Modelling multisera data: The estimation of new joint and conditional epidemiological parameters

Estimating HIV incidence over a decade in Zimbabwe: A comparison of the catalytic and Farrington models

Over the years, numerous modelling studies have been proposed to estimate HIV incidence. As a result, this study aimed to evaluate two alternative methods for predicting HIV incidence in Zimbabwe between 2005 and 2015. We estimated HIV incidence from seroprevalence data using the catalytic and Farrington-2-parameter models. Data were obtained from 2005-06, 2010-11, and 2015 Zimbabwe Demographic Health Survey (ZDHS). These models were validated at the micro and macro-level using community-based cohort incidence and empirical estimates from UNAIDS EPP/SPECTRUM, respectively. The HIV incidence for the catalytic model was 0.32% (CI: 0.28%, 0.36%), 0.36% (CI: 0.33%, 0.39%), and 0.28% (CI: 0.26%, 0.30%), for the years 2005-06, 2010-11, and 2015, respectively. The HIV incidence for the Farrington model was 0.21% (CI: 0.16%, 0.26%), 0.22% (CI: 0.20%, 0.25%), and 0.19% (CI: 0.16%, 0.22%), for the years 2005-06, 2010-11, and 2015, respectively. According to these findings, the catalytic model estimated a higher HIV incidence rate than the Farrington model. Compared to cohort estimates, the estimates were within the observed 95% confidence interval, with 88% and 75% agreement for the catalytic and Farrington models, respectively. The limits of agreement observed in the Bland-Altman plot were narrow for all plots, indicating that our model estimates were comparable to cohort estimates. Compared to UNAIDS estimates, the catalytic model predicted a progressive increase in HIV incidence for males throughout all survey years. Without a doubt, HIV incidence declined with each subsequent survey year for all models. To improve programmatic and policy decisions in the national HIV response, we recommend the triangulation of multiple methods for incidence estimation and interpretation of results. Multiple estimating approaches should be considered to reduce uncertainty in the estimations from various models.

The impact of behavioral interventions on co-infection dynamics: An exploration of the effects of home isolation

Behavioral epidemiology, the field aiming to determine the impact of individual behavior on the spread of an epidemic, has gained increased recognition during the last few decades. Behavioral changes due to the development of symptoms have been studied in mono-infections. However, in reality, multiple infections are circulating within the same time period and behavioral changes resulting from contraction of one of the diseases affect the dynamics of the other. The present study aims at assessing the effect of home isolation on the joint dynamics of two infectious diseases, including co-infection, assuming that the two diseases do not confer cross-immunity. We use an age- and time- structured co-infection model based on partial differential equations. Social contact matrices, describing different mixing patterns of symptomatic and asymptomatic individuals are incorporated into the calculation of the age- and time-specific marginal forces of infection. Two scenarios are simulated, assuming that one of the diseases has more severe symptoms than the other. In the first scenario, people stay only at home when having symptoms of the most severe disease. In the second scenario, twice as many people stay at home when having symptoms of the most severe disease than when having symptoms of the other disease. The results show that the impact of home isolation on the joint dynamics of two infectious diseases depends on the epidemiological parameters and properties of the diseases (e.g., basic reproduction number, symptom severity). In case both diseases have a low to moderate basic reproduction number, and there is no home isolation for the less severe disease, the final size of the less severe disease increases with the proportion of symptomatic cases of the most severe disease staying at home, after an initial decrease. This counterintuitive result could be explained by a shift in the peak time of infection of the disease with the most severe symptoms, resulting in a smaller number of people with less contacts at the peak time of the other infection. When twice as many people stay at home when having symptoms of the most severe disease than when having symptoms of the other disease, increasing the proportion staying at home always reduces the final size of both diseases, and the number of co-infections. In conclusion, when providing advise if people should stay at home in the context of two or more co-circulating diseases, one has to take into account epidemiological parameters and symptom severity.

