Background

Science is a process of systematically collecting and recording data about the physical world, then categorising and studying the collected data in an effort to explain observed natural phenomena. The scientific method provides a way of understanding the world around us and its purpose in medicine is to attempt to get at the truth. Clinicians with a clear understanding of how research can help them are better able to judge claims about treatments and the evidence supporting them. As new methods of diagnosis and treatment continue to be introduced, the need to question the validity of various therapies in order to determine what is science-based as opposed to what is quackery will exist. For example, whether alternative therapies such as acupuncture and homeopathy are as efficacious as more conventional medicine is a topical question. At the clinical level, the principles of epidemiology provide the tools for evaluating methods of diagnosis and treatment in the clinical environment, and these tools can help clinicians to develop and apply methods of clinical observation that will lead to valid conclusions by avoiding being misled by systematic error and chance.

The field of epidemiology is ancient, dating back to Hippocrates around 400 BC. The classical definition of epidemiology is that it 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 control of health problems' or, more simply, the 'study of the health status of populations'. Epidemiologists make observations about the world as it is (often with no hope of controlling anything) and accept that the interplay of many factors may influence the phenomenon they are trying to study. Epidemiologists rely on the theory of biology, using statistical methods and applying these methods to a variety of disease problems.

What Is Clinical Epidemiology?

As clinicians, you use various sources of information to answer questions that you might have about the clinical cases that you see on a daily basis. Questions such as:

 How well do the various tests for hyperadrenocorticism distinguish among the possible causes of Cushing's disease? in dogs? in cats?

 If pituitary-dependent Cushing's disease is diagnosed, how long can an owner expect their dog to live? without treatment? with treatment? which treatment is best?

 How likely is it that complications will arise? due to the disease? due to the treatment?

What sources of information would you use to help you answer these questions... your own experiences, the advice of colleagues, the medical literature?

Clinical epidemiology can be interpreted to mean 'clinical' as it seeks to answer clinical questions and to guide clinical decision making and 'epidemiology' as the care of individual patients in the context of the larger population and methods used to answer questions. The purpose of clinical epidemiology is to foster methods of clinical observation and interpretation that lead to valid conclusions in order to improve medical decision making. The term evidence-based medicine first appeared in the medical literature in 1992 and can be defined as a method of clinical practice in which the clinician is aware of, and uses, critical appraisal skills to evaluate the strength of the evidence in support of their practice. The way in which observations are made and interpreted determines whether the conclusions you reach are valid and thus, how helpful these conclusions will be to your patients and their owners.

The principles of epidemiology are indispensable in clinical medicine because there is wide variation in manifestations of a disease between individuals. If all individuals responded identically to a disease or its prevention, then we would be able to do detailed experiments on one animal and know how to diagnose, treat and prevent that condition in all others of that species, breed or group. Most diseases can result in a wide variety of severity of signs from death to clinical disease to subclinical disease. There are also differences in risk factors acting on individuals that contribute to the spectrum of disease that we see. Epidemiology is important to clinical practice because it provides a foundation for the logical diagnosis and treatment of disease problems in a group of animals or an individual patient. As a result of the fact that most, if not all, research studies are imperfect, clinicians commonly do not have perfect information on which to base clinical decisions. Veterinary clinicians often have to make decisions using evidence developed in human medicine and this reasoning by analogy offers many opportunities for error.

The Role of Chance

Why Should We Care About Statistics?

Observations about disease are made on a sample of patients rather than all those with the disease in question. These observations may misrepresent the situation in the population as a whole because of chance. The deviation of an observation on a sample from the true population value, due to chance alone, is called random variation (vs. systematic variation). Statistics are a form of mathematics that involves the collection, organisation and interpretation of numerical data, especially the analysis of population characteristics by inference from sampling, to evaluate the effects of chance. Statistical methods allow us to quantify the probability that chance (random variation) accounts for the observed clinical results. The results of statistical tests will be valid only to the extent that the design and execution of the study was able to minimise the occurrence of bias.

Validity of a Study

Internal validity is the truth within a study while external validity or generalisability is the truth beyond a study. Internal validity is the degree to which the results of a study are correct for the sample of patients being studied. A study is considered to be internally valid if the conclusions represent the truth for the individuals studied in as much as the results were not likely due to the effects of chance, bias, or confounding.

Bias, or systematic error (as opposed to random error), is any trend in the collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are systematically different from the truth. Bias cannot be avoided by using large sample sizes; bias can only be reduced through appropriate study design and execution. Almost all studies suffer from bias to some degree. The crucial question is whether or not the results of a study are considered to be invalid due to any identified bias. If enough bias has occurred to invalidate a study, then the statistical assessment of the effects of chance will be meaningless. Since all studies are imperfect, the critical question one must consider is whether or not any identified problems are great enough to conclude that the study results are invalid.

External validity, or generalisability, is the degree to which the results of a study hold true in other settings. Generalisability expresses the validity of assuming that the study patients are similar enough in important characteristics (such as age of patients, previous disease history, disease severity, nutritional status) with other patients in a population. Consideration must be given to the risk that conclusions may only be valid for the population under study. The degree to which a study is generalisable to the population of interest to you is a question only you can answer. A study must first be considered to be internally valid before external validity can be assessed. For an individual clinician, external validity is an answer to the question: 'assuming that the results of this study are true, do they apply to my patient as well?'

References

1.  Hulley SB, Cummings SR, et al. Designing clinical research: an epidemiologic approach (second edition). Philadelphia: Lippincott Williams & Wilkins, 2001.

2.  Rothman K. Epidemiology: An introduction to epidemiology. New York: Oxford University Press Inc, 2002.

3.  Sackett DL, Haynes RB, et al. Clinical epidemiology: how to do clinical practice research (third edition). Philadelphia: Lippincott Williams & Wilkins, 2004.

4.  Strauss SE, Richardson WS, et al. Evidence-based medicine: how to practise and teach EBM (third edition). London: Elsevier Churchill Livingstone, 2005.

5.  Vaillancourt JP, Toma B. Dictionary of veterinary epidemiology. Ames, Iowa: Iowa State University Press, 1999.