Secondary Logo

Journal Logo

Medical Education: Original Clinical Research Report

Do You Really Mean It? Assessing the Strength, Frequency, and Reliability of Applicant Commitment Statements During the Anesthesiology Residency Match

Moran, Kenneth R. MD, FASA*; Schell, Randall M. MD, MACM; Smith, Kathleen A. MD, FASA; Abdel-Rasoul, Mahmoud MS, MPH§; Lekowski, Robert W. Jr MD, MPH; Rankin, Demicha D. MD*; DiLorenzo, Amy MA; McEvoy, Matthew D. MD

Author Information
doi: 10.1213/ANE.0000000000004136

Abstract

KEY POINTS

  • Question: During the residency interview process, how frequently do ranked-to-match candidates who provide statements of strong interest, high rank, and guaranteed commitment actually end up matching to the program that received the statement?
  • Findings: Candidates who provided guaranteed commitment statements had a 95.7% match rate to the programs to which the statements were directed, in comparison to statements of high rank (25.6%), strong interest (14.6%), and a 5.9% match rate for those who provided no statement.
  • Meaning: While the ethical use of postinterview communication is debated, the strength of statements of commitment from ranked-to-match candidates was associated with an increased frequency of a successful match, particularly when candidates commit to rank a program #1.

Despite the critical nature of the residency interview process, few metrics have been shown to adequately assess the potential for an applicant’s success in a given program. While desired characteristics vary between specialties, US medical school graduation, high United States Medical Licensing Examination scores, and strong letters of recommendation are associated with a successful match.1–7 Communication during the interview process also plays an influential role and may be a predictor of clinical performance.8,9

The National Resident Match Program has outlined ethical communication and warns against untruthful statements.10 The 2014 Match Participation Agreement states “Both applicants and programs may express their interest in each other; however, they shall not solicit verbal or written statements implying a commitment.”11

While evaluating whether a candidate is a good match, bias can occur when applicants make commitment statements to a program.12–15 Frishman et al12 found that nearly one-third of 132 program directors rank a candidate more favorably when they express a high degree of interest. Fellow residents and faculty advisors commonly recommend that students notify their top choice of their commitment.16,17 Nagarkar and Janis18 found that half of 1179 surveyed residents admitted to exaggerating interest in statements made during the match. Many recognize the negative impact of this lapse of professionalism and have called for rules prohibiting communication outside of the match.10,13,14,19

We do not fully understand how an applicant’s perceived commitment to a program influences how they are ranked as a candidate.12,20 It is difficult to validate whether a sentiment is genuine or a deceptive strategy to increase the odds of matching. Unfortunately, residents may perceive pressure from programs to demonstrate commitment and, in response, provide misleading statements.15,16,18,21,22 The purpose of this study is to evaluate the frequency of various commitment statements from applicants to 5 anesthesiology departments during a single interview season, to report how often these statements are associated with a successful match to the recipient program, and to identify how frequently candidates provide a misleading statement indicating that they will rank a program #1.

METHODS

This prospective, multi-institutional study was conducted on data collected during the 2014 National Resident Match Program match for all applicants to 5 anesthesiology residency programs: The Ohio State University Wexner Medical Center, Vanderbilt University Medical Center, University of Kentucky College of Medicine, University of North Carolina, and Brigham and Women’s Hospital. The study was approved or exempted by each site’s institutional review board. Written consent for participation was not required because all data were deidentified and collected as part of the routine interview process and knowledge of the study may have affected the nature of the statements provided by participants.

Participating programs collected written and verbal communications from applicants made to members of their interview committee during the interview process up until the February 26, 2014 deadline for National Resident Match Program rank list submission. All statements were collected including handwritten letters, thank you cards, emails, or direct conversations. Recorded statements were unsolicited by the programs, and verbal statements were recorded verbatim to capture the candidates wording as exactly as possible. Each site created a rank list of candidates for submission to the National Resident Match Program according to their usual selection process. Biographical information submitted by applicants as part of the typical interview process was associated with each candidate and their commitment statements. The applicant’s age, sex, United States Medical Licensing Examination step 1 and 2 scores, couples match status, and whether or not they matched at that program were all associated with each candidate. The program then ordered each applicant and their associated data according to the rank list order submitted to the National Resident Match Program. Before submitting each institution’s data to the primary study site, the data were deidentified and a unique identification number was assigned to each applicant.

