A fundamental principle of pharmacology is that all drugs have multiple actions. Actions that are desirable in the treatment of disease are considered therapeutic, while those that are undesirable or pose risks to the patient are called "effects." Adverse drug effects range from the trivial, e.g., nausea or dry mouth, to the serious, e.g., massive gastrointestinal bleeding or thromboembolism; and some drugs can be lethal. Therefore, an effective system for the detection of adverse drug effects is an important component of the health care system of any advanced nation. Much of the research conducted on new drugs aims at identifying the conditions of use that maximize beneficial effects and minimize the risk of adverse effects.
The intent of drug labeling is to reflect this body of knowledge accurately so that physicians can properly prescribe the drug; or, if it is to be sold without prescription, so that consumers can properly use the drug.
The current system of drug investigation in the United States has proved very useful and accurate in identifying the common side effects associated with new prescription drugs. By the time a new drug is approved by the Food and Drug Administration, its side effects are usually well described in the package insert for physicians. The investigational process, however, cannot be counted on to detect all adverse effects because of the relatively small number of patients involved in premarketing studies and the relatively short duration of the studies.
Animal toxicology studies are, of course, done prior to marketing in an attempt to identify any potential for toxicity, but negative results do not guarantee the safety of a drug in humans, as evidenced by such well known examples as the birth deformities due to thalidomide.
This recognition prompted the establishment in many countries of programs to which physicians report adverse drug effects. The United States and other countries also send reports to an international program operated by the World Health Organization. These programs, however, are voluntary reporting programs and are intended to serve a limited goal: alerting a government or private agency to adverse drug effects detected by physicians in the course of practice. Other approaches must be used to confirm suspected drug reactions and to estimate incidence rates. These other approaches include conducting retrospective control studies; for example, the studies associating endometrial cancer with estrogen use, and systematic monitoring of hospitalized patients to determine the incidence of acute common side effects, as typified by the Boston Collaborative Drug Surveillance Program.
Thus, the overall drug surveillance system of the United States is composed of a set of information bases, special studies, and monitoring programs, each contributing in its own way to our knowledge about marketed drugs. The system is decentralized among a number of governmental units and is not administered as a coordinated function. Still, it would be inappropriate at this time to attempt to unite all of the disparate elements into a comprehensive surveillance program. Instead, the challenge is to improve each segment of the system and to take advantage of new computer strategies to improve coordination and communication.