arXiv subjects
Publications and source records attributed to Ralph Brinks.
This paper describes types of errors arising in a recently proposed method of incidence estimation from prevalence data. The errors are illustrated by a simulation study about a hypothetical irreversible disease. In addition, a way of obtaining error bounds in practical applications of the method is proposed.
Background: A common modelling approach in public health and epidemiology divides the population under study into compartments containing persons that share the same status. Here we consider a three-state model with the compartments: A, B and Dead. States A and B may be the states of any dichotomous variable, for example, Healthy and Ill, respectively. The transitions between the states are described by change rates (or synonymously: densities), which depend on calendar time and on age. So far, a rigorous mathematical calculation of the prevalence of property B has been difficult, which has limited the use of the model in epidemiology and public health. Methods: We develop an equation that simplifies the use of the three-state model. To demonstrate the broad applicability and the validity of the equation, it is applied to simulated data and real world data from different health-related topics. Results: The three-state model is governed by a partial differential equation (PDE) that links the prevalence with the change rates between the states. The validity of the PDE has been shown in two simulation studies, one about a hypothetical chronic disease and one about dementia. In two further applications, the equation may provide insights into smoking behaviour of males in Germany and the knowledge about the ovulatory cycle in Egyptian women. Conclusions: We have found a simple equation that links the prevalence of a dichotomous variable with the transmission rates in the three-state model. The equation has a broad applicability in epidemiology and public health. Examples are the estimation of incidence rates from cross-sectional surveys, the prediction of the future prevalence of chronic diseases, and planning of interventions against risky behaviour (e.g., smoking).
In this work we describe a non-parametric disease model that links the temporal change of the prevalence of an infectious disease to the incidence and the recovery rates. The model is only based on the common epidemiological measures incidence and recovery rate. As an application, the model is used to calculate the prevalence of influenza in Germany for a hypothetical birth cohort during 2001 and 2013.
The illness-death model of a chronic disease consists of the states 'Normal', 'Disease' and 'Death'. In general, the transition rates between the states depend on three time scales: calendar time, age and duration of the chronic disease. Previous works have shown that the age-specific prevalence of the chronic disease can be described by differential equations if the duration is negligible. This article derives a partial differential equation (PDE) in the presence of duration dependency. As an important application, the PDE allows the calculation of the age-specific incidence from cross-sectional surveys.
Secular trends of the incidence of NCDs are especially important as they indicate changes of the risk profile of a population. The article describes a method for detecting secular trends in the incidence from a series of prevalence data - without requiring costly follow-up studies or running a register. After describing the theory, the method is applied to the incidence of diabetes in Denmark.
Relevant events in a three state illness-death model (IDM) of a chronic disease are the diagnosis of the disease and death with or without the disease. In this article a simulation framework for populations moving in the IDM is presented. The simulation is closely related to the concept of Lexis diagrams in event history analysis. Details of the implementation and an example of a hypothetical disease are described.
This paper generalizes a previously published differential equation that describes the relation between the age-specific incidence, remission, and mortality of a disease with its prevalence. The underlying model is a simple compartment model with three states (illness-death model). In contrast to the former work, migration- and calendar time-effects are included. As an application of the theoretical findings, a hypothetical example of an irreversible disease is treated.
In this work we examine the properties of a recently described ordinary differential equation that relates the age-specific prevalence of a chronic disease with the incidence and mortalities of the diseased and healthy persons. The equation has been used to estimate the incidence from prevalence data, which is an inverse problem. The ill-posedness of this problem is proven, too.
In epidemiology, the person-years method is broadly used to estimate the incidence rates of health related events. This needs determination of time at risk stratified by period, age and sometimes by duration of disease or exposition. The article describes a fast method for calculating the time at risk in two- or three-dimensional Lexis diagrams based on Siddon's algorithm.
This article reformulates a common illness-death model in terms of a new system of stochastical differential equations (SDEs). The SDEs are used to estimate epidemiological characteristics and burden of systemic lupus erythematosus in England and Wales in 1995.
This paper descibes a new method of calculating the mean duration and mean age of onset of a chronic disease from incidence and mortality rates. It is based on an ordinary differential equation resulting from a simple compartment model. Applicability of the method is demonstrated in data about dementia in Germany.
This paper descibes a new method for deriving incidence rates of a chronic disease from prevalence data. It is based on a new ordinary differential equation, which relates the change in the age-specific prevalence to the agespecific incidence and mortality rates. The method allows the extraction of longtudinal information from cross-sectional studies. Applicability of the method is tested in the prevalence of dementia in Germany. The derived age-specific incidence is in good agreement with published values.