Background: Many studies have modeled and predicted the spread of COVID-19 (coronavirus disease 2019) in the U.S. using data that begins with the first reported cases. However, the shortage of testing services to detect infected persons makes this approach subject to error due to its underdetection of early cases in the U.S. Our new approach overcomes this limitation and provides data supporting the public policy decisions intended to combat the spread of COVID-19 epidemic. Methods: We used Centers for Disease Control and Prevention data documenting the daily new and cumulative cases of confirmed COVID-19 in the U.S. from January 22 to April 6, 2020, and reconstructed the epidemic using a 5-parameter logistic growth model. We fitted our model to data from a 2-week window (i.e., from March 21 to April 4, approximately one incubation period) during which large-scale testing was being conducted. With parameters obtained from this modeling, we reconstructed and predicted the growth of the epidemic and evaluated the extent and potential effects of underdetection. Results: The data fit the model satisfactorily. The estimated daily growth rate was 16.8% overall with 95% CI: [15.95, 17.76%], suggesting a doubling period of 4 days. Based on the modeling result, the tipping point at which new cases will begin to decline will be on April 7th, 2020, with a peak of 32,860 new cases on that day. By the end of the epidemic, at least 792,548 (95% CI: [789,162, 795,934]) will be infected in the U.S. Based on our model, a total of 12,029 cases were not detected between January 22 (when the first case was detected in the U.S.) and April 4. Conclusions: Our findings demonstrate the utility of a 5-parameter logistic growth model with reliable data that comes from a specified period during which governmental interventions were appropriately implemented. Beyond informing public health decision-making, our model adds a tool for more faithfully capturing the spread of the COVID-19 epidemic.

A well-designed clinical trial requires an appropriate sample size with adequate statistical power to address trial objectives. The statistical power is traditionally defined as the probability of rejecting the null hypothesis with a pre-specified true clinical treatment effect. This power is a conditional probability conditioned on the true but actually unknown effect. In practice, however, this true effect is never a fixed value. Thus, we discuss a newly proposed alternative to this conventional statistical power: statistical assurance, defined as the unconditional probability of rejecting the null hypothesis. This kind of assurance can then be obtained as an expected power where the expectation is based on the prior probability distribution of the unknown treatment effect, which leads to the Bayesian paradigm. In this article, we outline the transition from conventional statistical power to the newly developed assurance and discuss the computations of assurance using Monte Carlo simulation-based approach.

The cusp catastrophe model is an innovative approach for investigating a phenomenon that consists of both continuous and discrete changes in one modeling framework. However, its application to empirical health and behavior data has been hindered by the complexity in data-model fit. In this study, we reported our work in the development of a new modeling method—cusp catastrophe regression (RegCusp in short) by casting the cusp catastrophe into a statistical regression. With the RegCusp approach, unbiased model parameters can be estimated with the maximum likelihood estimation method. To validate the RegCusp method, a series of simulations were conducted to demonstrate the unbiasedness of parameter estimation. Since the estimated residual variance with the Fisher information matrix method was over-dispersed, a bootstrap re-sampling procedure was developed and used as a remedy. We also demonstrate the practical applicability of the RegCusp with empirical data from an NIH-funded project to evaluate an HIV prevention intervention program to educate adolescents in the Bahamas for condom use. Study findings indicated that the model parameters estimated with RegCusp were practically more meaningful than those estimated with comparable methods, especially the estimated cusp point.

A regional stock–recruitment meta-model is developed using a hierarchical Bayesian framework to combine information from multiple fish populations. The use of the meta-model is illustrated through analysis of the regional stock–recruitment parameters of the coho salmon (Oncorhynchus kisutch) within two large fisheries management units in southern and northern British Columbia. We construct our regional prior distribution from an analysis of all stock-recruitment data rather than by the more usual approach of assuming a prior distribution. That preliminary analysis indicated that the regional prior distribution for the two parameters of the Ricker model was bivariate normal–lognormal (NLN) with a high degree of correlation between the two Ricker parameters. Because this distribution had not been fully developed, we formulated the density function for the NLN distribution and proved some of its important properties. An empirical Bayesian approach was then used to estimate the regional distributions of the Ricker parameters and derived management parameters. Characterization of the distributional properties of productivity within management regions is a necessary step for resource managers seeking to prosecute mixed-stock fisheries while conserving population diversity.

Count data with structural zeros are common in public health applications. There are considerable researches focusing on zero-inflated models such as zero-inflated Poisson (ZIP) and zero-inflated Negative Binomial (ZINB) models for such zero-inflated count data when used as response variable. However, when such variables are used as predictors, the difference between structural and random zeros is often ignored and may result in biased estimates. One remedy is to include an indicator of the structural zero in the model as a predictor if observed. However, structural zeros are often not observed in practice, in which case no statistical method is available to address the bias issue. This paper is aimed to fill this methodological gap by developing parametric methods to model zero-inflated count data when used as predictors based on the maximum likelihood approach. The response variable can be any type of data including continuous, binary, count or even zero-inflated count responses. Simulation studies are performed to assess the numerical performance of this new approach when sample size is small to moderate. A real data example is also used to demonstrate the application of this method.

Background: Fatigue is a common symptom in adults with inflammatory bowel disease (IBD) and is influenced by many physiological, psychological, and situational factors. However, the influencing factors of fatigue associated with IBD have not been evaluated. Objective: This study aims to examine factors associated with fatigue during IBD and develop a parsimonious model that describes the influencing factors of fatigue. Methods: The study was a secondary analysis of cross-sectional data obtained from IBD Partners, an online cohort of adults with the disease, including 12,053 eligible participants. Data were collected using the Patient-Reported Outcomes Measurement Information System short-form scales measuring fatigue, sleep disturbances, pain interference, anxiety, depression, and satisfaction with social roles. Physical activity was measured using a single question. Demographic and clinical variables were collected. Path analysis was computed to identify the direct and indirect effects of situational, physiological, and psychological factors on IBD–fatigue based on the middle range theory of unpleasant symptoms’ conceptual framework. Results: Most of the participants were White females. The data best fit a model with situational factors (physical activity and satisfaction with social roles as the mediators). The direct effect of IBD activity, age, sleep disturbances, pain interference, anxiety, and depression on IBD–fatigue was significant. Significant indirect effects were noted on IBD–fatigue from sleep disturbances, pain interference, and depression via physical activity and satisfaction with social roles. Discussion: The study identified two important intervening variables from the tested model. In addition, other symptoms such as sleep, pain, anxiety, and depression are essential and also influence IBD–fatigue.