Eliminating Geographic Bias Improves Match Results: An Analysis of Program Preferences and Their Impact on Rank Lists and Results.

BACKGROUND: Previous studies have demonstrated that programs emphasize United States Medical Licensing Examination scores, publications, and geography in creating rank lists. The authors aimed to quantify the importance of geography and to determine how eliminating geographic preferences would affect Match outcomes. METHODS: The Match algorithm was implemented and validated on 6 years of deidentified data from the San Francisco Match (2009 to 2014). A "consensus" ranking was generated for each year-all applicants were ordered into a single list using Markov chain rank aggregation. Each program's rank list was reordered using the consensus list, and a new Match result was simulated. Statistical analysis was carried out with Microsoft Excel. RESULTS: Variation of program rank lists from the consensus rank list was driven by geography (training in the same medical center or state as the ranking program), "pedigree" (top 25 ranking of applicants' prior training), and foreign medical graduation status. Step 1 scores, publications, and medical school or residency region were not factors. The simulated Match resulted in a slight increase in the match rate. The median normalized number needed to match decreased from 6.7 to 6.5, and 80 percent of applicants had an unchanged or better result compared to the actual Match. CONCLUSIONS: Geography is the primary driver of variation between program rank lists. Removing this variation would result in fewer unfilled positions, no significant change in the average number needed to match, and improved Match outcomes for most applicants. Programs should critically evaluate whether their geographic biases reflect underlying information about applicant quality.

Defining Temporal Hairline Landmarks.

Brow dermoids are benign pediatric neoplasms. Endoscopic resection requires incisions placed within the temporal hairline, which is often difficult to reliably locate in infants. The authors studied adult and pediatric patients to define the location of the hairline in relationship to simple facial landmarks. Adult and pediatric patients who were seen in preoperative consultation for facial surgical procedures were identified and included in the study. Patients with a history of facial trauma, surgery, or congenital anomalies were excluded. Digital photographs were analyzed to measure corneal white-to-white diameter. Lines were drawn connecting the tragus to the lateral canthus and from the inferiormost aspect of the earlobe bisecting the first line (line A). The maximal distance between line A and the temporal hairline was recorded. One hundred sixteen patients met the inclusion criteria, 81 adults and 35 pediatric patients. Average age was 55.9 years in adults and 3.2 years in the pediatric group. Measurements were normalized to the corneal diameter. Average temporal hairline distance from line A was 25.0 mm in adults and 21.8 mm in the pediatric group. Hairline position was not correlated with age or sex. The temporal hairline can be reliably located relative to a line drawn from the inferior aspect of the earlobe to the midpoint of the line connecting the lateral canthus and tragus. The temporal hairline is within 30 mm of this line. When designing a temporal hairline incision in infants, it can be safely placed 30 mm or more posterior to this line to ensure a well-hidden scar.

Facial Nerve Axonal Analysis and Anatomical Localization in Donor Nerve: Optimizing Axonal Load for Cross-Facial Nerve Grafting in Facial Reanimation.

BACKGROUND: Donor nerve axonal count over 900 in two-stage reconstructions using cross-facial nerve grafts is possibly associated with improved outcomes in facial reanimation. Facial nerve axonal analysis was performed to determine the ideal location for optimizing axonal load. Correlation of axonal number, branch diameter, and age was also assessed. METHODS: Twenty-eight fresh unpreserved cadaveric hemifaces were dissected exposing the extracranial facial nerve branches. Axonal counts at 2-cm intervals from the pes anserinus along branches inserting into the zygomaticus major muscle were taken, noting position relative to the zygomatic arch, posterior ramus border, lateral border of the zygomaticus muscle, and anterior parotid gland border. Nerves were fixed, sectioned, and stained with SMI-31 antineurofilament stain for digital axonal analysis. RESULTS: All specimens had one or more intraparotid zygomatic branches with over 900 axons, and 96 percent had an extraparotid branch with over 900 axons. The likelihood that a zygomatic branch would have over 900 axons at its last intraparotid point (mean, 6 mm posterior to the parotid border) was 92 percent (range, 67 to 100 percent) in contrast to 61 percent (range, 25 to 100 percent) when sampled at the first extraparotid point (mean, 14 mm anterior to the parotid border). Nerve cross-sectional area was positively correlated to its axonal count (R° = 78 percent; p < 0.0001), with nerve diameter over 0.6 mm predicting over 900 axons. Age did not correlate with axonal counts. CONCLUSIONS: Branches with adequate axonal load were found in all specimens. The likelihood of adequate branch selection improved from 61 percent to 92 percent with short retrograde intraparotid dissection. Nerve diameter correlated with axonal load.

Public reporting of patient safety metrics: ready or not?

