Welcome back to the journal review. This month we're working through the October 2026 issue of Dermatologic Surgery, and I've pulled four pieces worth your time: a retrospective cohort study on biopsy-to-debulk concordance in basal cell carcinoma, a retrospective analysis of how access to dermatology affects outcomes in dermatofibrosarcoma protuberans, a research letter extending last issue's work on reflected surgical lighting and drape choice, and a single-arm pilot study on virtual reality for preprocedural anxiety. Let's get into it.

First up is a retrospective cohort study out of a tertiary dermatologic surgery center, looking at histopathologic concordance between preoperative biopsy and debulk specimens in basal cell carcinoma treated with Mohs. The clinical gap here is a familiar one: for cutaneous squamous cell carcinoma there's reasonably good evidence supporting routine debulk analysis, but for basal cell carcinoma, whether to even submit the debulk for permanent sectioning is left entirely to surgeon discretion, with no consensus guideline behind it. This group wanted to know how often biopsy and debulk histology actually disagree, and whether that disagreement changes anything that matters intraoperatively.

Methodologically, this was a single-center retrospective chart review of consecutive Mohs cases for basal cell carcinoma over one calendar year, five hundred fifty-nine patients total. They classified tumors as low-risk — superficial or nodular — versus high-risk, which captured infiltrative, morpheaform, basosquamous, or micronodular features, with any mixed tumor containing an aggressive component bumped into the high-risk bucket. Discordance was simply a change in risk category between the biopsy and the debulk, split into upstaging and downstaging, the latter including the scenario where the debulk showed only scar. A retrospective design makes sense here as essentially the only option — you can't ethically or practically randomize what tissue gets submitted for permanent sectioning, so this is a chart-mining exercise to generate hypothesis-level data before anyone invests in a prospective, costlier study.

Statistically, the main tool was the chi-square test, used to ask whether two categorical variables are associated — for example, whether submitting a debulk specimen is associated with needing more than one Mohs stage. In plain terms, chi-square compares the number of cases actually observed in each cell of a table against the number you'd expect if there were truly no relationship between the two variables, and asks whether that gap is bigger than chance would produce. It's the right tool here because both exposure and outcome are categorical — yes or no debulk, one stage or more than one stage — rather than continuous measurements.

On to the numbers. Debulking was performed in the large majority of cases, about ninety-three percent. Larger tumors were significantly more likely to be debulked, which tracks with what's been shown in squamous cell carcinoma, where bigger lesions more often harbor aggressive features. Biopsy type — punch versus shave versus other — did not significantly predict whether the risk category changed, so don't lean on biopsy technique as a proxy for how much subclinical disease you're dealing with. The headline discordance numbers: about eight percent of tumors were upstaged from low to high risk, roughly six percent were downstaged, and about eleven percent of debulk specimens came back essentially as scar only, meaning the biopsy itself may have been curative in a meaningful subset of patients. Add it all up and any stage change — up or down — occurred in about forty-one percent of cases, which is a strikingly high discordance rate.

But here's the clinically important twist: despite that discordance, upstaging did not significantly predict how many Mohs stages a case required, and reconstruction complexity didn't differ between upstaged and non-upstaged tumors either. Submitting a debulk specimen was significantly associated with needing more than one stage, but that's probably just a marker of tumor size and complexity rather than something upstaging itself was driving. Perineural invasion, while rare at about one and a half percent, was flagged as the one finding that genuinely changes downstream management — triggering multidisciplinary review and consideration of imaging or adjuvant radiation.

The authors are appropriately cautious in their discussion. They note the long, sometimes eighteen-month, interval between biopsy and surgery as a plausible driver of upstaging — tumors have time to evolve. Limitations are the usual retrospective suspects: incomplete data on things like tumor size and defect size, which were only available in a subset of patients, a surgeon-dependent decision about whether to even debulk in the first place, and no recurrence data to tell us whether any of this discordance actually matters for long-term control.

Practically, here's where I'd land this one for a working Mohs surgeon. This is not practice-changing — it doesn't tell you to start or stop routinely submitting debulk tissue for permanent sectioning. What it does is reinforce that histologic discordance between biopsy and debulk is common, finding it doesn't reliably change your intraoperative stage count or reconstruction plan, and the main actionable yield from debulk analysis is catching the rare but consequential finding of perineural invasion. Routine permanent-section debulk analysis for every basal cell carcinoma is probably low yield and resource-intensive; a more selective approach targeting larger or higher-risk-appearing tumors is reasonable until someone runs the prospective, cost-effectiveness-informed study the authors are calling for.