Commitment statements were temporarily disassociated from biographical data except for the associated identification number and were randomly sorted. Three residency program directors (K.R.M., R.M.S., and M.D.M.), from 3 institutions in this study, independently reviewed the statements to classify them as guaranteed commitment, high rank commitment, or strong interest. In cases where there was disagreement, but 2 of the 3 reviewers selected the same designation, the majority choice was assigned. Weighted κ coefficients were then used to measure interrater reliability beyond chance agreement between evaluators. Guaranteed commitment statements were defined as statements that definitively demonstrate that a candidate will be ranking a program #1 on their rank list. High rank commitment statements imply that a program will be ranked at or near the top of their rank list but do not guarantee the #1 position. Strong interest statements were defined as expressing great interest without making any reference to favorable ranking (Supplemental Digital Content 1, Table 1, http://links.lww.com/AA/C779). Communications that demonstrated gratitude without expressing interest in matching or attending the program were not included in analysis.

After the match, each institution was required to indicate the rank list position for the last candidate to match on their rank list. All candidates ranked at or above this position were considered ranked-to-match for the purpose of the study. Candidates not ranked-to-match were excluded from any analysis that compared commitment statements with the odds of a successful match because these candidates had no chance of matching. For 3 programs in the study with both advanced and categorical rank lists, ranked-to-match was defined as the last ranked-to-match position for the position type (categorical or advanced) that fell the shortest distance down their rank list. Candidates below this described cutoff were not included in data analysis of match frequency because applicants typically do not specify which position type they are referring to when making a commitment statement but will often rank both (Supplemental Digital Content 2, Appendix A, http://links.lww.com/AA/C778).

Couples match candidates and “R” position candidates were also excluded from analysis of match frequency. A successful match can be confounded by a partner with a poor ranking despite an honest commitment statement. Typically, R position candidates interview for advanced and categorical positions in addition to available R positions and can be ranked on all 3 rank lists for the same program. Unfortunately, because virtually all commitment statements did not specify which position they were committing to, R positions were not included in this portion of the analysis.

The collected data set included the associated rank list position, biographical information, match status at that institution, and any offered commitment statements. While deidentified before submission, the data from each candidate encounter remained associated with a rank list number and an indicator of match success.

For the estimation of interrater agreement of commitment statement level, all candidates were included in the analysis because the influence of external variables (couples match, R position, rank list position) on not matching was not considered relevant to differentiating among strong interest, high rank, and guaranteed commitment statements. Cohen weighted κ statistics and their respective 95% CIs were estimated to assess pairwise interrater agreement beyond chance among the 3 raters, respectively, on scoring the strength of commitment statements. A weight of 1 was assigned for perfect agreement, a weight of 0.5 was assigned for partial agreement, and a weight of 0 was assigned for no agreement when estimating Cohen weighted κ statistic.

Pearson χ2 or Fisher exact tests (where relevant) were used to test the association of candidate characteristics and the dichotomous outcome of successfully matching and to compare candidate characteristics between commitment statement groups. Normally distributed continuous variables were compared between commitment statement groups using ANOVA. Linear trends between ordinal categorical variables were tested using Cochran–Armitage χ2 tests. A multivariable logistic regression model was fit including an ordinal “strength of statement” variable, age, sex, United States Medical Licensing Examination 1 score, United States Medical Licensing Examination 2 score, and 2 interaction terms between strength of statement and each of the United States Medical Licensing Examination scores as potential factors associated with the outcome of a successful match. The guarantee statement and high rank statement groups were combined in the model due to sparse data in the guarantee statement group that did not successfully match. The final model included each of the main effects listed above, regardless of significance, and any interaction terms that were statistically significant (P < .05). Summary statistics for categorical variables were reported as frequencies and percentages. P values <.05 were considered to be statistically significant. All statistical analyses were conducted using SAS 9.4 (SAS Institute, Cary, NC).