In its 1999 report, the Institute of Medicine estimated that medical error leads to between 44,000 and 98,000 deaths per year. Given that statistic, public reporting of quality and safety metrics is a welcome response that may serve to reduce the rate of adverse events and restore patients' trust in the health care system. To ensure that any public reporting system fulfills its potential, several questions must be addressed: Are we measuring the right metrics? Are the metrics accurate, valid, and is their public reporting effecting change? Based on a review of the literature, it is clear that current metrics suffer from low reliability, low validity, and possibly minimal relevance to the intended consumer. To improve data collection and analysis, both physicians and health care consumers need to be involved in the design and collection of metrics. Until we have a valid, reliable, and actionable data set at our fingertips, it would behoove patients, providers, and institutions to look at outcome and safety metrics with a skeptical and discerning eye.

Insights from San Francisco match rank lists, part II: are programs doing it wrong?

BACKGROUND: It is well known that the ideal method of creating rank lists for the match is the "true-preference" strategy. However, there is anecdotal and survey-based evidence that programs and applicants often deviate from this strategy. We analyzed rank lists of applicants and programs participating in the Plastic Surgery San Francisco Match to investigate whether programs were following an optimal strategy. METHODS: We obtained deidentified program and applicant rank lists and their match results from SF Match for 4 years (2010-2013). Statistical analysis was carried out with Microsoft Excel. RESULTS: The number of applicants, applications submitted, interviews offered, and match rate were all relatively stable over this 4-year period (range, 117-138 applicants, 36-41 applications, 9.0-10.3 interviews per applicant, and 78%-86% match rate). The "number needed to match" for programs was 4 (range, 1-21). A subset of applicants had poor average ranks on program rank lists but was nevertheless ranked to match by one program. Forty-six percent of these applicants matched at their top choice compared to 20% of matched controls. CONCLUSIONS: The independent Plastic Surgery Match has become less competitive over the last decade. The low average number needed to match in the context of a high applicant match rate supports the hypothesis that programs and applicants may be modifying rank lists from a true-preference list. Noncompetitive applicants are occasionally ranked to match by a program, and these applicants tend to match at their top choice far more often than the average applicant-providing further support to this hypothesis.

Insights from the San Francisco Match rank list data: how many interviews does it take to match?

BACKGROUND: For many years, the independent plastic surgery match has been regarded as a competitive process. Applicants expend significant time and resources applying to, and interviewing with, many programs to increase their chance for success. Public data from the San Francisco (SF) Match provide no predictors of success in the Match. Previous survey-based studies have provided some data, but suffer from recall and sampling bias. The purpose of this study was to provide match participants with objective primary-source data that can aid them in making informed decisions with regard to planning their interviews. METHODS: Four years of fully deidentified individual-level and program-level data from the SF Match (2010-2013) were analyzed. Data included number of programs applied to, interview offers, and length of rank lists. For applicants who matched, data included the applicant's rank of program and the program's rank of applicant. RESULTS: During the 4 match years, 434 (86.3%) of 503 applicants received at least 1 interview offer. Of these candidates, 355 (82%) matched. Match rate increased with number of interviews, reaching 96% for those with 5 or more interview offers; 95% of applicants matched within their top 7 choices. On average, applicants matched at number 2.9 on their rank lists. CONCLUSIONS: Number of interview invitations is a strong predictor of success in the independent plastic surgery match, with the "magic number" being 5. Applicants rarely match to programs below number 7 on their rank lists. These data can aid applicants wishing to maximize their potential while minimizing unnecessary expenditures.

Fixing the match: a survey of resident behaviors.

BACKGROUND: The authors studied residency applicant attitudes toward rank list creation, communication with programs, and the impact of these factors on their performance in the Match. METHODS: An anonymous, 26-question, multiple-choice, online survey was distributed to the program coordinators of every Accreditation Council for Graduate Medical Education-accredited program participating in the National Resident Matching Program for whom e-mail addresses were available. The survey addressed five areas: (1) demographics and interview characteristics, (2) preinterview and interview factors, (3) postinterview contact, (4) importance of various factors in rank list creation, and (5) Match outcome. Survey responses were analyzed with Microsoft Excel. RESULTS: A total of 1179 responses were received. It was not possible to calculate a response rate, because the number of residents receiving the survey was not known. The majority of respondents (78 percent) reported postinterview contact with a program. A large portion of respondents (42 percent) considered such contact to be important in the creation of their rank lists. Half of all respondents admitted to exaggerating their interest in a program during or after an interview. The majority of respondents (87.5 percent) received no assistance in covering the costs of "second-look" visits to programs. CONCLUSIONS: Applicants may be modifying their rank lists in response to post-interview contact from programs; furthermore, they usually have no assistance in paying for the cost of second looks. To level the playing field for students and programs alike, the authors propose that the National Resident Matching Program modify residency interview rules to (1) disallow any postinterview contact between programs and students, and (2) disallow second looks.