Next, a retrospective review of dermatofibrosarcoma protuberans outcomes across a regional health system spanning twenty-four years, asking a pointed question: does who first sees the patient — a dermatologist versus a non-dermatologist — change the treatment trajectory for this rare, locally aggressive sarcoma.

The design was a retrospective records pull using diagnosis codes, narrowed from a hundred eight candidate patients down to seventy-two with complete, histologically confirmed records spanning the full treatment course. The authors pre-specified three null hypotheses tied to three sequential endpoints: whether dermatology versus non-dermatology first evaluation changes the rate of appropriate nonexcisional pretreatment biopsy, whether that biopsy changes the rate of getting appropriate definitive surgery — meaning peripheral and deep en face margin assessment or wide local excision with two-centimeter margins — and whether that appropriate surgery changes the rate of positive margins or recurrence. That's a nice chain-of-custody design: it's tracing the diagnostic and surgical pathway step by step to find where the system breaks down.

For statistics, they used Fisher's exact test for all three comparisons, with a chi-square test incorporating a Yates correction used separately to check whether tumor size was confounding the final endpoint. Fisher's exact test is the right choice here rather than a standard chi-square because several of their two-by-two tables have small cell counts — things like two patients in a cell — where the chi-square's approximation to the underlying distribution breaks down. Fisher's test instead calculates the exact probability of seeing a table as extreme as the one observed, under the assumption of no association, which makes it the more trustworthy tool when numbers are sparse, exactly the situation you get with a rare tumor at a regional center.

The results are about as clean a demonstration of a diagnostic pathway problem as you'll see. Only about half of all patients, forty-nine percent, received any nonexcisional pretreatment biopsy before definitive treatment. But that rate was hugely dependent on who saw the patient first: dermatologists obtained a pretreatment biopsy in roughly eighty-five percent of cases, compared with about forty-one percent for non-dermatologists — a statistically significant and clinically enormous gap. That initial biopsy decision then cascaded: among patients who got an appropriate pretreatment biopsy, ninety-four percent went on to appropriate definitive surgery, compared with just five percent of those who didn't — again, both statistically significant and clinically enormous. And appropriate definitive surgery, in turn, was strongly associated with better margin outcomes — about twenty-nine percent positive margins or recurrence in that group versus ninety-five percent in the group that received other surgical approaches. Reassuringly, tumor size at presentation didn't independently predict margin outcomes, so this isn't simply a story of bigger, harder tumors ending up with non-dermatologists. And there were no metastases anywhere in the cohort, which fits the generally indolent-but-locally-destructive biology of this tumor, even when things go wrong locally.

One nuance worth flagging for the discussion beat: a post hoc analysis found that failing to get appropriate surgery the first time around did not significantly increase the rate of positive margins or recurrence after a second treatment attempt — in other words, if you catch the inadequate initial excision and go back with proper margin assessment, you can often still salvage the outcome, just at the cost of extra procedures and, as the authors note, four patients in the no-pretreatment-biopsy group ultimately needed more than three lines of treatment.

Limitations are straightforward: retrospective, single regional system, modest absolute numbers once you stratify into subgroups, and some of those odds ratios come with very wide confidence intervals even though they cross the threshold of significance, which is the authors' own data telling you the point estimate is less precise than it looks.

Practically, this isn't new information for you as a fellowship-trained Mohs surgeon — you already know dermatofibrosarcoma protuberans needs a nonexcisional biopsy and margin-controlled excision. The actionable takeaway is upstream of your practice: this is ammunition for provider education. The leverage point for improving outcomes in this disease is training primary care, urgent care, and non-dermatology surgical colleagues to recognize a firm, indurated plaque for what it might be and get either a core or punch biopsy, or a dermatology referral, before anyone reaches for a scalpel with excisional intent.

Third, a research letter responding to and extending a prior paper on reflected surgical lighting at the Mohs surgeon's eye level. The original study, by the same senior author group, had shown that blue disposable drapes substantially cut reflected illuminance compared to white disposable drapes, with working distance being the dominant variable overall. The acknowledged gap was that reusable woven cloth towels weren't tested, despite woven and nonwoven materials having genuinely different optical reflectance properties.

