Analysis: Insights from AAMC Report on Residents

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

Lawpy

42% Full Member
7+ Year Member
SDN Ambassador
Joined
Jun 17, 2014
Messages
63,099
Reaction score
154,731
Some interesting observations found in the 2016 AAMC Report on Residents.

Comparison of Test Scores

From Table B1, the following charts observe the relationship among median MCAT, Step 1 and Step 2 CK scores for first-year (PGY1) residents by specialty in 2015-2016.

Figure B1.jpg


Figure B2.jpg


Figure B3.jpg


The more competitive specialties on average have higher entering MCAT scores and higher Step 1 and Step 2 CK scores. MCAT scores have greater predictive power for Step 1 performance than for Step 2 CK performance based on comparing the R^2 values. But note that doing well on Step 1 is strongly correlated with doing well on Step 2 CK.

Changes in Specialty Preference

Charts derived from Table A1 examine changes in specialty preference throughout medical school. These changes are measured by comparing specialty preferences reported on graduation questionnaire (GQ) survey with specialty preferences reported in matriculating student questionnaire (MSQ).

Figure B4.jpg


Figure B5.jpg


Roughly 25% of medical students reported on the surveys had kept their specialty preferences unchanged throughout medical school.

Enjoy! Feel free to post any interesting observations below and let me know what you think.

Members don't see this ad.
 
Last edited:
  • Like
Reactions: 9 users
Gender Balances and US/Canada Advantage

Table B3 data provide the charts illustrating the percentages of active residents by gender and medical school status (graduated from US/Canada medical schools vs graduated from international medical schools) for each specialty.

Figure B6.jpg


Figure B7.jpg


Figure B8.jpg


Note that women make up the majority of residents in family medicine, OB/GYN and pediatrics, while men make up the majority of residents in anesthesiology, emergency medicine and surgery. Also note the significant advantage towards US/Canadian residents rather than international (IMG) residents, with nuclear medicine being a small exception.

2013-2015 Trends in MD/PhD Residents

The following charts derived from Table B4 visualize changes in number of MD/PhD residents as percent of US MD residents by specialty over time.

Figure B9.jpg


Figure B10.jpg


The percent of active US MD residents who were MD/PhD graduates has been steadily increasing from 3.3% to 3.5% during the 2013-2015 interval.
 
Last edited:
  • Like
Reactions: 1 users
Practice in Medically Underserved Areas

Charts from Table C2 and Table C3 show the percent of physicians who completed training from 2006 to 2015 practicing in federally designated medically underserved areas by specialty and state.

Figure B11.jpg


Figure B12.jpg


The practicing rates in medically underserved areas are greater than 20% for many of the largest specialties, including family medicine, internal medicine, neurology and psychiatry. Nuclear medicine and thoracic surgery are smaller specialties with practicing rates close to 30%. Alabama, Louisiana and Mississippi have among the largest practicing rates (~45%-55%) in medically underserved areas.

Physician Retention Rates

Table C4, Table C5 and Table C6 provide the charts examining the percent of physicians practicing in the state of residency training completed from 2006 to 2015 by specialty, gender and state.

Figure B13.jpg


Figure B14.jpg


Overall, roughly 54% of physicians are practicing in the state of residency training, with female physicians generally having higher retention rates than male physicians across most specialties. Family medicine, pediatrics and integrated vascular surgery have among the highest retention rates ranging around 60%. Texas, Florida, Alaska and Montana have retention rates ranging around 60%, with California having the highest retention rate of about 78%.

Faculty Appointment Status at US MD Schools

The following chart from Table C7 and Table C8 illustrates the percent of physicians who completed training from 2006 to 2015 having a full-time faculty appointment at a US MD school by specialty and rank. Part-time faculty appointments are not included.

Figure B15.jpg


About 17% of physicians currently hold a full-time faculty appointment at a US MD school, the majority of which hold appointments at the assistant professor level. Roughly 76% of physicians have never had a full-time faculty appointment, while 7% had a prior appointment.
 
Last edited:
  • Like
Reactions: 4 users
Members don't see this ad :)
The fact that high MCAT scorers have high Boards scores is no surprise. But thanks for posting...it's good ammo for those loose cannons who refuse to accept that the MCAT has predictive powers.
 
  • Like
Reactions: 3 users
Bumping this to provide charts derived from Table B2. I wanted to see the percent of first-year residents who improved on Step 2 CK for a given Step 1 score. Note that the table and charts examine Step 1 and Step 2 CK score bins, and score improvement is based on Step 2 CK score bin being greater than Step 1 score bin. For example, someone with a Step 1 score of 225 and Step 2 CK score of 229 is considered to have the same performance since both scores are in the 220-229 score bin. However, someone with a Step 1 score of 224 and Step 2 CK score of 232 is considered to have improved (moving from 220-229 score bin to 230-239 score bin).

Below is the Step 1-Step 2 CK improvement curve for first-year residents of the listed 27 specialties in 2015-2016.

Figure B16.jpg


This improvement curve can be divided into large (>1000 residents), medium (100-1000 residents) and small (<100 residents) specialties. Note that the smaller the specialty, the greater the fluctuations seen in the improvement curves.

Large Specialties

Figure B17.jpg


Medium Specialties

Figure B18.jpg


Figure B19.jpg


Small Specialties

Figure B20.jpg


Hope you enjoy these and feel free to share your comments/feedback/thoughts etc.
 
Last edited:
  • Like
Reactions: 1 users
For those in a lower step 1 bracket, you would expect that they would score higher on step 2, no? Especially if they scored very low, there's no where else to go but up. Regression to the mean applies to those who score very high too, though it looks like the effect is less dramatic there.

Great posts btw, I wasn't sure how I felt seeing my specialty of interest among the lowest average step score, but I guess that'll make it easier to get in!
 
  • Like
Reactions: 1 user
The fact that high MCAT scorers have high Boards scores is no surprise. But thanks for posting...it's good ammo for those loose cannons who refuse to accept that the MCAT has predictive powers.
This just shut me up on my innate genius and my first MCAT of a 29 lol
 
  • Like
Reactions: 1 users
For those in a lower step 1 bracket, you would expect that they would score higher on step 2, no? Especially if they scored very low, there's no where else to go but up. Regression to the mean applies to those who score very high too, though it looks like the effect is less dramatic there.

Great posts btw, I wasn't sure how I felt seeing my specialty of interest among the lowest average step score, but I guess that'll make it easier to get in!

That's true although surprisingly, a small percent of those in low Step 1 bin actually scored similarly or worse in Step 2 CK (even in cases for below 190 bin that isn't 100% improvement). In addition to regression to the mean, I think there are two reasons for improvement: Step 2 CK may be significantly easier than Step 1, so people tend to score better on Step 2 CK. And people who did poorly on Step 1 may have worked very hard to improve on Step 2 CK to increase matching success. Likewise, people who did well on Step 1 may not spend much effort studying for Step 2 CK and hence experience a score decline.
 
  • Like
Reactions: 1 user
Top