4 ECTS credits
104 h study time
Offer 1 with catalog number 4007316FNR for all students in the 2nd semester at a (F) Master - specialised level.
In the introduction, some historical examples of research that made use of epidemiological methods are presented. The importance of these historical findings for the fight against diseases, though not directly leading to a real ‘understanding’ of the underlying biological mechanisms, has made 'epidemiology' very popular. There is still much confusion about what exactly is epidemiology, even so in epidemiologists. The student will be confronted with the definition of the theoretical epidemiology as a discipline focused on the study of the occurrence of phenomena that are relevant in the domain of health. Just as in the 'classical' science the study object is presented as a mathematical function that symbolizes the relationship between the measure of occurrence for illness (event or state), death or cure and its determinants. We resolutely choose for this mathematical representation as we feel it is enlightening and it also prepares for the various forms of analysis (mathematical models) that are frequently used in applied scientific research in the field of health and ill-health. In this introduction the scope is split into three major sections, which occur when we split the knowledge (gnosis) relevant to the domain of health care in diagnosis, prognosis and etiognosis.
We briefly address some frequently used measures for the display of the occurrence of events or states (illness, death, healing, relapse). In the study of the occurrence of events relevant to health and ill-health these measures are presented as the dependent variables in mathematical functions.
Diagnosis, the perception (gnosis, knowledge) that the health professional has about the presence of a disease in an individual, is here presented as a problem of uncertainty. In order to be able to handle this uncertainty different approaches are proposed. First, the approach based on the validity attributes of a (dichotomised) diagnostic test and the frequency of the disorder is proposed (Bayes’ theorem). This approach has a number of important draw-backs. However those can be overcome (at least in part) by considering the diagnostic problem as a state prevalence. This prevalence can then be studied as a function of diagnostic indicators (among other determinants of the presence of disease).
Predicting the future (prognosis) is an important part of the activities in modern scientific health care. In a self-study we explore survival analysis as a technique for displaying the frequency of occurrence of events on the basis of a numerical example that we derive from a study into the effectiveness of treatment for acute leukemia from Freireich (1963). With this example we seek to introduce the basic concepts of survival analysis and the graphical presentation of incomplete data in a convenient way. Emphasis is made on the appropriate design of a prognostic study with a focus on typical temporal aspects.
Further attention is paid to causality. Although one easily could argue that a causal relationship exists when it is not explained by any confounding factor, frequently a number of criteria are used to examine the causality of a relationship. The fact that in the domain of health and ill-health an event almost never occurs solely as a result of exposure to one single factor and that most factors which are supplementary required are not known, makes the understanding of causality very difficult. Rothman and Lanes have presented an elegant model for causality which is a good help in dealing with causality. The model also shows clearly the weaknesses of the different criteria used for causality. Based on the model of (multi-) causality we will indicate how the influence of a determinant on the occurrence of events can be quantified. There are various measures of association suggested that each have their own characteristics. We restrict ourselves here to the absolute effect, the relative effect and the ratio component of the relative effect.
This course highlights the importance of systematic errors for the interpretation of research results. We also make the distinction between the bias that can arise from the selection (with limited participation) of individuals from the study population, by the inadequate collection of information on exposures and events from the persons under study and finally by the mixing of the effects studied and the effects of other characteristics relevant to the study result (confounding). We also emphasise on the phenomenon of effect modification (interaction). We specify how stratified analysis and multiple regression analysis offer a solution for both confounding and effect modification.
Nihil
After completion of this course the student is able:
Definitions and objectives:
To define epidemiology
To appoint the various categories of knowledge (diagnosis, prognosis, etiognosis) relevant to the domain of health
To present and discuss the basic form of a study object deduced from a study question related to the domain of health
Occurrence:
To realize the limitations of using absolute numbers when presenting frequency of occurrence
To define and calculate different measures for disease frequency (prevalence, cumulative incidence, incidence density)
Diagnosis:
To demonstrate that determining the presence of a disease is a matter of uncertainty (probability setting)
To present and discuss a general model for the emergence and course of an illness
To formulate the basic form of a diagnostic study object
To describe the appropriate design of a diagnostic study
To indicate the restrictions of tackling the diagnostic problem through diagnostic testing based on cut-off points, the parameters for validity and the application of Bayes’ theorem and which solution can be proposed
Prognosis:
To formulate the basic form of the prognostic study object
To describe the appropriate design of a prognostic study (both descriptive and interventional)
To formulate the objectives of survival analysis
To perform a simple survival analysis based on a dataset including the display of an average survival time, an average hazard rate (h) and a survival curve (according to Kaplan-Meier and Cox regression).
Etiognosis:
To develop a general model for causality
To formulate the basic form of the etiognostic study object
To describe the appropriate design of a etiognostic study
To specify on the basis of a model for (multi-) causality under what condition the influence of a feature on the occurrence of events can be quantified
To discuss the different measures for the relationship between the characteristics studied (the dependent and independent variables), and to make the necessary calculations to quantify these relations
To describe and discuss the various forms of systematic error
To indicate what is meant with confounded study results.
To understand the difference between actual and potential confounding and to ascertain the presence of confounding in study results
To identify the presence of confounding in a study in simple exercises and to control for it on the basis of stratified analysis and multiple (logistic) regression
To indicate effect modification and how to deal with it
To determine and describe the presence of effect modification
The final grade is composed based on the following categories:
Written Exam determines 40% of the final mark.
Other Exam determines 40% of the final mark.
SELF Report determines 20% of the final mark.
Within the Written Exam category, the following assignments need to be completed:
Within the Other Exam category, the following assignments need to be completed:
Within the SELF Report category, the following assignments need to be completed:
Nihil
This offer is part of the following study plans:
Master of Molecular Biology: Standaard traject
Master of Applied Sciences and Engineering: Computer Science: Artificial Intelligence
Master of Applied Sciences and Engineering: Computer Science: Multimedia
Master of Applied Sciences and Engineering: Computer Science: Software Languages and Software Engineering
Master of Biology: Molecular and Cellular Life sciences