The 10,000 Hour Rule: Does It Apply to New Graduates of Residency

Hospital inpatients taken care of by older doctors have a higher mortality rate, a recent article in the American Journal of Medicine concluded. The authors investigated 6572 admissions cared for by 59 different physicians at Montefiore Medical Center in New York. They suggested that older MDs might benefit from remediation to improve their skills. They also speculate about why this seeming non-intuitive finding occurred.

Maxwell S. Kennerly, an attorney blogging on the KevinMD website, proposed that maybe recent graduates of residency programs, having spent more than 10,000 hours as residents, are already “experts” as defined by Malcolm Gladwell’s 10,000 hour rule. That is an intriguing thought. Let’s see if it is plausible.

One must first agree that 10,000 hours of practicing something would lead to an expert level of skill. In his book Outliers, Gladwell gives several examples to try to prove his point. The Beatles played music 8 hours/day and 7 days/week for the better part of three years when they first started. Bill Gates programmed “practically nonstop” for 7 years. The best musicians at the Berlin Academy of Music practiced for more than 10,000 hours as they progressed through the school while the merely average musicians only did 4,000 hours. These anecdotes, while not scientific, bolster Gladwell’s theory. Not mentioned are the negative results. How many people do something for 10,000 hours and never get better? For example, that might be true of my tennis game.

What about internal medicine residents? They train for 3 years less about 4 weeks of vacation/year. If one takes the remaining 144 weeks X 80 hours of work/week, the result is 11,520 hours. That makes them Gladwell-anointed experts, right?

I don’t think so. Unless they underwent particularly rigorous training, they must have slept for some parts of their 80 hour week. They very likely were on overnight call only one or two days/week and sleeping an average of at least four hours/night. Even if they were awake for 70 of the 80 hours/week they worked, they were not practicing medicine the entire time. There are massive amounts of paperwork. There are educational conferences. There is “scut work” [medical jargon for menial tasks] like looking up lab results, calling to arrange x-rays, calling consultants and many more. There are lunches, dinners, snacks and even occasional down time.

Medical residents participating in a study in 2000 said that they were not comfortable performing four of seven basic procedures [central venous line placement, knee joint aspiration, lumbar puncture, and thoracentesis] despite achieving minimum numbers set by the American Board of Internal Medicine.

Even surgical residents, whose training lasts five years [19,200 hours], feel somewhat uneasy at the prospect of working on their own. A study of surgical residents’ work flow revealed that more than 20% of their work time involved non-educational activities. That is the equivalent of one year of their training.

Another recent study showed that for 61 of 121 procedures thought to be essential by surgical residency program directors, the most common number [mode] performed by graduating surgical residents was “0.” After residency, they often take an extra year of training [fellowship] to increase their level of experience before assuming attending surgeon status.

Similarly, Canadian thoracic surgery residents were queried about their level of comfort in performing 18 core thoracic surgical procedures. The residents responded that they were not comfortable performing any of the 18 operations and even the faculty answered that at the end of their training, they were comfortable performing only 7 of the 18.

Assuming that practicing anything for 10,000 hours makes one an expert [a leap of faith at best], I do not think any new graduate of residency training qualifies.

Kennerly’s other theory is that maybe newly minted attending MDs have what is called a “beginner’s mind” and are more open to new ideas and other diagnoses. Although he presented no data to support this and I, of course, am close-minded about it, it is possible.

And as an older MD and a confirmed skeptic, I would like to see the Montefiore study findings replicated before I believe it.