What is the half-life? The old joke about that used to be an insult to various communities. In the old days, a "where do you live" response might be followed by a "well, if you call that living." It was a backhanded put-down of whatever community you were digging at the moment. A similar punchline was "Sounds more like a half-life to me."
The thought came to me in a recent presentation when the concept of our expanding knowledge arose. An attendee made a reference to the "half-life" of information that was inspiring. A "half-life" is usually associated with the rate of decay in radioactive elements or used to describe how long a drug or medication volume takes to be reduced by half in the human body.
However, in the context of medical knowledge, the question is "What is the half-life of what we know?"
An article in The American Clinical and Climatological Association Journal describes Challenges and Opportunities Facing Medical Education. The author contends that early in the 21st century, we met a crossroads in education and information. The result of tools we have built to expand, corral, and coalesce information.
Think for a moment of the inverse of a "half-life" that measures erosion or decay, and instead think of accretion (accumulation). The paper notes that:
"It is estimated that the doubling time of medical knowledge in 1950 was 50 years; in 1980, 7 years; and in 2010, 3.5 years. In 2020 it is projected to be 0.2 years—just 73 days."
That was five years ago. Amazing as the statistic is, imagine where we are today.
That estimation made me think of an exercise from gym class in middle school. The coach lined us all up on the back wall of the gym (I suspect, in retrospect, that it was raining, and the coach had not made a lesson plan). We were told to walk halfway to the other end, the wall. We did. Then we were told to walk halfway again. After three iterations, we were asked how many more iterations it would take before we reached the wall.
The answer seems simple, but after letting us flounder a bit, Coach's answer was "you never will." Because each iteration will only take you half the remaining distance after the last movement. Eventually, the movements will become so minute that it is likely irrelevant and, with a room full of middle schoolers, perhaps comical. Nonetheless, the ideas of halving or doubling were there for our consumption.
We have clearly and unequivocally established that I am no scientist. Thus, I proceed today on the foundation of Dr. Peter Densen (who wrote the article cited above).
The doubling between 1950 and 1980, from every 50 years to every 7, was a decrease of 86%. That is an 86% decrease achieved over 43 years, about an average of 2% per year in terms of delay awaiting the next doubling.
The doubling between 1980 and 2010, from every 7 years to every 3.5, was a decrease of 50%. That is a 50% decrease over 30 years, still about an average of 2% per year.
The doubling between 2010 and 2020, from every 3.5 years to every 0.2 years, was a decrease of 94%. That is a 94% decrease over 10 years, an average of about 9% per year.
If that pace is maintained, then by 2030 we may be doubling weekly. A 94% decrease would bring us to just over 4 days. At the rate illustrated between 1950 and 2010, 2% per year, in 2030 we will be doubling in just under 60 days. By 2050 or before, medical knowledge could actually be doubling each week. The potentials are simply astounding.
This pace described above was all before the advent of artificial intelligence (AI). Without question, the volume of medical knowledge and the pace of accumulation (accretion) has nonetheless been influenced and aided by the advent of other tools such as computers, innovative testing equipment, and brilliant minds.
This all reminded me of Moore's Law. The co-founder of Intel hypothesized in 1965 "that the number of transistors in an integrated circuit (IC) doubles about every two years." He essentially predicted that computers would become twice as productive and effective every two years. He was spectacularly wrong; the capacity and capability have doubled at a far greater pace, and innovation continues to occur daily.
So, how many iterations until we get to that far wall? I would suggest that there is merit in the term "zero" in this analysis. Many think that is a number, but that is fallacious at best. Zero is a placeholder representing a null set (a non-number). But, just as those middle schoolers can never reach the wall by making iterations that each take them half of the remaining distance, it is as unlikely that the downward trend in "days to double medical knowledge" can go on indefinitely.
Taken to the extreme, the knowledge might eventually reach a pace of doubling every second, or even every nanosecond (thousandth of a second). Nonetheless, it cannot double every negative one second or minute. Well, unless some physicist somewhere finds some way for time to work in a non-linear fashion.
The theme of Challenges and Opportunities (2011) is that the pace renders medical school almost insurmountable. This is not new. Over a decade ago, in 2011, the pace was too much. Like trying to drink from a fire hose, the volume and flow are simply unmanageable. The theme revolves around rethinking medical schooling, and it is an intriguing analysis.
In the broader context, this fits with the advice recently rendered by AI experts that obtaining a doctoral degree may be a poor investment of time and money. See HealthLeaders. We are rapidly approaching a moment in which the doctor today cannot necessarily rely on what she knew yesterday. And, as important, she cannot hope to master the knowledge of today before a new torrent will arrive tomorrow morning.
In keeping with the trend lines described above, we may reach a moment in which the scientific knowledge doubles between the time the doctor begins her patient's examination and the moment she concludes it. The potentials and the potential absurdity simply boggle the mind.
There are great advances being made persistently. There is vast knowledge being gathered. The challenge will be to capture, categorize, and organize that knowledge. The doctor of tomorrow (today?) needs to be able to access and leverage that information, contribute to that information without having to ever study or memorize that discrete information.
Medical training, and perhaps more, will have to evolve from a rote memorization or even familiarization of what one needs to know to a mastery of the learning, interpreting, and actualizing of the wealth of available unlearned information. There is a new day dawning. But lately one seems to be dawning every day, if you get my drift.
