Risk Assessment 2In order to assess risk, you need to understand the reliability of information provided by various sources. We all know, for example. that just because it is written in the newspaper, that does not make it so. This is especially important when considering medical information, because newspapers often sensationalize the results of medical trial or may completely misunderstand the scientific information they are trying to present.
One of the best explanations of evaluating risk that I have seen is Risk in Perspective: A Consumer Guide to Taking Charge of Health Information prepared by the Harvard Center for Risk Analysis. This is a six page brochure designed to help people make sense of conflicting sources of health information. It is a very valuable explanation of risk, and worth reading in full, especially because the cartoons are very funny. The text is serious, though. One section struck me as relevant for our discussions:
Health information can be based on untested claims, anecdotes, case reports, surveys, and scientific studies. Scientific studies, which take samples and apply the results to the whole population, often provide the best clues about health. Nonetheless, many studies are needed to be confident about an answer. The following are some factors that might help you judge information:If you read these guidelines, it is not difficult to understand that most of the homebirth literature and natural childbirth literature fall into the category of less reliable, and is almost always superceded by scientific evidence that is more reliable. So, for example, anecdotes are not reliable since they tell us nothing about what happens to most people. Information that is not published in peer review scientific journals is not reliable compared to information that has been published in a peer reviewed journal. Bits of information scavenged from a variety of studies that were unrelated to the claim being discussed are far less reliable than actual studies of the specific claim.
Less reliable (less certain) More reliable (more certain) One or a few observations Many observations Anecdote or case report Scientific study Unpublished Published and peer reviewed Not repeated Reproduced results Nonhuman subjects Human subjects Results not related to hypothesis Results about tested hypothesis No limitations mentioned Limitations discussed Not compared to previous results Relationship to previous studies discussed
It is important to keep all of this in mind as we move on to discussing the actual risks of various interventions and the actual risks of homebirth.