Welcome back to the journal review. This month we're working through the November 2026 issue of Dermatologic Surgery, and we've got four pieces on the docket — a fellowship match analysis, a systematic review on topical therapy for melanoma in situ and lentigo maligna, a workforce study on retirement patterns in Mohs surgeons, and a small technique-driven case series on nail reconstruction. Let's get into it. First up is a communications piece looking at the brand-new signaling system in the Mohs Micrographic Surgery and Dermatologic Oncology Fellowship Match. For anyone not knee-deep in fellowship administration, 2024 was the first year applicants could send ten preference signals to programs — essentially a formalized way of saying "I'm genuinely interested in you," meant to help programs sift signal from noise in an increasingly saturated applicant pool. This is a real original study, not just a survey write-up, so it's got a genuine analytic backbone. The authors ran what they call a multicycle analysis — comparing the 2024 signaling cycle against the two presignaling years, 2022 and 2023 — using applicant-level and program-level SF Match data plus a national program director survey. Methodologically, this retrospective multi-source design makes sense because you can't randomize something like a national match process; the only way to isolate the effect of a new policy is to compare the cohort before it existed to the cohort after. For their statistical modeling, they used negative binomial regression and log-binomial regression. Worth pausing on why: negative binomial regression is suited to count data — like number of programs applied to or ranked — that's overdispersed, meaning the variance is bigger than you'd expect from a simpler Poisson model, which is typically the case with human behavioral counts like this. Log-binomial regression, meanwhile, was used for the binary matched-or-not outcome specifically because it estimates a risk ratio directly, which is far more intuitive to communicate than an odds ratio from standard logistic regression — especially here, where the outcome isn't rare, so odds ratios would otherwise exaggerate the apparent effect size. On to results. In 2024, about two-thirds of applicants matched — 63 and a half percent — and among those who matched, about seven in ten matched to a program they had actually signaled. The vast majority of applicants, roughly 86%, used all ten of their signals, and doing so was associated with meaningfully higher odds of matching, around a 56% relative increase. Ranking more programs also helped in a dose-dependent way — each additional five programs ranked bumped match probability up by about 25%. From the program side, positions filled by a signaled applicant were more than twice as likely as positions filled by a nonsignaled one. And on the program director survey, about two-thirds reported signaling had a moderate-to-significant impact on interview and secondary application decisions, but more than three-quarters said it had little to no influence on final rank order — so signaling seems to function as a door-opener for interviews, not as a rank-list decider. Compare that to the presignaling years, where predictors of matching were more tied to traditional variables — allopathic training conferred a meaningfully higher match likelihood, and more recent graduates did better too. The authors note that the overall match rate didn't significantly change with signaling — still in that 62-to-64% range both before and after — but the pathway to matching shifted. Interestingly, those traditional predictors — degree type, graduation year — lost some of their predictive strength after signaling was introduced, which the authors interpret as a possible equity benefit: signaling may give nontraditional applicants a way to stand out that doesn't rely on pedigree markers. Limitations here are real and the authors are upfront about them — this is only one year of postsignaling data, program director survey responses carry self-report bias, and critically, the dataset can't capture whether a nonsignaled match reflected a home-institution placement, which would be a totally different story than genuine signaling "failure." So take the equity hypothesis as provocative but preliminary. Practically, for those of you mentoring fellowship applicants: the actionable takeaway is to tell them to use all ten signals, and to use them on programs where they're genuinely competitive rather than spreading them thin or saving them strategically — the data suggest full utilization correlates with better outcomes. For programs, the message is that signals are a legitimate screening tool for interview decisions, but you should keep holistic review intact for final ranking, since this data confirms a meaningful chunk of matches still land on nonsignaled programs. This isn't something that changes how you operate clinically, obviously, but if you sit on a fellowship selection committee, it's genuinely actionable guidance for this cycle. Second article: a systematic review on topical imiquimod for melanoma in situ and lentigo maligna. This is a review, so there's no new patient cohort being generated — it's a synthesis, and it updates a decade-old benchmark review by Mora and colleagues that looked at about 350 tumors through 2014 and found roughly 76% histologic clearance with a recurrence rate around 2%. The gap this new review addresses is that nothing has synthesized the literature since, despite continued off-label use, and the prior work leaned heavily toward lentigo maligna specifically rather than melanoma in situ at other sites. Methodologically, they followed PRISMA 2020 guidelines, prospectively registered the protocol, and searched PubMed, Web of Science, and Cochrane for imiquimod monotherapy studies — deliberately excluding any cohort where combination therapy with surgery, cryotherapy, or radiation was used, because you can't disentangle imiquimod's independent effect from an adjunct. That's a sound exclusion criterion worth noting: if a paper mixes modalities without