There is an emerging realization that as clinical providers, we need systems or methods to help sort and evaluate the utility of information before we apply it in clinical care. Without such systems, when we are facing an overload of information, most of us tend to take the first or the most easily accessed information—our clinical experience or the drug detailer’s latest data. Perhaps we overestimate the quality and utility of that information while we remain unaware of potentially more useful and relevant information. As a result, the quality of care we provide tends to become suboptimal and the cost of care increases.
Clinical practice is based on knowledge, skills, beliefs, and research. Knowledge can become outdated quickly and skills can get rusty. The traditional methods of improving skills and keeping current with new diagnostic and treatment approaches include reading professional materials, accumulating clinical experience, attending CME conferences, learning from colleagues, and conducting research. There are significant problems, however, with each of these strategies.
The inferences we draw from our individual successes and failures in clinical encounters, with or without evidence, shape our beliefs, which, in turn, influence our clinical practice. These inferences can be distorted by overvaluing treatment successes and undervaluing treatment failures. Evidence-based medicine does not consider one’s clinical experience to be listed among the levels of evidence, and expert opinion is either viewed as the lowest level of evidence1 or has fallen off the list completely.2
Attending CME events has been shown to have little or no effect on clinical practices.3,4 Learning from colleagues’ clinical experience can have similar recall problems. Conducting one’s own research is sel-dom done because it is too time-consuming. The challenge is that the knowledge base continues to grow at a rate that is impossible to keep up with unless we devise effective and efficient methods to identify, evaluate, consolidate, and distribute new knowledge and provide tools for real-time analysis and decision support.
Rate of growth in knowledge base
It is estimated that if we took all the knowledge that humankind had accumulated by the year 1 AD as equal to 1 unit of information, it probably took about 1500 years, or until the 16th century, for that amount of knowledge to double. The next doubling of knowledge (from 2 to 4 “knowledge units”) took only 250 years, until about 1750 AD. By 1900, 150 years later, knowledge had doubled again to 8 units.5
The rate at which information has been doubling continues to pick up speed. It is now estimated that the collective sum of all printed knowledge is doubling every 4 years.6 Between 1999 and 2002, the amount of new information stored on paper, film, magnetic tape, and optical media was estimated to have about doubled, with a growth rate of about 30% a year.7 President Clinton8 had the following to say about this subject in his 1998 remarksto the National Association of Attorneys General:
The sheer volume of knowledge is doubling every 5 years now. . . . because of human genome research, we are literally solving problems in a matter of days that took years to solve not long before I took office. The World Wide Web is growing by something like 65,000 websites an hour now. When I took office, there were 50—(laughter)—50.
Medical knowledge is now reported to be doubling every 8 years,9 and medical professionals are struggling more than they ever have to keep up. Of course, this is not something that is entirely new. Stewart10 wrote the following almost 50 years ago:
A brilliant student who might successfully master all the present facts and theories by graduation would be seriously out of date 10 years later, and hopelessly so by retirement age unless he continued his education while in practice.
