Background: Despite generally favorable outcomes of ligation of intersphincteric fistula tract for anal fistula, several patients experience persistent or recurrent disease.
Objective: This study aimed to identify patterns of treatment failure after ligation of intersphincteric fistula tract and evaluate their outcomes.
Design: Multicenter retrospective study.
Setting: Five university hospitals in Thailand.
Patients: Those diagnosed with persistent or recurrent cryptoglandular anal fistula after ligation of intersphincteric fistula tract between 2012 and 2020.
Main outcome measures: Fistula characteristics, patterns of treatment failure and outcomes of subsequent management were analyzed.
Results: A total of 200 patients were included, of whom 12.5% had persistent fistulas and 87.5% had recurrence. Median time to recurrence was 6 months (range, 1-48). Most treatment failures (n = 186, 93%) occurred within one year after surgery. Six distinct patterns of treatment failure were identified. The most common was type 1 (I-E, tract running from the previous internal opening (I) to an external opening (E) - known as an original fistula: n=116, 58%), followed by type 2 (I-L, tract running from the previous internal opening (I) to an unhealed intersphincteric wound - known as an intersphincteric fistula: n = 56, 28%). Type 1 failure was effectively treated by redo- ligation of intersphincteric fistula tract with an 87% success rate. Overall, the healing rate was 90.5% - with a median follow-up period of 18 months (range, 12-38).
Limitations: Exclusion of non-cryptoglandular fistulas may limit generalizability. The retrospective multicenter design and surgeon-dependent decision-making may have introduced selection bias and heterogeneity in surgical technique.
Conclusions: Most treatment failures occurred within one year after ligation of intersphincteric fistula tract. Understanding the six distinct failure patterns enables more appropriate surgical decision-making. The majority represented either recurrence of the original tract or persistence within the intersphincteric space. See Video Abstract.
Background: Standardized Letter of Recommendation was introduced to capture more meaningful fellowship applicant evaluations. There are concerns that this letter may introduce subjectivity to assessments traditionally measured by objective metrics, such as board scores and publications.
Objective: To evaluate the influence of individual application elements on ranking by referee.
Design: Collected data included applicant demographics and objective assessments of academic performance, including in-service exam percentiles, Step 1 and Step 2 scores, and publication count. Summary statistics and odds ratios from repeated measures logistic regression models using the method of generalized estimating equations are reported.
Settings: Standardized letters submitted to a single institution during the 2019 application cycle were analyzed.
Main outcome measures: To determine if objective applicant parameters are associated with being subjectively ranked #1 by the referee.
Results: Among 302 standardized LOR forms, 51.3% of the applicants were ranked #1 by the referee. Applicants ranked #1 had higher junior and senior in-training exam percentiles (60.3 vs 50.9, p = 0.0002 and 59.5 vs 48.0, p < 0.0001, respectively), Step 1 scores (235 vs 231, p = 0.0287) and average number of publications (5.9 vs 3.5, p = 0.0011). For each unit rise of in-training percentile and publication number, the odd ratio for being ranked #1 improved by 1.03 (p = 0.002) and 1.08 (p = 0.013), respectively. Ranks lower than #1 did not demonstrate the expected decline in objective performance metrics with decreasing rank except Step 1 percentile (p ≤ 0.001). Technical ability was the strongest predictor of a #1 ranking.
Limitations: Analyses limited to elements included in standardized letters.
Conclusions: Standardized letters retain the relationship between objective performance metrics and subjective ratings for applicants ranked #1 by referee. However, this relationship is lost when applicants are ranked lower than first place. Our findings provide valuable insights and inform best practices for employing standardized letters of recommendation in the context of fellowship selection. See Video Abstract.

