Safety Impacts of Transit Signal Priority Using a Full Bayesian Approach

Transit signal priority (TSP) is a strategy that prioritizes the movement of transit vehicles through a signalized intersection to provide better transit travel time reliability and minimize transit delay. Although TSP is primarily intended to improve the operational performance of transit vehicles, it may also have substantial safety benefits. This study explored the potential safety benefits of the TSP strategy deployed at various locations in Florida. An observational before–after full Bayes (FB) approach with a comparison group was adopted to estimate the crash modification factors (CMFs) for total crashes, rear-end crashes, sideswipe crashes, and angle crashes. The analysis was based on 12 corridors equipped with the TSP system and their corresponding 29 comparison corridors without the TSP system. The deployment of TSP was found to reduce total crashes by 7.2% (CMF = 0.928), rear-end crashes by 5.2% (CMF = 0.948), and angle crashes by 21.9% (CMF = 0.781), and these results are statistically significant at a 95% Bayesian credible interval (BCI) except for the rear-end crashes. On the other hand, sideswipe crashes increased by 6% (CMF = 1.060) although the increase was not significant at a 95% BCI. Overall, the results indicated that TSP improves safety. The findings of this study may present key considerations for transportation agencies and practitioners when planning future TSP deployments.

Simultaneous confidence intervals for all pairwise differences between the coefficients of variation of rainfall series in Thailand

The delta-lognormal distribution is a combination of binomial and lognormal distributions, and so rainfall series that include zero and positive values conform to this distribution. The coefficient of variation is a good tool for measuring the dispersion of rainfall. Statistical estimation can be used not only to illustrate the dispersion of rainfall but also to describe the differences between rainfall dispersions from several areas simultaneously. Therefore, the purpose of this study is to construct simultaneous confidence intervals for all pairwise differences between the coefficients of variation of delta-lognormal distributions using three methods: fiducial generalized confidence interval, Bayesian, and the method of variance estimates recovery. Their performances were gauged by measuring their coverage probabilities together with their expected lengths via Monte Carlo simulation. The results indicate that the Bayesian credible interval using the Jeffreys’ rule prior outperformed the others in virtually all cases. Rainfall series from five regions in Thailand were used to demonstrate the efficacies of the proposed methods.

Comparing Medical Care Costs using Bayesian Credible Intervals for the Ratio of Means of Delta-Lognormal Distributions

When considering the medical care costs data with a high proportion of zero items of two inpatient groups, comparing them can be estimated using confidence intervals for the ratio of the means of two delta-lognormal distributions. The Bayesian credible interval-based uniform-beta prior (BCIh-UB) is proposed and compared with the generalized confidence interval (GCI), fiducial GCI (FGCI), the method of variance estimates recovery (MOVER), BCIh based on Jeffreys’ rule prior (BCIh-JR), and BCIh based on the normal-gamma prior (BCIh-NG). The coverage probability (CP), average length (AL), and lower and upper error rates were the performance measures applied for assessing the methods through a Monte Carlo simulation. A numerical evaluation showed that BCIh-UB had much better CP and AL than the others even with a large difference between the variances and with a high proportion of zero. Finally, to illustrate the efficacy of BCIh-UB, the methods were applied to two sets of medical care costs data.

The Support Interval

A frequentist confidence interval can be constructed by inverting a hypothesis test, such that the interval contains only parameter values that would not have been rejected by the test. We show how a similar definition can be employed to construct a Bayesian support interval. Consistent with Carnap’s theory of corroboration, the support interval contains only parameter values that receive at least some minimum amount of support from the data. The support interval is not subject to Lindley's paradox and provides an evidence-based perspective on inference that differs from the belief-based perspective that forms the basis of the standard Bayesian credible interval.

