Do Younger or Older Doctors Get Better Results?

From a Wall Street Journal essay by Anupam B.Jena and Christopher Worsham headlined “Do Younger or Older Doctors Get Better Results?”:

Imagine you’ve been admitted to the hospital and you’re meeting the physician taking care of you for the first time. Who are you hoping walks through that door? Would you rather they be in their 50s with a good amount of gray hair, or in their 30s, just a few years out of residency?

In a study published in 2017, one of us (Dr. Jena) and colleagues set out to shed some light on the role of age when it came to internists who treat patients in hospitals. These physicians, called hospitalists, provide the majority of care for elderly patients hospitalized in the U.S. with some of the most common acute illnesses, such as serious infections, organ failure and cardiac problems.

In much of medical care, patients choose their doctors based on things like bedside manner, perceived expertise, responsiveness, and attributes that are impossible to deduce. Patients who are hospitalized, however, don’t get a say in which hospitalist will treat them—they’re cared for by whichever doctor happens to be on duty at the time. Those doctors tend to be scheduled to cover the hospital in blocks, perhaps one or two weeks at a time.

Using data from Medicare on patients over age 65 and a database containing doctors’ ages, we identified about 737,000 non-elective hospitalizations managed by about 19,000 different hospitalists from 2011-13. We divided patients into four different groups based on the age of the doctor who treated them: doctors aged less than 40, 40-49, 50-59, and 60 and above.

Older doctors obviously had more years of experience since completing residency, with doctors under 40 having an average of 4.9 post-residency years of experience, increasing to 28.6 years for doctors over 60. Older doctors were also more likely to be male: 61% of doctors under 40 were men, compared to 84% of doctors over 60, reflecting the shift in gender makeup that has occurred in our profession in recent decades.

Some percentage of hospitalized patients will survive or die no matter who their doctor is, but for others, their doctor’s clinical judgment, decision-making, and technical skill could be the difference between life and death. The next step, therefore, was to compare 30-day mortality rates between the four age different groups. Our statistical model found that as doctors got older, their patients had higher mortality rates. The rate for under-40 doctors was 10.8%, increasing to 11.1% in the 40-49 group, 11.3% in the 50-59 group, and 12.1% in the over-60 group.

To put these numbers in perspective, the results suggested if the over-60 doctors took care of 1,000 patients, 13 patients who died in their care would have survived had they been cared for by the under-40 doctors. We repeated the analysis using 60- and 90-day mortality rates, in case longer term outcomes might have been different, but again, the pattern persisted: Younger doctors had better outcomes than their more experienced peers. The inevitable question followed: Why?

There are two possible explanations. The first is that there is a true age effect, wherein simply being older leads to changes in how a doctor practices, resulting in higher mortality. Perhaps older doctors are overly confident in their experience, feeling they have “seen a case like this a million times,” and thus miss tricky diagnoses.

The other, which we think is more likely, is that there are things that older doctors and younger doctors do differently simply because they were trained at different times. Younger doctors possess clinical knowledge that is more current. If older doctors haven’t kept up with the latest advances in research and technology, or if they aren’t following the latest guidelines, their care may not be as good as that of their younger peers.

One way that doctors stay up to date is simply by taking care of patients. When patients come to us with a given diagnosis, it may prompt us to check out the latest research, guidelines, or recommendations for that condition. Medications are the internist’s primary tool; since newer and better drugs are developed at a (relatively) rapid pace, seeing a high volume of patients is a good way to keep up.

To see if this might be the case, we repeated the analysis but this time divided doctors based on both age and case volume. We found that for “low volume” doctors, older doctors had higher mortality. For “medium volume” doctors, the pattern was less pronounced. And for “high-volume” doctors, the pattern went away altogether. In practical terms, as long as a doctor is seeing a sufficiently large number of patients, the doctor’s age is irrelevant to the care they give.

Does this mean that, on balance, younger doctors are “better” than older ones? This study suggests that if “better” is defined as a hospitalist having lower 30-day patient mortality, then we would have to say yes. But what about surgeons, who in addition to their diagnostic skills require technical abilities that depend on experience and muscle memory?

To find out, a separate study by Dr. Jena and colleagues looked at about 900,000 Medicare patients who underwent common non-elective major surgeries (for example, emergency hip fracture repair or gall bladder surgery) performed by about 46,000 surgeons of varying age. We chose non-elective surgery since patients don’t have a whole lot of control over their surgeon when they come in with an urgent or emergent problem. As with hospitalists, they’ll end up assigned to the surgeon on duty in an as-good-as-random fashion. Just as before, patients were divided into four groups based on the age of their surgeon, and we used a statistical model to calculate the 30-day mortality rate following surgery.

The results showed that unlike hospitalists, surgeons got better with age. Their patient mortality rates had modest but significant declines as they got older: mortality was 6.6% for surgeons under 40, 6.5% for surgeons age 40-49, 6.4% for surgeons age 50-59, and 6.3% for surgeons over age 60.

Clearly something different was happening here. It may be that for hospitalists, the benefit of steadily increasing experience starts to be outweighed by their waning knowledge of the most up-to-date care. It’s different for surgeons, though, who hone many of their skills in the OR. Surgeons build muscle memory through repetition, working in confined spaces with complex anatomy. They learn to anticipate technical problems before they happen and plan around them based on prior experience. Over time, they build greater technical skills across a wider variety of scenarios, learn how to best avoid complications, and choose better surgical strategies.

What does this mean for all of us as patients when we meet a new doctor? Taking studies of hospitalists and surgeons together, it’s clear that a doctor’s age isn’t something that can be dismissed out of hand—age does matter—but nor can it be considered in isolation. If we’re concerned about the quality of care we’re receiving, the questions worth asking aren’t “How old are you?” or even “How many years of experience do you have?” but rather “Do you have a lot of experience caring for patients in my situation?” or “What do you do to stay current with the research?”

Drs. Jena and Worsham are researchers at Harvard and practice medicine at Massachusetts General Hospital. This essay is adapted from their new book, “Random Acts of Medicine: The Hidden Forces That Sway Doctors, Impact Patients, and Shape Our Health,” published July 11 by Doubleday.