lesion-level breakdown, you simply can't attribute the outcome to either one. For quality appraisal, they used the Joanna Briggs Institute checklist for case series and case reports, and the QUIPS tool for the one biomarker-based prognostic study — these are standardized risk-of-bias frameworks designed for exactly this scenario, where you're synthesizing predominantly observational, non-randomized, case-level data and need a structured way to judge how much weight each study deserves. They ended up with thirteen studies, 129 total cases, spanning 2000 through early 2025. Combined histologic clearance landed at about 84 and a half percent — meaningfully higher than the older Mora benchmark, though cross-study comparison here is soft given the heterogeneity. Anatomically, the overwhelming majority of lesions were facial, with periocular lesions actually showing the highest clearance at just under 92%, compared to about 82% for nonperiocular facial sites. There was a statistically significant dose-response relationship between treatment duration and clearance — short courses of eight weeks or less cleared only about 74%, standard nine-to-sixteen week courses hit about 85%, and extended courses beyond sixteen weeks reached 90%. That's a trend worth remembering if you're counseling a patient toward this route: duration matters, and skimping on it measurably underperforms. Lesion size showed a similar directional trend favoring smaller lesions under two centimeters, though that didn't reach statistical significance — so take it as a plausible pattern, not an established cutoff. Prior treatment history didn't meaningfully affect clearance either way. The one factor that was both statistically significant and clinically telling was the development of a visible inflammatory response during treatment, which strongly correlated with success — this fits the known mechanism, since imiquimod's antitumor effect is mediated through that induced local immune response via Toll-like receptor 7 activation. On recurrence, the numbers are reassuring but come with a big asterisk: long-term follow-up was only available for about 40% of successfully treated cases, with a mean follow-up around a year and a half. Within that subset, recurrence was low, under 5%, and in the full cohort the overall recurrence rate was under 2% — both recurrences occurred in lesions larger than four centimeters, reinforcing that lesion size is where the real risk concentrates. Adverse events were generally mild — erythema and local irritation as expected — though periocular cases carried specific risks like ectropion and conjunctival chemosis in a small minority. The authors' own conclusion is measured: imiquimod is a reasonable, efficacious nonsurgical option, particularly for anatomically difficult sites, but the heterogeneous protocols, limited long-term data, and small per-study sample sizes limit how confidently we can generalize, and they explicitly call for standardized treatment protocols going forward. For your practice, I'd frame this as interesting and reinforcing rather than practice-changing. It doesn't touch what you do with a biopsy-proven lentigo maligna that's surgically resectable — Mohs still carries clearance rates near 98% and remains the standard. But for that genuinely difficult subset — elderly patients, periocular or other cosmetically sensitive sites, patients who decline or can't tolerate surgery — this gives you a somewhat more robust evidence base than before to support a shared decision-making conversation, with the specific, quotable numbers being roughly 85% clearance and under 2% recurrence with adequate treatment duration, understanding that long-term follow-up remains thin. Third article, and this one's a genuine retrospective cohort study: retirement patterns among Mohs micrographic surgeons versus general dermatologists, using national Medicare Part B and Part D data. The clinical question is straightforward — dermatology has a reputation as a low-burnout, high-satisfaction specialty, but nobody had actually characterized whether surgical subspecialists retire at different rates than general dermatologists, and that matters for workforce planning as skin cancer incidence keeps climbing. The design: they identified board-certified dermatologists practicing between 2012 and 2021 using Medicare billing and prescribing data, defining "retirement" operationally as two consecutive years of zero billing and zero prescribing with no return — a deliberately conservative definition, and the authors note this is stricter than a prior study that used a single CPT code and captured more retirements as a result. Methodologically, two techniques are doing the heavy lifting here. First, propensity score matching: because Mohs surgeons and general dermatologists differ systematically at baseline — more male, more years in practice — the authors matched each Mohs surgeon one-to-one with a general dermatologist with a similar predicted probability of being a Mohs surgeon based on those covariates. This is essentially an attempt to simulate randomization after the fact, balancing known confounders so that any remaining difference is less likely to just be an artifact of who ends up choosing each path. Second, multivariable Cox proportional hazards regression — this is your classic survival analysis tool, and it's the right choice here because it directly compares the instantaneous risk of retiring between groups at every time point among physicians still actively practicing, which elegantly handles the fact that people entered the data at different career stages and were followed for different lengths of time without requiring everyone to reach the same endpoint. The authors also checked the proportional hazards assumption using Schoenfeld residuals and log-log survival plots — a good methodological touch, since Cox models only give valid comparisons if the relative hazard between groups stays roughly constant over time, and verifying that assumption is often skipped but shouldn't be. Results: among just over eleven thousand dermatologists, roughly 