GFR Slope as a Surrogate End Point for Kidney Disease Progression in Clinical Trials: A Meta-Analysis of Treatment Effects of Randomized Controlled Trials

BackgroundSurrogate end points are needed to assess whether treatments are effective in the early stages of CKD. GFR decline leads to kidney failure, but regulators have not approved using differences in the change in GFR from the beginning to the end of a randomized, controlled trial as an end point in CKD because it is not clear whether small changes in the GFR slope will translate to clinical benefits.MethodsTo assess the use of GFR slope as a surrogate end point for CKD progression, we performed a meta-analysis of 47 RCTs that tested 12 interventions in 60,620 subjects. We estimated treatment effects on GFR slope (mean difference in GFR slope between the randomized groups), for the total slope starting at baseline, chronic slope starting at 3 months after randomization, and on the clinical end point (doubling of serum creatinine, GFR<15 ml/min per 1.73 m2, or ESKD) for each study. We used Bayesian mixed-effects analyses to describe the association of treatment effects on GFR slope with the clinical end point and to test how well the GFR slope predicts a treatment’s effect on the clinical end point.ResultsAcross all studies, the treatment effect on 3-year total GFR slope (median R2=0.97; 95% Bayesian credible interval [BCI], 0.78 to 1.00) and on the chronic slope (R2 0.96; 95% BCI, 0.63 to 1.00) accurately predicted treatment effects on the clinical end point. With a sufficient sample size, a treatment effect of 0.75 ml/min per 1.73 m2/yr or greater on total slope over 3 years or chronic slope predicts a clinical benefit on CKD progress with at least 96% probability.ConclusionsWith large enough sample sizes, GFR slope may be a viable surrogate for clinical end points in CKD RCTs.

Decline towards extinction of Mexico's vaquita porpoise ( Phocoena sinus )

The vaquita ( Phocoena sinus ) is a small porpoise endemic to Mexico. It is listed by IUCN as Critically Endangered because of unsustainable levels of bycatch in gillnets. The population has been monitored with passive acoustic detectors every summer from 2011 to 2018; here we report results for 2017 and 2018. We combine the acoustic trends with an independent estimate of population size from 2015, and visual observations of at least seven animals in 2017 and six in 2018. Despite adoption of an emergency gillnet ban in May 2015, the estimated rate of decline remains extremely high: 48% decline in 2017 (95% Bayesian credible interval (CRI) 78% decline to 9% increase) and 47% in 2018 (95% CRI 80% decline to 13% increase). Estimated total population decline since 2011 is 98.6%, with greater than 99% probability the decline is greater than 33% yr −1 . We estimate fewer than 19 vaquitas remained as of summer 2018 (posterior mean 9, median 8, 95% CRI 6–19). From March 2016 to March 2019, 10 dead vaquitas killed in gillnets were found. The ongoing presence of illegal gillnets despite the emergency ban continues to drive the vaquita towards extinction. Immediate management action is required if the species is to be saved.

Facility and Regional Factors Associated With the New Adoption of Electronic Medical Records in Japan: Nationwide Longitudinal Observational Study (Preprint)