Measures for concordance and discordance with applications in disease control and prevention

Bivariate binary response data appear in many applications. Interest goes most often to a parameterization of the joint probabilities in terms of the marginal success probabilities in combination with a measure for association, most often being the odds ratio. Using, for example, the bivariate Dale model, these parameters can be modelled as function of covariates. But the odds ratio and other measures for association are not always measuring the (joint) characteristic of interest. Agreement, concordance, and synchrony are in general facets of the joint distribution distinct from association, and the odds ratio as in the bivariate Dale model can be replaced by such an alternative measure. Here, we focus on the so-called conditional synchrony measure. But, as indicated by several authors, such a switch of parameter might lead to a parameterization that does not always lead to a permissible joint bivariate distribution. In this contribution, we propose a new parameterization in which the marginal success probabilities are replaced by other conditional probabilities as well. The new parameters, one homogeneity parameter and two synchrony/discordance parameters, guarantee that the joint distribution is always permissible. Moreover, having a very natural interpretation, they are of interest on their own. The applicability and interpretation of the new parameterization is shown for three interesting settings: quantifying HIV serodiscordance among couples in Mozambique, concordance in the infection status of two related viruses, and the diagnostic performance of an index test in the field of major depression disorders.

Modelling time varying heterogeneity in recurrent infection processes: an application to serological data

Frailty models are often used in survival analysis to model multivariate time‐to‐event data. In infectious disease epidemiology, frailty models have been proposed to model heterogeneity in the acquisition of infection and to accommodate association in the occurrence of multiple types of infection. Although traditional frailty models rely on the assumption of lifelong immunity after recovery, refinements have been made to account for reinfections with the same pathogen. Recently, Abrams and Hens quantified the effect of misspecifying the underlying infection process on the basic and effective reproduction number in the context of bivariate current status data on parvovirus B19 and varicella zoster virus. Furthermore, Farrington, Unkel and their co‐workers introduced and applied time varying shared frailty models to paired bivariate serological data. In this paper, we consider an extension of the proposed frailty methodology by Abrams and Hens to account for age‐dependence in individual heterogeneity through the use of age‐dependent shared and correlated gamma frailty models. The methodology is illustrated by using two data applications.

Prenatal diagnosis and prevention of toxoplasmosis in pregnant women in Northern Vietnam: study protocol

Background

In Vietnam, no systematic prenatal toxoplasmosis screening is in place, and only few studies have assessed the prevalence and importance of this zoonotic parasite infection. In addition, no studies have been conducted to assess the risk factors associated with toxoplasmosis. This study protocol was developed to determine the seroprevalence of toxoplasmosis in pregnant women in Hanoi and Thai Binh, Northern Vietnam, and to evaluate the association with risk factors and congenital toxoplasmosis. The protocol was developed in a way that it could potentially evolve into a countrywide prenatal diagnosis and prevention program, with the main focus on primary prevention.

Methods

The collaborating gynaecologists will invite eligible pregnant women attending antenatal care for the first time to participate in the study. At first consult, information about toxoplasmosis and its prevention will be provided. All participants will be asked to fill in a questionnaire, which is designed to analyse socio-demographic and biologically plausible risk factors associated with toxoplasmosis, and blood samples will be collected to determine the seroprevalence of toxoplasmosis in pregnant women. In case there is suspicion of a primary infection during pregnancy, the concerned women will be followed-up by the gynaecologists according to a predefined protocol. Every participant will be informed on her serological status, risk factors and prevention measures and is offered appropriate medical information and medical follow-up if required.

Discussion

The hypothesis is that congenital toxoplasmosis is an important but currently under-diagnosed public health problem in Vietnam. This study can strengthen sustainable control of toxoplasmosis in Vietnam, provide a protocol for prenatal diagnosis, boost overall awareness, improve the knowledge about toxoplasmosis prevention and can be essential for evidence-based health policy.

Electronic supplementary material

The online version of this article (doi:10.1186/s12879-017-2446-1) contains supplementary material, which is available to authorized users.

A Bivariate Mixture Model for Natural Antibody Levels to Human Papillomavirus Types 16 and 18: Baseline Estimates for Monitoring the Herd Effects of Immunization

Post-vaccine monitoring programs for human papillomavirus (HPV) have been introduced in many countries, but HPV serology is still an underutilized tool, partly owing to the weak antibody response to HPV infection. Changes in antibody levels among non-vaccinated individuals could be employed to monitor herd effects of immunization against HPV vaccine types 16 and 18, but inference requires an appropriate statistical model. The authors developed a four-component bivariate mixture model for jointly estimating vaccine-type seroprevalence from correlated antibody responses against HPV16 and -18 infections. This model takes account of the correlation between HPV16 and -18 antibody concentrations within subjects, caused e.g. by heterogeneity in exposure level and immune response. The model was fitted to HPV16 and -18 antibody concentrations as measured by a multiplex immunoassay in a large serological survey (3,875 females) carried out in the Netherlands in 2006/2007, before the introduction of mass immunization. Parameters were estimated by Bayesian analysis. We used the deviance information criterion for model selection; performance of the preferred model was assessed through simulation. Our analysis uncovered elevated antibody concentrations in doubly as compared to singly seropositive individuals, and a strong clustering of HPV16 and -18 seropositivity, particularly around the age of sexual debut. The bivariate model resulted in a more reliable classification of singly and doubly seropositive individuals than achieved by a combination of two univariate models, and suggested a higher pre-vaccine HPV16 seroprevalence than previously estimated. The bivariate mixture model provides valuable baseline estimates of vaccine-type seroprevalence and may prove useful in seroepidemiologic assessment of the herd effects of HPV vaccination.