Our post hoc power analysis estimates that we had >80% power to detect effect sizes of 0.14 and 0.15 with our sample size of 410 for 2 × 2 (1 degree of freedom χ2 test) and 2 × 3 (2 degrees of freedom χ2 test) contingency tables, respectively, at an α of .05 (PASS 12; NCSS, LLC, Kaysville, UT).

Protocol regarding how interviewers respond to candidate commitment statements varied between programs. When asked post hoc, 3 programs indicated that they rarely respond to commitment statements by telling candidates that they would be ranked favorably, while 2 programs never reciprocate with similar statements. Two of the programs sometimes increase a candidate’s rank list position after receiving a commitment statement, 2 programs rarely do so, and 1 of the programs never alters their rank list.

RESULTS

Interrater Agreement on the Assignment of the “Level” of Commitment for Statements

Five institutions submitted rank lists totaling 632 positions. Of those applicants, 226 (35.8%) made 1 of the 3 types of commitment statements. Interrater agreement on the level of each commitment statement was high across the raters. The weighted κ coefficient for reviewers A and B was 0.9 (95% CI, 0.9–1.0), and between A and C was 0.9 (95% CI, 0.8–0.9). The κ coefficient of 0.9 (95% CI, 0.9–1.0) between C and B was the highest measure of agreement beyond chance among the rater combinations.

Table 1.
Table 1.:
Commitment Level Assignments and Frequency of Interrater Agreement

All 3 raters agreed on 199 (88.1%) of the 226 statements. Of the 27 statements that did not have uniform agreement, all 27 were assigned the same designation by 2 of the 3 evaluators. Instances of disagreement never differed by >1 level of commitment. For example, none of the statements were assigned as both guaranteed rank and strong interest simultaneously and none of the statements were given different assignments by all 3 evaluators. Evaluators were approximately 3 times more likely to disagree between strong interest and high rank statements than between high rank and guaranteed statements. For guaranteed commitment statements, all 3 evaluators agreed in 33 of 34 instances (97.1%) (Table 1).

Frequency of Commitment Statements Among All Applicants

Overall, 226 (35.8%) of the 632 applicants provided a statement that demonstrated at least strong interest. Specifically, 124 (19.6%) gave strong interest statements, 68 (10.8%) provided high rank statements, and 34 (5.4%) gave guaranteed commitment statements.

Commitment Statements and Their Association With Successfully Matching to the Recipient Program

After eliminating couples match, R position, and non–ranked-to-match candidates from the 632 rank list positions, 410 study candidates remained. Of these 410 ranked-to-match candidates, 255 (62.2%) did not provide a statement, 89 (21.7%) provided statements of strong interest, 43 (10.5%) high rank, and 23 (5.6%) made statements of guaranteed commitment.

Table 2.
Table 2.:
Match Frequency Associated With Guaranteed, High Rank, and Strong Interest Statementsa
Figure.
Figure.:
Elimination of non–ranked-to-match, couples, and “R” position candidates to determine the frequency that guaranteed commitment statement candidates failed to match due to misrepresentation. Only 1 of the 23 did not match, suggesting that they misrepresented their statement of commitment or changed their mind.

For the 255 interview encounters that did not offer a statement of any kind, only 5.9% (15/255) matched. In contrast, 29.7% (46/155) of candidates who gave any type of statement matched (P < .001). For each increasing level of commitment statement given, the frequency of successfully matching to the program receiving the statement also progressively increased in a linear trend (Cochran–Armitage χ2 test; P < .001). There was a 14.6% (13/89) successful match rate for strong interest statements, 25.6% (11/43) for high rank statements, and 95.7% (22/23) for guaranteed commitment statements (P < .001; Table 2). For guaranteed commitment statements, it can be concluded that 1 of the 23 candidates (4.3%) incorrectly represented themselves when they stated that they would be ranking the program #1 (Figure).

