ABSTRACT
Essential ingredients of an experiment are a purported manipulable cause and an anticipated effect. The relationship is speculative-if already confirmed there is no need for an experiment. An experiment is a procedure to verify or falsify a causal relationship by introducing the purported cause and observing the effect. This definition requires that a cause must be hypothesized and its possible outcomes visualized in advance. Observations are geared to measure this anticipated outcome. Broadly speaking, the occurrence or nonoccurrence of different maternal compli-
cations in anemic and nonanemic women is an experiment albeit performed by nature, so is the occurrence of goiter of various grades in areas with iodine deficiency in water. When a rare or unique opportunity is available to observe or to study the effects of specific events as a result of naturally occurring changes, it is called a natural experiment. For example, the tsunami of 2005 provided a rare opportunity to study the health consequences of such a disaster. John Snow’s classical discovery that cholera is a waterborne disease was the outcome of a natural experiment. Snow identified two mixed populations, alike in many important respects but different in the sources of water supply to their households. The large difference in the occurrence of cholera among these two populations gave a clear indication that cholera is a waterborne disease. This was demonstrated in the year 1854, long before the advent of the bacteriological era. Such natural experiments generally come under the domain of observational studies as described in the preceding chapter. This chapter concerns experiments with human intervention for changing the
course of events. They are therefore called intervention studies. By exercising control on the extraneous factors that can affect the outcome, experiments are the most direct method to study the cause-effect relationship. Experimental evidence is generally more compelling than that available from observational studies. A medical experiment can be carried out in a laboratory, clinic, or community.
The subjects for experiment in the clinic or community are human beings, and such an experiment is generally termed a trial. Laboratory experiment, on the other hand, may involve inanimate entities such as physical forces or chemicals: in the context of medicine, laboratory experiments are generally conducted on biological material or animals. Laboratory experiments often provide important clues to the potential of the intervention for formulating into a therapeutic agent. When successful, they pave the way for human studies. Thus, such experiments have a special place in medical studies. Even if you do not plan to conduct an experiment yourself, the details
