Hajat Cother
Public Health Programmes, Health Authority-Abu Dhabi, Dubai, United Arab Emirates.
Methods Mol Biol. 2011;713:27-39. doi: 10.1007/978-1-60327-416-6_3.
Epidemiology as defined by Last is "the study of the distribution and determinants of health-related states or events in specified populations and the application of this study to the prevention and control of health problems". Traditional epidemiological studies include quantitative and qualitative study designs. Quantitative study designs include observational and interventional methodology. Observational methods describe associations that are already present at population (descriptive) or individual (analytical) level. Although they form the mainstay of epidemiological studies, observational methods are prone to bias and confounding. These can be dealt with by various means involving both the study design and statistical analysis. Interventional methods involve changing variables in one or more groups of people and comparing outcomes between those with the changed and unchanged variable. Interventional studies can more readily account for bias (such as through randomisation) and confounding (such as through controlling) as is seen in randomised, controlled trials. Qualitative studies employ non-numeric methods to obtain "richer" information on how people perceive or experience situations. Much of epidemiology and epidemiological methods have been stable for many years. There are, however, emerging issues in epidemiology, including those of causal inference, counterfactuals and Mendelian randomisation, among others. There are also several modern and emerging uses of traditional epidemiological techniques in the fields of infectious disease, environmental, molecular and genetic epidemiology.
正如拉斯特所定义的,流行病学是“对特定人群中与健康相关的状态或事件的分布及决定因素的研究,以及将该研究应用于健康问题的预防和控制”。传统的流行病学研究包括定量和定性研究设计。定量研究设计包括观察性和干预性方法。观察性方法描述了在人群(描述性)或个体(分析性)层面已经存在的关联。尽管观察性方法是流行病学研究的主要支柱,但它们容易出现偏差和混杂因素。这些可以通过涉及研究设计和统计分析的各种方法来处理。干预性方法涉及改变一组或多组人群中的变量,并比较变量改变者和未改变者之间的结果。干预性研究能够更轻松地应对偏差(如通过随机化)和混杂因素(如通过控制),就像在随机对照试验中看到的那样。定性研究采用非数字方法来获取关于人们如何感知或体验情况的“更丰富”信息。许多流行病学及流行病学方法多年来一直保持稳定。然而,流行病学中也出现了一些新问题,包括因果推断、反事实和孟德尔随机化等问题。在传染病、环境、分子和遗传流行病学领域,传统流行病学技术也有一些现代的和新出现的应用。