Association of Biographical Characteristics With Commitment Statements

Table 3.
Table 3.:
Association Between Biographical Factors and Providing Commitment Statements

There were no significant differences in the frequency of commitment statements among age, sex, couples match status, R position status, or rank list position above or below the “ranked-to-match” cutoff groups (P > .05, all comparisons). There were overall differences in mean United States Medical Licensing Examination 1 and mean United States Medical Licensing Examination 2 scores between strength of commitment statement groups (P = .044 and P = .003, respectively; Table 3).

Multivariable Logistic Regression Model Assessing the Association of Biographical Characteristics With Match Outcome Among Those Ranked High Enough to Match

A multivariable logistic regression model predicting the dichotomous “matched” outcome (yes/no) was fit using data from 374 participants (36 missing data). The model included the main effects of sex, and strength of statement as categorical predictor variables, along with age, United States Medical Licensing Examination 1, and United States Medical Licensing Examination 2 as continuous predictor variables, and an interaction term between strength of statement and United States Medical Licensing Examination 2 score (Table 4). Strength of statement was included as an ordinal variable with 3 levels (1: no statement, 2: strong interest, 3: high rank/guarantee). The high rank and guarantee categories were combined due to sparse numbers in the guarantee group leading to almost complete separation. The model had an area under the curve of 0.82 (95% CI, 0.8–0.9) and a Hosmer–Lemeshow goodness-of-fit P value of .895. Age, sex, and United States Medical Licensing Examination 1 score were not significantly associated with a successful match (P > .05).

Table 4.
Table 4.:
Multivariable Logistic Regression Model of Successfully Matching

There was a significant interaction effect between strength of statement and United States Medical Licensing Examination 2 scores (P = .048). Participants with United States Medical Licensing Examination 2 scores of 239, for example, who also provided high rank/guarantee statements had 3.8 (95% CI, 2.4–9.2) times greater odds of matching than participants with the same United States Medical Licensing Examination 2 score providing a strong interest statement (P = .003) and had 31.4 (95% CI, 9.9–99.3) times greater odds of matching than participants with the same United States Medical Licensing Examination 2 score who did not provide any interest statement (P < .001). Participants with United States Medical Licensing Examination 2 scores of 239 who provided a strong interest statement had 8.3 (95% CI, 2.6–26.9) times greater odds of matching than participants with the same United States Medical Licensing Examination 2 score who did not provide any interest statement (P < .001).

Participants with United States Medical Licensing Examination 2 scores of 260, as another example, who provided a high rank/guarantee statement had 11.2 (95% CI, 3.7–34.5) times greater odds of matching than participants with the same United States Medical Licensing Examination 2 score providing a strong interest statement (P < .001) and had 19.4 (95% CI, 7.8–48.2) times greater odds of matching than participants with the same United States Medical Licensing Examination 2 score who did not provide any interest statement (P < .001). Participants with United States Medical Licensing Examination 2 scores of 260 who provided a strong interest statement had 1.7 (95% CI, 0.6–5.1) times greater odds of matching than participants with the same United States Medical Licensing Examination 2 score who did not provide any interest statement, but this difference was not statistically significant (P = .327).

Couples Match, R Position Candidates, and Candidates Not Ranked-to-Match on Both Advanced and Categorical Rank Lists

These groups were analyzed separately to demonstrate their odds of not matching despite a guaranteed commitment statement. For couples match candidates who gave guaranteed commitment statements, 3 of 4 (75%) did not match, and 1 of 2 candidates (50%) who were not ranked-to-match on both advanced and categorical rank lists did not match.

Table 5.
Table 5.:
Match Failure Rates for Ranked-to-Match Applicants Who Provided Guaranteed Commitment Statementsa

The rate of failed guaranteed commitment statement matches for couples match positions (75%) and for candidates not ranked-to-match on both advanced and categorical rank lists (50%) was significantly higher than the rate of failed match for the 23 other candidates who offered guaranteed statements (4.3%, 1/23; P = .003). None of the R positions offered guaranteed statements (Table 5).