2,800 were Mohs surgeons and about 8,000 were general dermatologists. In the matched cohorts, Mohs surgeons retired at a rate of about 1.6%, compared to 2.6% among general dermatologists. After adjustment in the multivariable Cox model, Mohs surgeons had a 43% lower risk of retirement — an adjusted hazard ratio of 0.57 — and this was both statistically significant and, frankly, a fairly large effect size, so this clears the bar for clinically meaningful, not just statistically detectable. Male sex was independently associated with lower retirement risk as well, around a 35% reduction, which the authors note separately from the subspecialty finding. Discussion-wise, the authors speculate — and they're appropriately tentative about this — that lower retirement risk among Mohs surgeons could reflect unmeasured factors like income structure, lower burnout, or the ability to sustain income on comparatively fewer clinical hours, rather than something directly attributable to being a surgeon per se. They also raise an interesting downstream implication: if established Mohs surgeons are staying in practice longer, that could contribute to workforce saturation in already-dense metropolitan markets while shortages persist in rural areas — a point worth sitting with if you're advising fellows on practice location. Limitations are exactly what you'd expect from claims-based data: it can't distinguish someone who went part-time or moved off Medicare from someone who truly stopped practicing, it can't separate voluntary retirement from retirement due to death or disability, and it has no data on practice setting, income, or geography — all plausible confounders that simply weren't available. Practically, this isn't something that changes how you run your clinic tomorrow, but it is a genuinely useful data point for workforce conversations — fellowship program planning, projections about procedural capacity, conversations with trainees about career longevity in this subspecialty. File it under informative and well-supported, not practice-changing. Last one is a short communications piece — really a small case series describing a novel surgical technique — for post-traumatic onychodystrophy, specifically the subset with lateral nail-plate long-axis deviation, thickening, and discoloration. The clinical problem the authors are addressing is that existing options are unsatisfying: total nail extraction with matricectomy permanently sacrifices the nail, while simply trimming the nail shorter gives only temporary relief without correcting the underlying anatomic distortion. Their proposed technique is a combination procedure: partial nail matricectomy, realignment of the nail plate's long axis to match the long axis of the distal phalanx, partial excision of hypertrophied lateral nail-fold soft tissue on both sides, and widening of the nail bed. Technically, it starts with nail plate extraction using a Freer septum elevator, then a U-shaped incision around the toe tip with dissection of the nail bed and matrix off the underlying phalanx, oblique incisions on the proximal nail fold to expose the matrix, and partial excision of the lateral horns of the nail matrix — this step is doing important mechanical work, since it disconnects the matrix from the surrounding ligamentous and soft tissue anchoring, which is what allows the matrix-bearing flap to actually rotate without springing back. The flap is then elevated, medially rotated into alignment with the distal phalanx's long axis, the hypertrophied lateral nail fold tissue is trimmed to widen the bed, and the flap is sutured into its corrected position. This was performed and followed in three patients, each followed for at least a year — patients with onychomycosis or shorter follow-up were excluded. Outcomes were tracked across five indices: deviation angle measured objectively between the nail plate and phalanx axes, plus patient-reported numeric ratings for thickness, discoloration, pain, and overall satisfaction. All three patients improved across all five measures, and as of this report, there have been no relapses and no complications. The authors' proposed mechanism is worth understanding even from this tiny series: they hypothesize that trauma disrupts the ligamentous structures supporting the nail unit, producing the axis deviation, while nail avulsion itself triggers hypertrophy of the lateral and distal nail folds that override and narrow the nail bed — which in turn drives the upward growth, thickening, and discoloration. They draw an explicit parallel to pincer nail pathophysiology, suggesting techniques from that literature may be transferable here, which is a reasonable extrapolation given the shared anatomic logic, though it remains speculative. The limitation is simply the nature of the report — three patients, no comparator, no long-term durability data beyond the current follow-up window, and the authors themselves call for larger studies and longer follow-up before this can be validated. So frame this clearly as hypothesis-generating and technically interesting, not something to adopt wholesale into practice based on this alone. If you do encounter the specific clinical picture they describe — deviated, thickened, discolored nail plate after trauma, in a patient who wants to avoid permanent nail sacrifice — this gives you a conceptually sound reconstructive option worth knowing exists, but I wouldn't call it practice-changing until a larger series with longer follow-up confirms durability. That wraps the four pieces for this issue. To summarize the through-line: the match signaling data gives fellowship programs and applicants genuinely actionable guidance for this cycle, the imiquimod review strengthens — but doesn't revolutionize — the nonsurgical option you already reach for in select LM and MIS cases, the retirement data is a solid workforce-planning insight rather than a clinical one, and the onychodystrophy technique is a promising but very early reconstructive idea worth filing away rather than adopting outright. Thanks for listening, and I'll see you next month.