The Conundrum of Divergent Signaling by Couples Applicants

Program directors (PDs) have experienced an explosion in the number of residency applications, making it difficult to determine who to interview. The Association of American Medical College’s (AAMC’s) Electronic Residency Application System (ERAS) allows applicants to send a limited number of “signals” to programs through a supplemental application. The intention of signals was to help programs “identify applicants who are genuinely interested,”¹ and PDs reported a high rate of using signals in deciding who to interview.² Signaling by couples, however, has created a conundrum. How do 2 PDs in different specialties at the same institution interpret intentions when one applicant in a couple sends a signal to one PD’s program but the second does not send a corresponding signal to the other, particularly if the couple then attempts to leverage the first member’s signal to obtain an interview for the second?

There are 2 major contributors to this situation. One is the uneven number of signals between specialties, ranging from 2 to 30, with the most frequent being 5.³ Members of a couple may have a different number of signals, and the partner with a lower number may have difficulty expressing their level of interest to some of the programs their partner signaled. As an example, partner A applies to obstetrics and gynecology (OB/GYN), which allows 15 signals, and partner B applies to anesthesiology, which allowed 5 in 2022. How would an anesthesiology program interpret an email from a candidate who did not send them a signal but is requesting an interview because their partner received an interview with the institution’s OB/GYN program after signaling it? They might wonder, “Are we number 6 on this candidate’s list or 15?” The different number of signals between specialties (and gold versus silver signals in OB/GYN) has potentially created less certainty, rather than more, regarding an applicant’s interest when couples matching. What if an OB/GYN applicant sends an institution a silver signal but their partner applying for anesthesiology sends one of their 5 signals? Should the OB/GYN program consider that silver signal equivalent to a gold signal based on their partner’s signal to anesthesiology?

A second potential contributor may be dyssynchronous signaling by couples to obtain a competitive advantage. Many applicants believe that more interviews mean an increased chance of matching, and early data suggest that signaling increases the chances for obtaining an interview.⁴ Consider again the aforementioned partners A and B applying to OB/GYN and anesthesiology, respectively. Partner A sends a signal to an OB/GYN program and is offered an interview, but partner B does not signal the institution’s anesthesiology program. Partner B requests an interview based on partner A’s interview offer. Partner B may obtain an interview before other applicants who also did not send a signal to the anesthesiology program. In this scenario, the signaling process did not help programs judge interest level and may have conferred an unfair advantage to this couple. Each partner has the freedom to train where they choose, but the intention of couples matching is to be placed together. Signaling should reflect this.

The AAMC’s introduction of signaling is a wonderful adaptation of the application process. As with all process changes, there usually are unanticipated issues to address after initial implementation. There should be transparency in signaling by couples due to the potential scope of the issue, possible advantage conferred to applicants, and potential harm to programs. Per the National Resident Matching Program (NRMP), 5.8% of applicants were part of a couple in 2022.^5,6 This is a large enough group to merit vigilance. Applicants may perceive an advantage to applying as a couple. Per the NRMP, 81.1% of all active applicants (US and non-US medical school graduates) matched to a training position, while 87.8% of couples matched together (all active coupled applicants), and only 6.9% of applicants in a couple did not match to any position (P<.001).^6,7 Unfortunately, the distribution of US to non-US graduates among couples and their respective match rates is not currently reported. The match rate was higher among US graduates, so if there are different proportions of US graduates among couples than among noncouples, then the comparisons of unadjusted rates is limited. Despite this limitation, the comparison may still suggest to applicants that there is an advantage and tempt some to misuse couples matching.

In addition to the potential scope and competitive advantage, there may be asymmetric harm to programs. There is a large deviation in the number of signals programs receive. For some specialties, there is a deviation of nearly 50% of the average number of signals received (university-based anesthesiology programs mean=121, SD=65).⁷ Programs receiving a lower number of signals may be especially susceptible to applicants who can send “extra” signals as half of a couple. These programs may be devoting limited interview spots to candidates with less interest than others who did not have the same potential advantage in signaling but who might be more interested in the program. This can harm, to a greater degree, programs that receive fewer signals.

The integrity of the signaling process is important to programs and couples who are attempting to match together. The AAMC should report degree of concordance of signals of couples. It should also assess the impact on program decision-making, positively or negatively, if an institution received a signal for one half of a couple but not the other in its self-reported questionnaire to PDs. Programs need granular data to assure them of the integrity of couples signaling.

There are some possible interventions if the data suggest couples signaling is being exploited. One possible change might be requiring a degree of concordance of signals between couples (Figure). Defining concordance will be a challenge since it may be by geographical area as opposed to institution. It is understandable if couples don’t want to train in the exact same institution but want to be geographically collocated. The AAMC would have to define concordance. Standardizing the number of signals across disciplines might be another step. This change would involve the AAMC potentially dictating the number of signals to the different specialties and stripping them of some independence to determine how to govern their respective application processes. While each possible change has a potential drawback, the current system of signaling for couples may impart an unfair advantage to applicants and may asymmetrically harm programs. Further, ERAS has no requirement for applicants to declare they are part of a couple, other than self-reporting. It is an advantage that more applicants may try to exploit going forward.

View Figure

• Download as Powerpoint
• Download Figure

In addition to considerations of couples preferences signaling, there are other potential signaling abuses, including applying to more than one specialty. Applicants can now send the maximal number of signals for each specialty, but limiting the overall number of signals or limiting each applicant to one personal statement, which presumably would reflect an applicant’s true preference, may help curb this potential abuse. The impact would be equal on all applicants, including couples.

Couples signaling is different from other potential methods of gaming the application process. The match process should strive to keep couples together and likely must accept some degree of potential for abuse of the signaling process to facilitate this goal. Any restriction may preferentially harm couples trying to remain together. Yet, applicants in the high-stakes match may perceive a favorable gain to consequence balance by falsely declaring themselves couples or sending divergent signals. The AAMC must provide more transparent data on signaling patterns overall and among couples in particular, to improve confidence in this promising approach.