DISCUSSION

The findings of this study can help to better understand commitment statements from applicants and how they might affect the match process. It found that program directors consistently differentiated between different types of commitment statements, that the greater the level of commitment the more frequently a successful match occurred, and that guaranteed commitment statements were very frequently associated with a successful match and rarely misrepresented.

To interpret the reliability of a commitment statement, interview committees must be able to differentiate between levels of commitment. Guaranteed commitment statements were recognized by all 3 program directors 97.1% of the time. For all types of statements, there was 88.1% unanimous agreement demonstrating that the assignment of lower levels of commitment was also reliable.

While commitment statements are quite common, those that guarantee a #1 ranking occurred relatively infrequently but were associated with a high rate (95.7%) of a successful match. Even high rank and strong interest statements showed an increased frequency of matching, 25.6% and 14.6%, respectively, when compared to those who did not offer any statements (5.9%). Based on these data, commitment statements did increase the odds of a successful match and the stronger the statement of commitment, the higher the frequency of a successful match.

Two independent studies assessed postinterview communications.16,17 They provide self-reported data about commitment statements but did not study corresponding match rates. Applicants (59.9%) in 1 study and 36.5% in the other reported telling >1 program they would be ranked highly. This evidence is consistent with our finding that high rank statements only resulted in a 25.6% match rate (11/43) to that program. In 1 study, 1.1% of survey responders reported telling multiple programs that each will be ranked #1. In comparison, we found that 4.3% (1/23) of guaranteed commitment statements were misrepresentations and did not result in a successful match.

In contrast, a study by Swan and Baudendistel20 suggested that 31% of ranked-to-match applicants who offered guaranteed commitment statements lied about ranking their program #1. However, their study was designed to demonstrate whether commitment statements affect rank list position, not to identify misrepresentation. The high rate of perceived misrepresentation may have resulted from not removing variables such as couples match, R position candidates, or applicants not ranked-to-match for both preliminary and categorical programs. By eliminating these variables, our study demonstrates a much lower rate of misrepresentation. This is also supported by the low match rate for couples match applicants (25%, 1 of 4) and candidates not ranked-to-match on both advanced and categorical lists (50%, 1 of 2) despite guaranteed commitment statements. When compared to the 95.7% match rate for the remaining candidates who offered guarantee statements, it is clear that these variables resulted in a much lower match rate despite promises to rank #1. The assumption that ranked-to-match candidates, who rank a program #1, will always match is supported by the National Resident Match Program match algorithm description that explains that the system always favors the interests of the candidate.23

Age, sex, couples match, R position, and ranked-to-match status did not affect the frequency of commitment statements. While mean United States Medical Licensing Examination 1 and 2 scores were statistically different overall between strength of statement groups, the difference in mean scores was minimal for all practical purposes. For example, mean United States Medical Licensing Examination 1 and 2 scores for candidates offering guaranteed commitment statements compared to candidates who offered no statement were 233.7 vs 237.0 for United States Medical Licensing Examination 1 and 248.4 vs 248.0 for United States Medical Licensing Examination 2, respectively.

Study Limitations

Non–couples match, non–R position, ranked-to-match candidates who provide a guaranteed commitment statement but do not match likely demonstrates dishonesty because variables that could contribute to a failed match have been removed. However, the assumption that those who did match were “honest” does not take into account a dishonest candidate who actually ranked another program #1, but failed to match there. In addition, unmatched ranked-to-match candidates may have made a guaranteed commitment statement and subsequently changed their mind without intending to be dishonest. In this context, it is important to understand that this study does not evaluate the subjective topic of “honesty,” but instead associates each statement with the frequency of a successful match.

The data collected represent individual encounters but do not imply that statements all came from different people. It is likely that candidates made statements to >1 institution. However, understanding the correlation of a statement to a successful match regardless of how many times a candidate has given the statements is still relevant.

In addition, programs were instructed to record, verbatim, the most committed statement given by the candidate in any form of communication including written, electronic, or verbal. They were not asked to report which form of communication was used. It is conceivable that verbal communications would be more difficult to recall verbatim and may be more subject to bias than occurs with written statements.

