Similarly, I'm pretty sure that "code written by programmers who don't program often is more likely to have bugs", and "pilots who don't fly often are more likely to crash".
(Strictly speaking, you'd want to normalize the risk of bugs / crashes per some unit of output e.g. per hour flown for pilots.)
I don't think 'slightly higher' is accurate. The report says, "We classified physicians into thirds of patient volume: low (estimated number of total admissions <90 per year), medium (91-200 admissions), and high (>201 admissions)". So you have two-thirds of doctors whose median volume is 90 per year, and one third of doctors whose median volume is well above 201 per year. That seems to be a significant difference.
> That in turn suggests the effect was pretty small.
For two-thirds of doctors the effect is: "adjusted 30 day mortality rates were 10.8% for physicians aged <40 (95% confidence interval 10.7% to 10.9%)...and 12.1% for physicians aged ≥60 (11.6% to 12.5%)". I'd say that's a significant effect - certainly something that should be included in the title.
(I don't mean to state the obvious but this wasn't made explicitly clear enough to me.)
Edit: My initial comment here is probably wrong based on the detailed section "Adjustment variables":
> Patient characteristics included age in five year increments, sex, race or ethnic group (non-Hispanic white, non-Hispanic black, Hispanic, other), primary diagnosis (diagnosis related group), 27 comorbidities (Elixhauser comorbidity index22), median household income of zip code (in 10ths), an indicator for dual Medicare-Medicaid coverage, day of the week of the admission date (to account for the possibility that severity of illness of patients could be higher on specific days of the week), and year indicators.
Another possibility is that younger staff are more likely to be questioned about things. "Have you thought about X?" causes them to rethink something and make a revision. If older staff are just assumed to know what they're doing, they might be questioned less.
The fact that differences weren't present among physicians with a large volume also makes me wonder if this is just a fishing expedition that wouldn't replicate. Not to cast aspersions on the authors; just to say that if you slice up any dataset enough you can find something.
Finally, the differences aren't huge...
This group of authors is well respected and known for doing studies like this. However, if their chief interest is observing the effect of physician age on patient outcome, physician age should clearly be treated as a continuous or truncated variable here.
Edit: In the supplement [1. Table B], they do perform the calculation with physician age as a continuous variable, and the effect stands. Good on them for doing the math in this way.
1 = http://www.bmj.com/content/bmj/suppl/2017/05/15/bmj.j1797.DC...
Physician age was modeled both as a continuous linear variable and as a categorical variable (in categories of <40, 40-49, 50-59, and ≥60) to allow for a potential non-linear relation with patient outcomes.
can you explain what it means? Specificially, how does making it categorical allow for it, and keeping it continuous prevent it?
When categorizing in buckets, they probably do an ANOVA. This technique posits that the average does vary per category exactly as measured, and asks the question: If I tell you the category, how much is the variance of your data reduced? If the variance falls a lot (relatively to what it was), it means there's a statistically significant effect between the category and your variable.
And, in their defense, they can't really go fishing for different continuous relationships once they have the data, as that'd reduce their statistical power.
Of interest too is the number of adjustable parameters of the model:
If instead of four age categories you use, say, four hundred, you end up splitting each doctor into one category. The predictive power of that model is greatest, with very good statistical significance, but you have achieved no insight at all.
Similarly when taking age as continuous; if instead of a straight line you fit a curve with four hundred free parameters, you overfit it to the point of destroying any insight.
So in that sense it's "unfair" that they used four age categories, vs two free parameters of a linear regression. And there would need to be some explanation as to the age ranges they used for each category.
"JENA: Exactly. So patients more or less end up getting quasi-randomized to physicians with different characteristics. So for example if you happen to get hospitalized in the first week of May, you may be treated by a group of doctors who on average have five years’ less experience than if you happen to get hospitalized in the second week of May. And we can basically see what happens if a patient happens to be treated by a doctor who is 20 years out of residency versus 5 years out of residency. And what we find is that if you happen to be treated by a doctor who is 10 years or 15 years out of residency, your mortality within thirty days of being hospitalized is higher."
"The effect of senior obstetric presence on maternal and neonatal outcomes in UK NHS maternity units: a systematic review and meta-analysis"
> Fifteen studies fulfilled the inclusion criteria, presenting data from 125 856 births. Overall, there was no significant difference between lesser and increased consultant presence for any outcome. When data were stratified by comparison type, the likelihood of emergency caesarean section was significantly lower (odds ratio, OR 0.91; 95% confidence interval, 95% CI 0.86–0.96) and the likelihood of non-instrumental vaginal delivery was significantly higher (OR 1.07; 95% CI 1.02–1.12) when the rostered hours of consultant presence per week were increased.
There can be all different sorts of anecdotes, but the link was about a study with more than 100k doctors, which would be a stronger evidence that the probability of survival is better for patients with younger doctors.
I daresay you drew the wrong conclusions from this incident.
All things being equal, a doctor with more experience and a lower rate of complications should be your preferred choice for any medical procedure.
They lose enthusiasm and energy. They are inclined to study less. They perform fewer procedures due to diminishing motor, visual-spatial and cognitive skills.
They suffer the same ailments as their patients - hypertension, diabetes, arthritis, dementia; divorce, overwork, errant children (that's a whole category of grievance and burden !) as well as sundry personal issues.
There is the drudgery. Paperwork. After a few decades you have seen it all: the rare cases don't excite as much. They are just another presentation of the human condition.
Finally,medicine is a hierarchical discipline, younger practitioners are discouraged from questioning (dubious) decisions made by their seniors.
Older practitioners compensate for these deficits with an accumulated body of experience that a younger doctor simply does have.
The best age(s) to be a doctor is mid-thirties to early fifties. Just like most other professions.
"patients’ adjusted 30 day mortality rates were 10.8% for physicians aged <40 … and 12.1% for physicians aged ≥60 … Among physicians with a high volume of patients, however, there was no association between physician age and patient mortality."
All of the patients were 65 or older with a medical condition. Perhaps the oldest and sickest are regularly routed to older doctors? As ever, correlation does not imply causation:
So it's not just age per se, but age plus lack of case quantity.
Which now raises the questions:
1) How are patients assigned to doctors? Is there something else that effects outcomes?
2) Why are some docs carrying less patients? Is that the cause, and age the correlation?
Even in my professional work experience, the new problems that may fall across multiple domains tend to be identified/solved by generalists rather than specialist.
I don't know if that is still the case or indeed was ever a 'real' thing.
Authors' conclusions reproduced below:
"Within the same hospital, patients treated by older physicians had higher mortality than patients cared for by younger physicians, except those physicians treating high volumes of patients."