Journal of Biometrics & Biostatistics

Background: Designing a study for independent confirmation of a treatment effect is sometimes not practical due to required large sample size. Post hoc pooling of studies including those for learning purposes is subject to selection bias and therefore not scientifically solid. We propose a Bayesian approach which calibrates the role of prior information from historical studies for learning and confirming purposes. The method is illustrated in the analysis of mortality data for the pirfenidone NDA. Methods: The pirfenidone NDA includes three placebo-controlled studies to demonstrate efficacy for idiopathic pulmonary fibrosis (IPF), a rare and ultimately fatal lung disease with no approved treatment in the US at the time of NDA. The results of two earlier conducted studies PIPF-004 and PIPF-006 suggested that pirfenidone might reduce mortality risk. We used a Bayesian analysis to synthesize mortality results from the subsequent confirmative Study PIPF-016 and the combination of Studies PIPF-004 and PIPF-006. Results: Pirfenidone’s treatment effect on mortality rate reduction for Study PIPF-016 is statistically significant with discounts of historical evidence from PIPF-044 and PIPF-006 for both all-cause mortality and treatment-emergent IPF-related mortality. Conclusions: The Bayesian analysis provides a formal method to calibrate the role of information from historical evidence in the overall interpretation of results from both historical and concurrent clinical studies. The increased efficiency of using all available data is especially important in drug development for rare diseases with serious consequences, where limited patient source prohibits large trials, and unmet medical needs demand rapid access to treatment options.

Introduction: Coronavirus disease 2019 (COVID-19) is one of the serious infectious diseases that is caused by a specific virus called syndrome coronavirus 2 viruses (SARSCoV-2). The rapid spread of COVID19 raises serious concerns about the globally growing death rate. Currently, cases are doubled in one week around the world. Recorded data shows that COVID-19 does not infect all patients equally. This opportunistic virus can affect people of any age and gender. Information about the reason for high mortality in the age group 60 and older is limited. The gender differences among all deceased are poorly known. To understand more about COVID-19, this study aims to examine the different age groups among the death and focuses on comparing genders between males and females. Method: Statistical analysis including Pearson’s Chi-squared (χ2 ) and binary logistic regression was conducted based on existing data to examine factors relating to death, such as age and gender. Adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for death. Results: The results show that males were 2.51 more likely to die of coronavirus COVID-19 than females. Moreover, the study found a significant increase in death for patients age 60 and older compared to patients age less than 40. Thus, males of 80+ age were found to be highly associated with death. Conclusions: Older people and male are more susceptible to death from COVID-19,we should pay more attention to the elderly people and male with COVID-19. This imposes providing careful health care for this population.

Background: Promotion of smoking cessation has been proposed as one of the primary areas of focus for tobacco control in developing countries as prevalence is high over there. This paper aimed to analyze statistically quitting method followed by the smokers who wanted to quit tobacco use in the past 12 months of the survey. Methods: The paper was based on secondary data of size 9629 collected from people aged 15 years and above by the Global Adult Tobacco Survey (GATS), 2010. Descriptive analysis and binary logistic regression had been performed using STATA-13 to analyze the data. Outcome variable was whether quitting method(s) was (were) followed by the tobacco user (1. Tobacco smoker, and 2. Smokeless tobacco user) who wanted to quit tobacco use in the past 12 months of the survey and independent variables were age, gender, residential status, education, occupation and wealth index. Results: It had been found that 47.38% of smoker respondents tried to quit tobacco smoking and among them 27.13% used any method to quit. It had been also found that among the smokeless tobacco users, 31.89% tried to quit and among them 24.83% used any method to quit. Among the quitting methods, counselling was the most used method. From the logistic regression to methods used to quitting tobacco use, it had been found that age, education and wealth index were significantly associated with the use of methods to quit tobacco smoking; whereas, gender, age and wealth index were statistically significant to the use of methods to quit smokeless tobacco. Conclusions: This study suggests that more active quitting methods should be invented targeting male, younger, lower educated and poorer tobacco users to make the cessation successful in Bangladesh.

Context: There is a lack of research that prospectively estimates the age of initiation of electronic cigarette use in U.S. youth. Younger ages of initiation of tobacco product use are associated with greater exposure to nicotine, and recently e-cigarette use has been associated with subsequent cigarette initiation. This study sought to estimate the distribution of the age of first reporting of e-cigarette use outcomes among youth never e-cigarette users overall, by sex and by race/ethnicity, prospectively. Methods: Secondary analysis of the Population Assessment of Tobacco and Health (PATH) youth dataset (ages 12-17) across waves 1 (2013-2014), 2 (2014-2015), 3 (2015-2016), and 4 (2016-2017) were conducted. Four outcomes are presented, age of first report of: (i) susceptibility to use, (ii) ever, (iii) past 30-day use, and (iv) “fairly regular” e-cigarette use. Each outcome was prospectively estimated using participant age when they entered the study and the number of weeks between the last report of never use and the first report of each outcome across waves. Weighted survival analyses for interval censoring accounting for the complex survey design were implemented. Results: Among youth non-susceptible to e-cigarettes, 50.2% became susceptible to e-cigarette use by age 18. There were no statistically significant differences in the age of first report of susceptibility to e-cigarette use by sex or by race/ethnicity in this nationally representative sample of U.S. youth. Among never users, 41.7%, 23.5% and 10.3% initiated ever, past 30-day and “fairly regular” e-cigarette use by the age of 18, respectively. Less than 10% initiated ever e-cigarette use between the ages of 18 and 21. Boys had a higher risk of first reporting ever, past 30-day and “fairly regular” e-cigarette use at earlier ages than girls. Non-Hispanic Blacks and Other racial/ ethnic groups were less likely than Non-Hispanic Whites to initiate ever e-cigarette use at earlier ages, and there was no difference between Non-Hispanic Whites and Hispanics. Hispanic, Non-Hispanic Black and Other racial/ethnic youth were less likely to first report past 30-day use and “fairly regular” e-cigarette use at earlier ages than Non-Hispanic White youth. Conclusion: This paper provides information on specific ages of the first report of e-cigarette use behaviors by sex and by race/ethnicity that can be used to tailor culturally e-cigarette interventions on specific windows of opportunity before youth begin using e-cigarettes or escalating their use.

Bivariate correlation analysis is one of the most commonly used statistical methods. Unfortunately, it is generally the case that little or no attention is given to sample size determination when planning a study in which correlation analysis will be used. For example, our review of clinical research journals indicated that none of the 111 articles published in 2014 that presented correlation results provided a justification for the sample size used in the correlation analysis. There are a number of easily accessible tools that can be used to determine the required sample size for inference based on a Pearson correlation coefficient; however, we were unable to locate any widely available tools that can be used for sample size calculations for a Spearman correlation coefficient or a Kendall coefficient of concordance. In this article, we provide formulas and charts that can be used to determine the required sample size for inference based on either of these coefficients. Additional sample size charts are provided in the Supplementary Materials.