The Gastroenterology Fellowship Match—The First Two Years
Article Outline
After a nearly 10-year period in which the market for gastrointestinal (GI) fellows was decentralized and chaotic, a fellowship Match was organized and conducted in June 2006. After the second year of operation of this new Match, which was conducted in June 2007, we surveyed fellowship directors to assess participation in and satisfaction with the Match. This survey concentrates on the current usage of the Match, and on how it has changed the interviewing and hiring process that developed in the years immediately before the Match as revealed in our 2005 survey of fellowship directors; that is, we compare the market with the Match to the market just before the decision to reinstitute a fellowship Match.1
The previous GI fellowship Match, started in 1986, collapsed in the late 1990s as fewer and fewer programs participated.2 In the following years, interviews and offers of employment became earlier and earlier from year to year, with fellowship programs eventually hiring at dispersed times and very quickly. The result was that both candidates and fellowship directors were faced with very few choices and scheduled interviews were often cancelled when candidates were faced with exploding offers from a program that interviewed them earlier.3, 4 (This “unraveling” in time is common in decentralized medical and other entry-level labor markets, and was part of the history leading up to the creation of the residency Match.5 In the case of GI fellowships, it led to fragmentation of the market into many local markets,6 although with no apparent effect on fellowship stipends.7) The new survey, therefore, asked about levels of current participation in the Match, timing of interviews, and related hiring experience in 2007.
The current survey shows that the Match succeeded in moving the interview process later in applicants' residency, and making it more orderly and predictable. Fellowship directors and applicants do not have to decide whom to hire before assessing many possibilities, and there is overall high satisfaction with the Match.
The survey was distributed in August 2007 via e-mail to 155 nonmilitary Gastroenterology Fellowship Program Directors on the American Gastroenterological Association Institute list, taking them to the following URL, hosted at the Harvard Business School site (https://surveys.hbs.edu/perseus/se.ashx?s=381B5FE506823F00). Two follow-up e-mails were sent in September 2007.
Sixty-nine program directors replied in a way that allowed us to identify them. One was from Puerto Rico, which faces a different market than fellowship programs on the mainland, so our analysis here is of the remaining 68 program directors.
Participation in the Match
There were a total of 123 programs with 150 tracks in the Gastroenterology Fellowship Program Match for positions to start July 2008. There were 622 applicants for 325 positions with 313 positions matching through the NMRP/SMS. This contrasts with the prior year (July 2007 start date) where there were 112 programs with 134 tracks in the NRMP/SMS; 578 applicants for 283 positions and 276 positions matching through the NRMP/SMS. Of the survey respondents, 61 (90%) of programs participated in the Match, advertising a total of 183 positions in the Match, 179 of which (98%) were filled through the Match. There was general satisfaction with the Match. Of the 50 respondents who answered the question “Do you think it was good for the Gastroenterology profession as a whole to reinstitute the Match?” 50 responded yes and only 4 responded no (3 of those 4 negative responses were from programs that did not participate in the Match).
Overall, the 61 programs in the Match hired 192 fellows (these include 13 positions filled outside of the Match by 11 [18%] of the 61 programs, in addition to the 179 positions these programs filled in the Match). So, responding programs in the Match hired just >3 fellows per program. The 7 programs outside of the Match hired only 5 fellows in total. Thus, among the responding programs, those not participating are among the smaller programs.
Overall, there was a shortage of Basic Science research fellows (19 were sought, but only 11 were hired), and Research fellows (25 were sought, only 19 were hired). Of the 19 responding programs that did not wish to hire only Clinical fellows, 6 (32%) used the “reversion” capability of the Match8, 9 to specify that if they failed to hire a research or clinical research fellow, they instead wished to fill the position with a Clinical fellow, and for 4 of these programs the reversion of ≥1one position took effect.
Timing of Interviews
Sixty-one programs gave us their starting date for interviews (including 6 programs that did not participate in the Match.) Of those, 47 gave us exact interview dates; 14 gave us a starting month. Because we argue that the Match made interviews happen later, we assume that a program that started interviewing in January 2007 actually started as early as possible in January, namely January 1. For our data from before 2006, we make the opposite assumption, namely that someone interviewing in January actually started interviewing only on January 31. (This is the most conservative set of assumptions; it makes it harder to find that interviews are now conducted later.)
The graphs in Figure 1 pool all Gastroenterology programs, because those that did not participate in the Match interview at similar times as those that do. It shows the cumulative distribution of start dates for positions beginning in 2003, 2005, and 2006 (none of which used a Match), and 2008 (these latter positions were filled using the June 2007 Match). Notice that interview times moved steadily earlier for 2003 to 2006 positions, and that they have moved back dramatically in the 2007 Match. For example, the point at which 50% of programs had begun interviewing came around December (of 2001) for the 2003 positions, and had moved to October (of 2004) for the positions starting in 2006. In contrast, 50% of the interviews for the positions starting in 2008 had not begun until February (of 2007). Note further that interviews before a Match do not result in exploding offers that have to be answered quickly, so the graph understates how much the hiring process moved later into applicants' residency career.
Figure 1.