Future Directions

Although commitment statements are common, little has been done to adequately demonstrate whether program directors are biased by postinterview communication. While some medical specialties have forbidden postinterview communication with applicants, others welcome it. This provides an opportunity to compare how frequently highly committed applicants match despite not expressing commitment in comparison to similar applicants who are allowed to communicate. Longitudinal studies demonstrating whether highly committed applicants are successful in a program and whether they maintain a positive opinion of the program would be useful.

CONCLUSIONS

The match is a complex process, and participants are motivated to maximize the chances of a successful match. This environment often produces commitment statements that have the potential to bias a program’s rank list and may not accurately represent the candidate’s submitted National Resident Match Program rank list. The results of this study confirmed that program directors reliably differentiated between levels of commitment statements. In general, positive interview or postinterview communication of any type did increase the frequency of a successful match at that institution with higher match rates associated with stronger commitment statements. While statements that an applicant will rank a program #1 are much less frequent, they were very reliable and resulted in a high frequency of match when ranked-to-match by the program. Guaranteed commitment statements were associated with very high odds that they will match if they are not a couples match or R position candidate and are ranked-to-match on both advanced and categorical rank lists when applicable. Programs can benefit from the understanding that despite providing guaranteed commitment statements, couples match candidates are affected by the ability of their “couple” to match and those not ranked-to-match on all lists may inadvertently also mislead a program if they do not specify ranking one or both of the advanced and categorical or R programs as first choices specifically.

We do not fully understand how an applicant’s perceived commitment to a program influences how they are ranked as a candidate. It is difficult to validate whether the sentiment is genuine or a deceptive strategy to increase the odds of matching, particularly when candidates perceive pressure from programs to demonstrate commitment. While the ethics of postinterview communication is debated, the results from this study may assist anesthesiology programs as they debate whether to join other fields who have already created rules that forbid postinterview communication.24,25 It is the hope of the authors that this study provides useful data to the discussion of how these interactions affect the interview process and assists future research and policy to improve how we accurately and ethically evaluate candidates during the interview process.

DISCLOSURES

Name: Kenneth R. Moran, MD, FASA.

Contribution: This author helped design and conduct the study; collect and analyze the data; and write, edit, and approve the final manuscript.

Name: Randall M. Schell, MD, MACM.

Contribution: This author helped design and conduct the study; collect and analyze the data; and write, edit, and approve the final manuscript.

Name: Kathleen A. Smith, MD, FASA.

Contribution: This author helped conduct the study; collect the data; and critically review, edit, and approve the final manuscript.

Name: Mahmoud Abdel-Rasoul, MS, MPH.

Contribution: This author helped provide statistical analysis for the study and critically review, edit, and approve the final manuscript.

Name: Robert W. Lekowski Jr, MD, MPH.

Contribution: This author helped conduct the study; collect the data; and critically review, edit, and approve the final manuscript.

Name: Demicha D. Rankin, MD.

Contribution: This author helped conduct the study; collect the data; and critically review, edit, and approve the final manuscript.

Name: Amy DiLorenzo, MA.

Contribution: This author helped conduct the study; collect the data; and critically review, edit, and approve the final manuscript.

Name: Matthew D. McEvoy, MD.

Contribution: This author helped design and conduct the study; collect and analyze the data; and write, edit, and approve the final manuscript.

This manuscript was handled by: Edward C. Nemergut, MD.

