People Treated by Younger Doctors Are Less Likely to Die

According to a new study of 730,000 patients.

Doctors matter. It's an obvious point: Who delivers your health care may affect your treatment. Recently we've seen studies that suggest that patients treated by foreign doctors are less likely to die; female doctors, too, are less likely than their male counterparts to see their patients die. And now research links treatment by younger doctors with lower death rates among patients.

The authors of the new study, published today in The BMJ, caution that their findings should be considered exploratory—as yet, there's no evidence that death rates are higher because doctors are older. That'd be a causal relationship, and so far there's only a link between the two. Interestingly, the link disappears among physicians who treat a high volume of patients; the authors suggest that treating a greater number of people could be "protective" of clinical skills.

To tease out the differences among physicians of varying ages, researchers from the Harvard T.H. Chan School of Public Health looked at data on Medicare patients aged 65 and over. They analyzed 30-day mortality (deaths within 30 days of admission to a hospital), readmissions, and costs of care for a random sample of more than 730,000 patients. Those patients were treated by nearly 19,000 general internists in hospitals across the United States between 2011 and 2014. The physicians' average age was 41.

After adjusting for variables like patient risk factors and time of admission, the 30-day mortality rates for physicians younger than 40 was 10.8 percent; for those over 60, it was 12.1 percent. (The mortality rates for the two other groups, doctors in their 40s and those in their 50s, were 11.1 percent and 11.3 percent, respectively.) Even within the same hospital, people treated by older doctors had higher mortality rates than those seen by younger ones. Readmissions didn't vary with doctor age, but costs of care were slightly higher among older doctors.


Adjusted association between physician age and patient mortality with linear spline model. Multivariable logistic regression model with linear splines was used with knots placed at physician age of 40, 50, and 60, adjusted for patient and physician characteristics and hospital fixed effects. Solid line represents point estimates, and shaded area represents 95% CI around these estimates. (The BMJ)

A 1.3-percent spread in mortality rates looks like a small difference. In practice, though, if the link proves causal, it would mean that for every 77 patients treated by doctors over age 60, one fewer patient would die within 30 days of admission if those patients had been cared for by a doctor younger than 40. Though, again, that difference disappears among doctors treating a high volume of patients. The study didn't account for differences in nurse staffing and previous studies have found that lower mortality rates in high-volume settings is contingent on nurse staffing.

The study authors stress that it's hard to draw conclusions about cause and effect from this observational data. And without those conclusions it's difficult to suggest improvements; in that way, it's similar to the studies about outcomes among patients of women and international doctors.


An editorial response to the study notes that younger doctors performed better when it came to mortality rates, perhaps because they had more recent training, but older, veteran doctors might do better on measures not included in the study, "particularly communication with patients and decisions informed by experience, leading perhaps to fewer invasive medical procedures at the end of life." Still, perhaps more continuing education on clinical guidelines should be in order for older docs.

The editorial authors also write that this kind of research into patient outcomes "is providing much needed evidence to inform clinical practice, educational innovation, organizational redesign, and healthcare policy." The challenge lies in deciding how to act on that evidence.

​

Source