Cumulative distribution of GI and Match programs that started interviewing by the time of any given 2-week period. Start 03 and Start 05: Start dates of interviews for GI fellowship positions starting in 2003 and 2005 respectively, from FREIDA (and Niederle and Roth3). Start 06 Survey: The replies from the survey of GI program directors to the question of when they started interviewing for 2006 positions. Start 08 Survey: The answers to the question of when GI program directors started interviewing for the June 2007 fellowship match.
Although we do not show the data, interview dates (in terms of the timing of the first interview) for positions in 2008 (the second year of the reinstitution of the GI Match) are basically identical to interview dates we found previously for internal medicine subspecialties that use the NRMP/SMS (see also Niederle et al1).
The 65 programs that provided information interview about 21 fellows each. Thirty-eight percent of programs indicated that they interviewed more fellows than before, whereas only 17% indicated that they now interviewed fewer.
In terms of the quality of the interview, 41% said that they were of higher quality now, and only 3% (2 of 66 programs) indicated that they are now of worse quality (exact wording of the question: “Was there any difference in the quality of interviews compared to 2 years ago?” With three choices as follows: 1) better quality (more relaxed, informative) interviews; 2) worse interviews; 3) no difference). Program directors made the following comments on this issue:
More relaxed interview on both sides. There was no need to speculate whether or not an offer was going to be made and if it was going to be accepted,
Also, the applicants were more forthcoming with their questions and I felt more honest in trying to find out what our program was all about and if it was the appropriate fit for them, instead of worrying whether or not I was going to make an offer to them.
It is still the case that most programs experience some cancellations: 57 of the 67 programs that answered experienced cancellations, for about 2.5 canceled interviews per program. However, programs are in general not asked to speed up the timing of an interview because of a pending deadline (only 8 of the 67 programs do, about 12%), and in general those requests were basically not accommodated. Note that this is in sharp contrast with the situation before the Match, when 43% of programs sped up offers, not merely interviews (because the candidate had another offer in hand). In the first year of the Match, quite a few programs interviewed a candidate with an outside offer: 19 (of 65, 29%), but only 5 of those were asked to make an early offer, not one of which was accommodated.
Of the 7 programs that did not participate in the Match, only 3 made offers before the Match (for a total of 6 offers). Of the 58 programs in the Match that answered this question, 5 (9%) made offers before the Match (a total of 7 offers). Some of the fellows hired outside of the Match were hired before, and some after, in a variety of circumstances as listed below. There were a variety of reasons given by program directors as to why they hired fellows outside of the Match.
Some of these early hires were genuine early hires: 4 of those were research (clinical or basic science research) fellows, some of whom were hired out of concern they would take other positions, some because of prior involvement in the laboratory. One clinical fellow was also hired early (“One clinical fellow, wanted a sure thing, not the uncertainty of the Match”). Two of the early hires were deals made years before. Four more fellows had unusual circumstances (“The U.S. Air Force funded the training of an individual they wanted trained”; “One fellow will replace 2nd year fellow leaving early due to prior experience”; “To apply for an AASLD liver fellow grant, we had to promise the applicant a spot for gastroenterology training”; and “Internal research fellow short tracked” [short tracking is done after completing 2 years of internal medicine residency, instead of the usual 3 years. Almost always the resident remains at the home institution. This is a legitimate reason to take someone outside of the Match]).
Two more fellows were hired after the Match, and thus not in competition with the Match, one because the position was only approved later, and “We hired one fellow from our residency program who failed to match through NRMP.”
When asked “Do you think that your fellows are glad that Gastroenterology rejoined the Match?” of the programs in the Match that answered the question, 47 of 49 (96%) answered affirmatively and 92% when we include all programs (48 of 52). Similarly, when asked “Do you think it was good for the Gastroenterology profession as a whole to reinstitute the Match?”, 48 of 49 (97%) of programs in the Match answered affirmatively, and 93% when we include programs not in the Match (50 of 54).
In conclusion, the survey results suggest that the Match makes hiring and interviewing later and more orderly. There seems to be general satisfaction with the Match. It will be a good idea to monitor it regularly, so that any problems can be identified early and addressed promptly.
References
- . What will be needed for the new GI fellowship match to succeed?. Gastroenterology. 2006;130:218–224
- . 2005 (The collapse of a medical labor clearinghouse (and why such failures are rare)). American Economic Review. 2005;95:878–889
- . The gastroenterology fellowship match: how it failed, and why it could succeed once again. Gastroenterology. 2004;127:658–666
- . The gastroenterology fellowship market: should there be a match?. American Economic Review Papers & Proceedings. 2005;95:372–375
- . The origins, history, and design of the resident match. JAMA. 2005;289:909–912
- . Unraveling reduces mobility in a labor market: gastroenterology with and without a centralized match. Journal of Political Economy. 2003;111:1342–1352
- . Relationship between wages and presence of a match in medical fellowships. JAMA. 2003;290:1153–1154
- . The redesign of the matching market for American physicians: some engineering aspects of economic design. American Economic Review. 1999;89:748–779
- . Competitive wages in a match with ordered contracts. American Economic Review. 2007;97:1957–1969
PII: S0016-5085(08)01197-9
doi:10.1053/j.gastro.2008.06.075
© 2008 AGA Institute. Published by Elsevier Inc. All rights reserved.