BACKGROUND The rate of adoption of electronic medical record (EMR) systems has increased internationally, and new EMR adoption is currently a major topic in Japan. However, no study has performed a detailed analysis of longitudinal data to evaluate the changes in the EMR adoption status over time. OBJECTIVE This study aimed to evaluate the changes in the EMR adoption status over time in hospitals and clinics in Japan and to examine the facility and regional factors associated with these changes. METHODS Secondary longitudinal data were created by matching data in fiscal year (FY) 2011 and FY 2014 using reference numbers. EMR adoption status was defined as “EMR adoption,” “specified adoption schedule,” or “no adoption schedule.” Data were obtained for hospitals (n=4410) and clinics (n=67,329) that had no adoption schedule in FY 2011 and for hospitals (n=1068) and clinics (n=3132) with a specified adoption schedule in FY 2011. The EMR adoption statuses of medical institutions in FY 2014 were also examined. A multinomial logistic model was used to investigate the associations between EMR adoption status in FY 2014 and facility and regional factors in FY 2011. Considering the regional variations of these models, multilevel analyses with second levels were conducted. These models were constructed separately for hospitals and clinics, resulting in four multinomial logistic models. The odds ratio (OR) and 95% Bayesian credible interval (CI) were estimated for each variable. RESULTS A total of 6.9% of hospitals and 14.82% of clinics with no EMR adoption schedules in FY 2011 had adopted EMR by FY 2014, while 10.49% of hospitals and 33.65% of clinics with specified adoption schedules in FY 2011 had cancelled the scheduled adoption by FY 2014. For hospitals with no adoption schedules in FY 2011, EMR adoption/scheduled adoption was associated with practice size characteristics, such as number of outpatients (from quantile 4 to quantile 1: OR 1.67, 95% CI 1.005-2.84 and OR 2.40, 95% CI 1.80-3.21, respectively), and number of doctors (from quantile 4 to quantile 1: OR 4.20, 95% CI 2.39-7.31 and OR 2.02, 95% CI 1.52-2.64, respectively). For clinics with specified EMR adoption schedules in FY 2011, the factors negatively associated with EMR adoption/cancellation of scheduled EMR adoption were the presence of beds (quantile 4 to quantile 1: OR 0.57, 95% CI 0.45-0.72 and OR 0.74, 95% CI 0.58-0.96, respectively) and having a private establisher (quantile 4 to quantile 1: OR 0.27, 95% CI 0.13-0.55 and OR 0.43, 95% CI 0.19-0.91, respectively). No regional factors were significantly associated with the EMR adoption status of hospitals with no EMR adoption schedules; population density was positively associated with EMR adoption in clinics with no EMR adoption schedule (quantile 4 to quantile 1: OR 1.49, 95% CI 1.32-1.69). CONCLUSIONS Different approaches are needed to promote new adoption of EMR systems in hospitals as compared to clinics. It is important to induce decision making in small- and medium-sized hospitals, and regional postdecision technical support is important to avoid cancellation of scheduled EMR adoption in clinics.

Emergence of the East-Central-South-African genotype of Chikungunya virus in Brazil and the city of Rio de Janeiro may have occurred years before surveillance detection

Brazil, which is hyperendemic for dengue virus (DENV), has had recentZika(ZIKV) and (CHIKV)Chikungunyavirus outbreaks. Since March 2016, CHIKV is the arbovirus infection most frequently diagnosed in Rio de Janeiro. In the analysis of 1835 syndromic patients, screened by real time RT-PCR, 56.4% of the cases were attributed to CHIKV, 29.6% to ZIKV, and 14.1% to DENV-4. Sequence analyses of CHIKV from sixteen samples revealed that the East-Central-South-African(ECSA) genotype of CHIKV has been circulating in Brazil since 2013 [95% bayesian credible interval (BCI): 03/2012-10/2013], almost a year before it was detected by arbovirus surveillance program. Brazilian cases are related to Central African Republic sequences from 1980’s. To the best of our knowledge, given the available sequence published here and elsewhere, the ECSA genotype was likely introduced to Rio de Janeiro early on 2014 (02/2014; BCI: 07/2013-08/2014) through a single event, after primary circulation in the Bahia state at the Northestern Brazil in the previous year. The observation that the ECSA genotype of CHIKV was circulating undetected underscores the need for improvements in molecular methods for viral surveillance.

The Support Interval

A frequentist confidence interval can be constructed by inverting a hypothesis test, such that the interval contains only parameter values that would not have been rejected by the test. We show how a similar definition can be employed to construct a Bayesian support interval. Consistent with Carnap’s theory of corroboration, the support interval contains only parameter values that receive at least some minimum amount of support from the data. The support interval is not subject to Lindley’s paradox and provides an evidence-based perspective on inference that differ from the belief-based perspective that forms the basis of the standard Bayesian credible interval.