to estimate RR because how closely the calculated OR approximates the RR depends largely on the outcome rate. The literature shows that when the incidence of outcomes exceeds 10%, ORs greatly overestimate RRs. Consequently, in addition to logistic regression, other regression methods to accurately estimate adjusted RRs have been explored. One method of interest is Poisson regression with robust standard errors. This generalized linear model estimates RR directly vs logistic regression that determines OR. The purpose of this study was to empirically compare risk estimates obtained from logistic regression and Poisson regression with robust standard errors in terms of effect size and determination of the most likely source in the analysis of a series of simulated single-source disease outbreak scenarios. Methods: We created a prototype dataset to simulate a foodborne outbreak following a public event with 14 food exposures and a 52.0% overall attack rate. Regression methods, including binary logistic regression and Poisson regression with robust standard errors, were applied to analyze the dataset. To further examine how these two models led to different conclusions of the potential outbreak source, a series of 5 additional scenarios with decreasing attack rates were simulated and analyzed using both regression models. Results: For each of the explanatory variables-sex, age, and food types-in both univariable and multivariable models, the ORs obtained from logistic regression were estimated further from 1.0 than their corresponding RRs estimated by Poisson regression with robust standard errors. In the simulated scenarios, the Poisson regression models demonstrated greater consistency in the identification of one food type as the most likely outbreak source. Conclusion: Poisson regression with robust standard errors proved to be a decisive and consistent method to estimate risk associated with a single source in an outbreak when the cohort data collection design was used.

Background Changes in the epidemiology of lip, oral cavity, and pharyngeal (LOCP) cancers have been reported in the United States. This study aimed to examine recent trends in LOCP cancer mortality in the United States from 1999 to 2019. Methods National mortality data were extracted from CDC WONDER, 1999–2019. International Classification of Diseases Codes, 10th Revision—C00‐C14, were used to identify decedents of malignant neoplasms of the lip, oral cavity, and pharynx. LOCP cancer mortality trends were assessed by fitting a Joinpoint regression model overall, and by race/ethnicity, sex, age, and US Census Region. Annual Percentage Changes (APC) were derived to estimate variations in mortality trends over time. Results The age‐adjusted mortality rate (AAMR) for LOCP cancers was 2.5 per 100 000 (95% CI: 2.5–2.5), equivalent to 180 532 deaths during 1999–2019. Overall mortality trends have stabilized since 2009 (APC = 0.3; 95% CI: −0.1, 0.7), but an examination by subtype revealed rising mortality trends from cancers of the lip and oral cavity (APC = 1.2; 95% CI: 0.7, 1.6) and pharynx (APC = 3.2; 95% CI: 1.7, 4.8), and declining trends in malignancies of other and ill‐defined areas of the lip, oral cavity, and pharynx (APC = −2.7; 95% CI: −3.4, −2.0). Trend variations were also noted by sex, age, US Census Region, and race/ethnicity. Conclusions There are differential trends in mortality from LOCP cancers in the United States. Investigating the biological, individual, and contextual factors related to LOCP cancers would guide effective public health intervention efforts.