Mollaret (H H)
Mollaret (H H)
ANIMAL EPIDEMIOLOGIC SURVEYS
Bénet (J-J), Sanaa (M), Dufour (B) & Toma (B)
The aim of this paper is to present the main principles of analytical and descriptive epidemiological methods to a non-specialist public to help him for a critical reading of epidemiological papers. First, the authors make a review of the basic principles of quantification in epidemiology, focusing on the notion of epidemiological indicator, the construction of which is presented as well as its difficulties. Then the common methodological components of the procedures for collecting data and their quality criteria are listed. The presentation of deceptive methods deals mainly on the different sampling principles. Analytical methods are compared to Claude Bemard’s experimental model: case control study and cohort study as well as cross sectional study. The way to interpret a causal association is also developed. At the end, the modalities of strategic choices between the different kinds of surveys are discussed.
Sanaa (M), Gerbier (G), Eloit (M) & Toma (B)
The aim of this paper is to describe the different steps of a sampling plan within the context of descriptive studies in animal epidemiology: definition of the population to study, choice of the sampling method and calculation of the size of the sample. The choice of the sampling plan is linked to the objectives of the study, to the observation tools and to the available funds. The principles of sampling within the frame of the qualification of a herd or of a region are also developed. The calculation of the size of the sample and the interpretation of the results is only possible, within this qualification frame, with an estimation of the sensitivity and the specificity of the qualification process at the herd level.
Carlier (V), Vanelle (A-M) & Chaubert Duffour (C)
General considerations on sampling and control plans in food hygiene are followed by more specific elements on sampling food for bacteriological, parasitological, physical, biochemical or histological examination.
Lassauzet (M-L), Lopez (C), Faye (B) & Madec (F)
Sampling methods used in analytical epidemiological studies are reviewed with respect to sampling schemes and samples size, according to the type of study, prospective or retrospective, and to the study objective, exploratory or explanatory. Since, in veterinary medicine, most analytical studies present deficiencies in their sampling methods for practical reasons, some validation procedures are recommended. Examples from studies performed in France are presented.
The quality of the results of a study is closely linked to the quality of the questionnaire. In this paper we discuss the positive and negative aspects of the different kinds of questionnaires (open, closed, mixed). The way the questions are asked (interview, mail, phone), the formulation of the questions, the presentation of the questionnaire and its evaluation are also presented.
Sulpice (P), Lescourret (F), Calavas (D), Bugnard (F) & Faye (B)
Data used in animal epidemiology surveys are complex, not always a priori well formalized and they have various origins. Their management is one of the key point of a good standard study; it has to be organised and realised from the beginning of the study planning. The data management involves, throughout the study, the data collection planning and the checking of the data quality (i.e. to check the different procedures used to produce the data). These two aspects of the data management are both required to obtain reliable and accurate data, critical point to perform good quality epidemiological research. This paper describes some methods and means useful to fulfil the goals of a good data management.
Dufour (B), Maurice (Y) & Idriss (A)
All the bias that can be found during studies in animal epidemiology have been classified following the chronology of a study. The first of the two papers presents this classification, gives definitions and examples coming from developed and developing countries for conception bias: financial bias, project first evaluation bias, objective bias and strictly speaking conception bias, sampling bias: representativeness bias, discrepancy bias, bias linked to the person who answers, censoring bias and entering bias.
Maurice (Y), Dufour (B) & Idriss (A)
Conception and sampling bias have been presented in a first paper. Here this second paper gives definitions and examples coming from developed and developing countries for measurement bias: strictly speaking measurement bias, recording bias, and analysis bias; bias in sampling data: bias in ordering, observations, confounding factor bias, and bias in the choice of treatment methods; interpretation bias: subjectivity bias and data interpretation bias. Otherwise, an attitude for the people in charge of the study in front of these different bias is proposed.
This presentation deals with the main principles followed to analyse the results of an analytical study. First the different measures and the association tests between a studied factor and the disease are presented. Then the notions of interaction factor and of confounding factor and how to take them into account during the analysis are developed. The main statistical models used in epidemiology and their adaptation to animal epidemiology are rapidly presented.