Charting Outcomes in the Match v 2

[achybrachy]

---

Members do not see this ad.

Many of you have probably already seen; the newest edition of "Charting Outcomes." http://www.nrmp.org/data/chartingoutcomes2007.pdf

I'd say all in all it looks about the same as last year, maybe a little tougher for radonc. There are a couple interesting new figures though, like the number of distinct specialties applied to. I didn't realize how few people rank a backup program.

 

[Gfunk6]

Thanks for posting this very interesting data. It seems that the SD for Step I scores has tightened up. Also, it was nice that they included "Top 40 NIH-funded med school" as another variable. It does not seem to have made a big statistical difference in matching however.

Other random thoughts,

* It looks like ranking at least 7 places may be the magic # to really improve your chances of a successful match.
* For US Seniors w/ Step 1 scores b/w 191-220 significantly MORE than 50% matched. I guess this demonstrates the relatively low importance of Step I in RadOnc matching.
* More that half of applicants did NOT submit Step 2 scores in time for the Match and this did not seem to adversely affect their outcome.

 

[achybrachy]

AOA dropped even further in importance; that is, match success for non-AOA applicants jumped from 69% to 78%. On the other hand, those who were AOA were still basically guaranteed to match.

As for the number of ranks, I always thought that was a fishy statistic, or at least not very helpful. I am guessing that most people rank all the places they interview, or maybe leave one off if it really stinks. Thus all this statistic says to me is that the more competetive applicants (who get more interviews) have a better chance of matching. What I would like to know is, if I get 15 offers but only interview at 10 of them, how different are my chances compared to a person who interviews at 15?

 

[deleted4401]

In comparing to other specialties ...

- No other specialty was close in % of PhDs (21% vs 11% in Derm, and everyone else in low single digits)
- Only plastics had more 'pedigreed' students (60% vs. 56%)
- Boards were lower than in Derm, ENT, Plastics, but same as Rad-Dx
- No other specialty had as much research publications (or even close; 2-3 vs. 6)
- AOA rates were lower than in Derm, ENT, Plastics, Ortho, similar to Rad-Dx

Numbers don't lie ... top school, research, and a PhD are crucial ... All the deans at med schools need to read this stuff to give applicants proper information.

-S

 

[stephew]

a phd is not crucial.

 

[Brim]

That document has a wealth of information in it. Very cool.

I think Chart 16 is the most telling.

The whole linear regression thing, however, is a bit silly.

Unfortunately, in their regression model, the best predictor variable is the number of "contiguous ranks" -- (are there non-contiguous ranks??) -- but isn't this number really a surrogate for how many interviews candidates are getting? Obviously, the more desirable candidates are getting more interviews, and thus ranking more programs. Those who rank 4 or fewer programs have a lower chance of matching, but it's not because they only decide to rank 4 programs; it's because they only interview at 4 programs. However, the NRMP doesn't see this logic and states:
"The principal message of these graphs is that applicants with longer rank order lists are more successful than those with shorter ones. The NRMP has been recommending longer list for many years, but many applicants apparently do not heed the advice."

Having the number of interviews be the #1 predictor variable is poor. Most people want to know what their chance of matching in a specialty is BEFORE they apply. This is like not knowing how sucky the odds of winning are until after I spend my entire paycheck on lotto tickets. Not very helpful.

Anyway... The point is, if anyone actually computes an odds ratio to predict their chance of matching, they should be slapped. And then slapped again ... with a Cerrobend gauntlet.

 

[Thaiger75]

a phd is not crucial.

Yes, a phd is NOT crucial, but pretty much nearly guarantees matching success. What the stats do say is that if you're AOA or a PhD, you have a 97% chance of matching.

Only plastics and derm have higher unmatched rates than we do.

Again, according to these stats, there were 200 applicants combined, but that's not including those who didn't rank programs or get interviews. I'd say from what I hear, there are close to 250-300 total applicants a year who attempt to apply.

Another take home is that those who have below average board scores can and do match (30-40% if I remember correctly). So it's not hopeless! These candidates I'm sure have other assets that make up for their scores.

 

[stephew]

hang on there. think there might be a selection bias with regard to PhD's who look towards radonc? and yes, there are plenty of people with below median usmle scores- and I mean of all usmle takers, not just radonc applicants- who get spots. there are other things in the application but dont count yourself out just because you haven't reached the 230 mark.

 

[achybrachy]

Looking at the charts again I was struck by the difference between charts 12 and 13; nearly all applicants, matched or unmatched, reported some research (98% for matched vs. 96% for unmatched), so you can't really say that simply reporting research helped anyone per se, it's pretty much just expected that you'll have it. On the other hand, 93% of those matching vs. 75% of those not matching had some sort of publication, abstract, or poster; these people did have an advantage on match day. It just goes to show again that it's worth it to write up that little summer project you did, even if it's just a poster or an abstract. As they say, if you didn't publish it, it didn't happen!