Last month, my friend Nikhil Kansal detailed in Vascular Specialist why the Association for Program Directors in Vascular Surgery (APDVS) elected to implement Preference Signaling in the upcoming Integrated Vascular Surgery Match. Here I will explain why I think this decision was well intended, inevitable and wrong. Let’s break it down FAQ style.
What is Preference Signaling?
Preference signaling is a relatively new feature in the residency Match that lets applicants explicitly show interest in specific programs during the ERAS application process. Applicants get a limited number of “signals” they can send to the programs they are most interested in. Some specialties use a tiered system (gold, silver) to indicate levels of interest. Prior to sending interview invites, programs can see if an applicant signaled them, but cannot see where else that applicant sent signals.
Do all specialties participate in signaling?
No. Otolaryngology was an early adopter in 2021. Since then, it has become increasingly popular with over 25 specialties opting in for the 26-27 cycle.
What problems does signaling address?
- Application inflation: In some specialties programs routinely receive over 1,000 applications per cycle. Holistic review becomes impossible and arbitrary numeric cutoffs are used to screen applicants (Step 2 score, class rank, AOA status, number of publications). Signaling creates a scarcity mechanism to combat this inflation.
- Interview hoarding: Strong applicants accumulate disproportionate numbers of interviews while weaker applicants struggle to get enough. This unequal distribution of opportunities can lead to a downstream inefficiency in ranking/matching. This would manifest by programs going unfilled in the Match despite interviewing an appropriate number of applicants. Let’s say there were 200 applicants for 100 positions. If every program interviewed the same top 30 applicants there could be 70 unfilled positions and 170 unmatched applicants.
- Randomness and inefficiency in the Match: Without signaling, Matches are influenced by timing, luck in getting interviews and arbitrary screening cutoffs. This can lead to less optimal applicant–program alignment which would manifest through high attrition numbers. (i.e. applicants are matching to programs where they are not happy).
Which of these problems is present in the current integrated vascular match?
None. But in life, the easiest problems to fix are ones that don’t exist.
In 2025, the average vascular program received 140 applications. A manageable number for holistic review. If interview hoarding occurs, it is having no effect on the efficiency of the Match — 487 of 489 positions have filled over the past five years. Finally, the current Match process appears highly successful in generating applicant-program concordance. The cumulative attrition rate in vascular residencies is only 4.1% over the past decade.
Wait, what? Then why are we doing this?
Insecurity. The Match is a near perfect system. The algorithm behind it won the Nobel Prize. When all parties follow the rules, the system cannot be gamed. To play fair, however, requires a leap of faith. After the interview cycle, program directors must submit a rank list of applicants in the order of their preference without knowing how the applicants ranked them. To many of us this is untenable. What if I match someone and I was their tenth choice? The horror! We need to be wanted. Signaling is reminiscent of a middle school strategy to find a date for the dance. We slip the applicants a note — “Do you like me? Check Yes or No.” (There were no DMs to slide into back then kids).
Shouldn’t programs want to interview applicants who are genuinely interested in their program?
Studies on the effects of signaling have consistently shown one thing — when introduced, signaling becomes the predominant driver of interview offers. Interviews are obviously the main determinant of rank order lists and therefore the ultimate match. The signal would not be a big deal if program directors were mature enough to ignore it. But, unless vascular surgery is unique among all specialties, they are not.
So, we have placed the most important decision in this process in the hands of the people who know little, at a point in time in which they know the least. What are they signaling based on? Mentor advice? Reputation? Program website? Reddit?? We are anchoring match outcomes to preferences formed at the least informed point in the process — before interviews, before fit assessment and often based on incomplete or secondhand impressions.
What’s the downside? Signaling seems harmless.
Prior to COVID, the integrated vascular matches were highly successful. Usually only one or two positions went unfilled. In my role at the APDVS, I would often reach out to these unmatched programs to debrief. One common theme stood out — the creation of artificial scarcity. Despite receiving enough applications to comfortably fill their spots, the program director had either interviewed or ranked too few applicants. An unforced error. Then, with COVID and the institution of virtual interviews, a peculiar thing happened. All vascular positions filled. For the first time ever, the match was now at 100% efficiency for both integrated residencies and fellowships. Virtual interviews had eliminated the key drivers of scarcity — applicant finances, applicant time and program resources.
Despite the gradual reintroduction of in-person interviews and the annual increase in positions offered, the integrated vascular match has remained near perfect including a 100% fill rate in 2025. By introducing signaling we risk shrinking interview pools — reintroducing the same artificial scarcity that historically led to unfilled positions. We spent years trying to eliminate scarcity. It would be an impressive achievement to bring it back voluntarily.
How will the applicants really use signaling?
Now that we have placed the applicants in a high-stakes game of strategic resource allocation they will naturally develop their own tactics to achieve their desired outcome — a maximum number of interview invitations. Ideally, programs and applicants would engage in behavior that is mutually beneficial. Program directors should have the maturity to use signals as one small part of the holistic review process. Applicants should truly signal the programs they are most interested in and not use them to amass the highest possible number of interviews. Neither approach is enforceable. In game theory this is known as the Prisoner’s Dilemma. As a general rule of thumb, I try to avoid situations where the closest comp is incarceration.
In the end, it is hard to blame the applicants if they approach signaling with strategy rather than authenticity. The consequences are asymmetric. A poor signaling plan may cost an applicant interviews they would have otherwise received. Meanwhile, programs face minimum downside.
Why was signaling inevitable?
Program directors are running out of ways to evaluate student applications. Step 1 scores are hidden. Many schools are now pass/fail. Dean’s Letters obfuscate through implausible Likert scales — “Here at the Medical School of the Everglades we rank our students in five tiers from Exceptional to Outstanding.” Even once reliable letters of recommendation now often read like someone trying to send coordinates to a secret location through a random series of numbers (I have known Marie for 29 weeks. She is among the top 13.4% of medical students I have worked with).
In the absence of other consistent metrics, program directors will naturally rely on signals for their interview invites. But if gauging an applicant’s interest pre-interview is deemed necessary, there are more reliable options for individual programs to employ. In person interviews and secondary applications both require extra time and effort from the applicant. Some programs even schedule their interview dates late in the cycle when many are tired of traveling and spending (Buyer beware: this tactic led to an unmatched program in at least one instance).
As signaling is implemented, it should be treated as an experiment — not a solution. Track the outcomes that matter; diversity, geographic distribution, attrition and, of course, Match efficiency. Manufacturing scarcity for every program through signaling is likely folly. There is absolutely nothing to gain. We will soon discover what can be lost.
