A simple method for correcting for the Will Rogers phenomenon with biometrical applications.

In its basic form, the Will Rogers phenomenon takes place when an increase in the average value of each of two sets is achieved by moving an element from one set to another. This leads to the conclusion that there has been an improvement, when in fact essentially nothing has changed. Extended versions of this phenomenon can occur in epidemiological studies, rendering their results unreliable. After describing epidemiological and clinical studies that have been affected by the Will Rogers phenomenon, this paper presents a simple method to correct for it. The method involves introducing a transition matrix between the two sets and taking probability weighted expectations. Two real-world biometrical examples, based on migration economics and breast cancer epidemiology, are given and improvements against a naïve analysis are demonstrated. In the cancer epidemiology example, we take account of estimation uncertainty. We also discuss briefly some limitations associated with our method.

Analysis of paediatric visual acuity using Bayesian copula models with sinh-arcsinh marginal densities.

We analyse paediatric ophthalmic data from a large sample of children aged between 3 and 8 years. We use a Bayesian additive conditional bivariate copula regression model with sinh-arcsinh marginal densities with location, scale, and shape parameters that depend smoothly on a covariate. We perform Bayesian inference about the unknown quantities of our model using a specially tailored Markov chain Monte Carlo algorithm. We gain new insights about the processes, which determine transformations in visual acuity with respect to age, including the nature of joint changes in both eyes as modelled with the age-related copula dependence parameter. We analyse posterior predictive distributions to identify children with unusual sight characteristics, distinguishing those who are bivariate, but not univariate outliers. In this way, we provide an innovative tool that enables clinicians to identify children with unusual sight who may otherwise be missed. We compare our simultaneous Bayesian method with a two-step frequentist generalised additive modelling approach.

Covid lockdowns in the UK: Estimating their effects on transmission.

UK citizens have lived through three lockdowns over the course of the Covid-19 pandemic. What effect have these restrictions had on the transmission of the virus? Emily Prestige, Julian Stander and Yinghui Wei investigate.

Estimating the number of livebirths to Hepatitis C seropositive women in England in 2013 and 2018 using Bayesian modelling.

BACKGROUND: The UK National Screening Committee currently recommends against antenatal screening for Hepatitis C virus (HCV) infection in England due to lack of HCV prevalence data and treatment licensed for use in pregnancy. We aimed to produce regional and national estimates of the number and proportion of livebirths to HCV seropositive women in England in 2013 and 2018. METHODS: A logistic regression model fitted in the Bayesian framework estimated probabilities of HCV seropositivity among 24,599 mothers delivering in the North Thames area of England in 2012 adjusted by maternal age and region of birth. These probabilities were applied to the underlying population structures of women delivering livebirths in England in 2013 and 2018 to estimate the number of livebirths to HCV seropositive women in these years nationally and by region. The Bayesian approach allowed the uncertainty associated with all estimates to be properly quantified. RESULTS: Nationally, the estimated number of livebirths to women seropositive for HCV for England was 464 (95% credible interval [CI] 300-692) in 2013 and 481 (95%CI 310-716) in 2018, or 70.0 (95%CI 45.0-104.1) per 100,000 and 76.9 (95%CI 49.5-114.4) per 100,000 in these years respectively. Regions with the highest estimated number of livebirths to HCV seropositive women in 2013 and 2018 included London with 118.5 and 124.4 and the South East with 67.0 and 74.0 per 100,000 livebirths. CONCLUSION: Few previous studies have investigated HCV among pregnant women in England. These findings complement and supplement existing research by providing national and regional estimates for the number of livebirths to HCV seropositive women in England. Bayesian modelling allows future national and regional estimates to be produced and the associated uncertainty to be properly quantified.

Contrasting patterns in the small-scale heterogeneity of human helminth infections in urban and rural environments in Brazil.

Marked heterogeneity exists in the patterns of parasitic infection between individuals, households and communities. Analysis of parasite distributions within populations is complicated by the fact that parasite distributions are highly aggregated and few studies have explicitly incorporated this distribution when investigating small-scale spatial heterogeneities. This study aimed to quantify the small-scale (within- and between-household) heterogeneity of helminth infection in an area of Minas Gerais State, Brazil, with rural and urban sectors. Parasitological data from a cross-sectional survey of 1,249 individuals aged 0-86 years from 242 households were analysed. Within-household clustering of infection was assessed using random effect logistic regression models and between-household spatial heterogeneity was assessed using a Bayesian negative binomial spatial model. The overall prevalence of hookworm (Necator americanus) was 66.9%, the prevalence of Schistosoma mansoni was 44.9% and the prevalence of Ascaris lumbricoides was 48.8%. Statistical analysis indicated significant (within) household and (between household) spatial clustering of hookworm in both rural and urban areas and of S. mansoni in rural areas. There was no evidence of either household or spatial clustering of S. mansoni in urban areas. The spatial correlation of S. mansoni was estimated to reduce by half over a distance of 700 m in the rural area. Rural hookworm had a much smaller half-distance (28 m) and urban hookworm showed an even smaller half-distance (12 m). We suggest that such species-specific differences in patterns of infection by environment are primarily due to variation in exposure and parasite life cycle, although host genetic factors cannot be ruled out.

Spatial variation of Anopheles-transmitted Wuchereria bancrofti and Plasmodium falciparum infection densities in Papua New Guinea.

The spatial variation of Wuchereria bancrofti and Plasmodium falciparum infection densities was measured in a rural area of Papua New Guinea where they share anopheline vectors. The spatial correlation of W. bancrofti was found to reduce by half over an estimated distance of 1.7 km, much smaller than the 50 km grid used by the World Health Organization rapid mapping method. For P. falciparum, negligible spatial correlation was found. After mass treatment with anti-filarial drugs, there was negligible correlation between the changes in the densities of the two parasites.