REFERENCES

1. de Oliveira GS Jr, Akikwala T, Kendall MC, et al. Factors affecting admission to anesthesiology residency in the United States: choosing the future of our specialty. Anesthesiology. 2012;117:243–251.
2. Fraser JD, Aguayo P, St Peter S, et al. Analysis of the pediatric surgery match: factors predicting outcome. Pediatr Surg Int. 2011;27:1239–1244.
3. McCaskill QE, Kirk JJ, Barata DM, Wludyka PS, Zenni EA, Chiu TT. USMLE step 1 scores as a significant predictor of future board passage in pediatrics. Ambul Pediatr. 2007;7:192–195.
4. McClintock JC, Gravlee GP. Predicting success on the certification examinations of the American Board of Anesthesiology. Anesthesiology. 2010;112:212–219.
5. Perez JA Jr, Greer S. Correlation of United States medical licensing examination and internal medicine in-training examination performance. Adv Health Sci Educ Theory Pract. 2009;14:753–758.
6. Stratman EJ, Ness RM. Factors associated with successful matching to dermatology residency programs by reapplicants and other applicants who previously graduated from medical school. Arch Dermatol. 2011;147:196–202.
7. Thundiyil JG, Modica RF, Silvestri S, Papa L. Do United States Medical Licensing Examination (USMLE) scores predict in-training test performance for emergency medicine residents? J Emerg Med. 2010;38:65–69.
8. Brothers TE, Wetherholt S. Importance of the faculty interview during the resident application process. J Surg Educ. 2007;64:378–385.
9. Swanson WS, Harris MC, Master C, Gallagher PR, Mauro AE, Ludwig S. The impact of the interview in pediatric residency selection. Ambul Pediatr. 2005;5:216–220.
10. Carek PJ. Postinterview communications between residency programs and candidates: what are the rules? J Grad Med Educ. 2012;4:263–264.
11. National Residency Match Program (NRMP). Terms and Conditions of the Match Participation Agreement Among Applicants, the NRMP, and Participating Programs. Match Participation Agreement For Applicants and Programs For the 2014 Main Residency Match. 2014. Available at: http://www.nrmp.org/policies/match-participation-agreements-and-policies/. Accessed January, 24, 2017.
12. Frishman GN, Matteson KA, Bienstock JL, et al. Postinterview communication with residency applicants: a call for clarity! Am J Obstet Gynecol. 2014;211:344.e1–350.e1.
13. List JM. Honesty and fairness in the residency match. The virtual mentor: VM. 2012;14:1011–1015.
14. Wu AJ, Vapiwala N, Chmura SJ, et al. Taking “the Game” out of the match: a simple proposal. Int J Radiat Oncol Biol Phys. 2015;93:945–948.
15. Holliday EB, Thomas CR Jr, Kusano AS. Integrity of the national resident matching program for radiation oncology: national survey of applicant experiences. Int J Radiat Oncol Biol Phys. 2015;92:525–531.
16. Jena AB, Arora VM, Hauer KE, et al. The prevalence and nature of postinterview communications between residency programs and applicants during the match. Acad Med. 2012;87:1434–1442.
17. Ratcliffe TA, Durning SJ, Jena AB, et al. Postinterview communication between military residency applicants and training programs. Mil Med. 2012;177:54–60.
18. Nagarkar PA, Janis JE. Fixing the match: a survey of resident behaviors. Plast Reconstr Surg. 2013;132:711–719.
19. Opel D, Shugerman R, McPhillips H, Swanson WS, Archibald S, Diekema D. Professionalism and the match: a pediatric residency program’s postinterview no-call policy and its impact on applicants. Pediatrics. 2007;120:e826–e831.
20. Swan EC, Baudendistel TE. Relationship between postinterview correspondence from residency program applicants and subsequent applicant match outcomes. J Grad Med Educ. 2014;6:478–483.
21. Miller JB, Schaad DC, Crittenden RA, Oriol NE, MacLaren C. Communication between programs and applicants during residency selection: effects of the match on medical students' professional development. Acad Med. 2003;78:403–411.
22. Sbicca JA, Gorell ES, Peng DH, Lane AT. A follow-up survey of the integrity of the dermatology National Resident Matching Program. J Am Acad Dermatol. 2012;67:429–435.
23. National Residency Match Program (NRMP). NRMP Match Algorithm. 2017. Available at: http://www.nrmp.org/matching-algorithm/. Accessed June 16, 2017.
24. Fisher CE. A piece of my mind. Manipulation and the match. JAMA. 2009;302:1266–1267.
25. Nagarkar PA, Janis JE. Fixing the “match”: how to play the game. J Grad Med Educ. 2012;4:142–147.

Supplemental Digital Content

Copyright © 2019 International Anesthesia Research Society