Childhood varicella-zoster virus vaccination in Belgium: cost-effective only in the long run or without exogenous boosting?

AIM

To assess the effectiveness and cost-effectiveness of a universal childhood varicella-zoster vaccination programme in Belgium (1) using the most recent Belgian data on varicella-zoster burden, (2) exploring different options for the timing of the second dose, (3) obtaining results with and without exogenous natural boosting, and (4) investigating the possible additional benefit of zoster booster vaccination for adults at age 50 or 60 y.

METHODS

An extensively studied and improved dynamic model is used to estimate primary and breakthrough chickenpox and zoster cases over time. For a range of vaccination options, we compared the direct costs (health care payer perspective) and health outcomes (including Quality-Adjusted Life-Years (QALYs) lost) associated with chickenpox and herpes zoster. Estimates of social contact patterns, health care use, costs and QALY losses are almost exclusively based on Belgian databases and surveys.

RESULTS AND CONCLUSIONS

If exogenous natural boosting exists, a net loss in QALYs is expected for several decades after implementing a universal chickenpox vaccination programme, due to an increase in zoster mainly in persons aged 50-80 y. This result holds also for scenarios that minimise or counteract the expected increase in zoster incidence (e.g. additional booster vaccinations in adults). However, if the boosting hypothesis is not true or if costs and QALYs are cumulated over at least 33 to more than 100 y after vaccination (depending on the assumptions made), different options for universal 2-dose vaccination against chickenpox in Belgium would be cost-effective at a vaccine price of €43/dose or lower.

Modelling distortions in seroprevalence data using change-point fractional polynomials

This paper shows how to model seroprevalence data using change-point fractional polynomials (FPs). The inclusion of a change point in the FP framework allows to detect distortions arising from common (often untestable) assumptions made in the estimation of the age-specific prevalence and force of infection from cross-sectional data. The method is motivated using seroprevalence data on the parvovirus B19 and the varicella zoster virus in Belgium.

Seventy-five years of estimating the force of infection from current status data

The force of infection, describing the rate at which a susceptible person acquires an infection, is a key parameter in models estimating the infectious disease burden, and the effectiveness and cost-effectiveness of infectious disease prevention. Since Muench formulated the first catalytic model to estimate the force of infection from current status data in 1934, exactly 75 years ago, several authors addressed the estimation of this parameter by more advanced statistical methods, while applying these to seroprevalence and reported incidence/case notification data. In this paper we present an historical overview, discussing the relevance of Muench's work, and we explain the wide array of newer methods with illustrations on pre-vaccination serological survey data of two airborne infections: rubella and parvovirus B19. We also provide guidance on deciding which method(s) to apply to estimate the force of infection, given a particular set of data.

Using empirical social contact data to model person to person infectious disease transmission: an illustration for varicella

With the aim to improve dynamic models for infections transmitted predominantly through non-sexual social contacts, we compared three popular model estimation methods in how well they fitted seroprevalence data and produced estimates for the basic reproduction number R(0) and the effective vaccination level required for elimination of varicella. For two of these methods, interactions between age groups were parameterized using empirical social contact data whereas for the third method we used the current standard approach of imposing a simplifying structure on the 'Who Acquires Infection From Whom' (WAIFW) matrix. The first method was based on solving a set of differential equations to obtain an equilibrium value of the proportion of susceptibles. The second method was based on finding a solution for the age-specific force of infection using the formula of the mass action principle by means of iteration. Both solutions were contrasted with observed age-specific seroprevalence data. The best fit of the WAIFW matrix was obtained with contacts involving touching, and lasting longer than 15min per day. Plausible values for R(0) for varicella in Belgium ranged from 7.66 to 13.44. Both approaches based on empirical social contact data provided a better fit to seroprevalence data than the current standard approach.