This follow-up is a bench experiment, not a clinical trial, so there's no patient population or inferential statistics to walk through — it's a repeat of the identical simulated measurement protocol, same three working distances, same two lamp orientations, same four lamp intensities, same three simulated Fitzpatrick skin tones, same calibrated lux meter, just swapping in a reusable, lint-free, one-hundred-percent-cotton woven blue surgical towel as the drape material.

The results were consistent and, frankly, additive to what they'd already shown. At the reference twenty-inch working distance, white disposable drapes produced about five hundred twenty lux of reflected illuminance, blue disposable drapes brought that down to about two hundred fifty, and the reusable woven blue cloth brought it down further still, to about two hundred fifteen — roughly a fifty-nine percent reduction compared to white disposable overall, with an incremental fourteen percent reduction beyond what blue disposable drapes alone achieved. And this wasn't a one-off: the woven blue cloth beat disposable blue drapes in all seventy-two of seventy-two matched conditions tested, which, while purely descriptive rather than a formal hypothesis test, is about as consistent a signal as a bench study can produce.

The proposed mechanism is a materials-science one — woven fabric has a different fiber architecture and bidirectional reflectance pattern than nonwoven disposable material, favoring more diffuse rather than specular scattering of light back toward the surgeon's eyes. The authors also layer on a sustainability argument: the World Health Organization already considers sterile reusable woven and disposable nonwoven drapes equivalent from an infection-control standpoint, which means the choice between them can legitimately be driven by other factors — and reusable textiles carry meaningfully lower water, energy, and waste footprints alongside this newly identified ocular exposure benefit.

There's no limitations section to speak of beyond what's implicit — this is simulated bench data, not a measurement of actual cumulative surgeon exposure over a career, and it doesn't address wavelength-specific or photobiologic effects, which the authors flag as the next logical step.

Practically, this is a nice, low-friction, low-cost adjustment: if your practice is choosing between disposable and reusable draping materials, this gives you a reason, beyond cost and sustainability, to favor woven blue cloth — less reflected light hitting your eyes across a career of cases, with working distance still mattering more than anything else you can control. It's a sensible occupational-health tweak rather than a practice-changing mandate.

Last, a single-arm pilot study testing virtual reality for preprocedural anxiety in general dermatologic surgery — excisions and biopsies outside the face, specifically, since facial procedures would interfere with headset placement. The gap being addressed is that VR has already shown benefit specifically within Mohs surgery, but its use across the broader spectrum of dermatologic procedures was still largely unexamined.

The design was a single-arm, within-subject pre-post study: twenty-eight patients at one academic center each underwent a five-minute VR-guided meditation immediately before their procedure, with anxiety measured right before and right after using the STAI-6, a validated six-item anxiety inventory. The authors used each patient as their own comparator rather than a separate control group, explicitly to avoid disrupting clinical flow — a pragmatic, feasibility-oriented choice that makes sense for a pilot, though it does mean you can't fully separate the effect of the VR itself from the simple passage of time, a bit of extra counseling attention, or regression toward the mean as patients settle into their visit.

For statistics, they used a one-tailed Wilcoxon signed-rank test for paired samples. This is the nonparametric cousin of the paired t-test, appropriate here because the sample is small and the outcome is an ordinal anxiety score rather than something you can confidently assume is normally distributed. Rather than comparing means directly, it ranks the magnitude of each patient's individual before-and-after difference and tests whether those ranks are systematically shifted in one direction — it's a good fit for pilot-sized behavioral data where you don't want to lean on distributional assumptions you can't verify.

The results: of forty-one interested patients, twenty-eight met inclusion criteria and completed the intervention, mean age about sixty, evenly split by sex. Mean STAI-6 anxiety dropped from about nine-point-seven before to about seven-point-eight after, a reduction of roughly two points, or about twenty percent relative reduction, and this was statistically significant. Looking at the individual sub-questions, feeling calm, tense, upset, relaxed, and worried all improved significantly; feeling content was the one item that didn't reach significance, worth noting since it suggests the effect may be more about dampening acute distress than genuinely elevating mood.

The authors are candid about the limitations: no true control arm, a modest sample from a single academic center, self-selected participants with no blinding, and — importantly — a study population that explicitly excluded patients with generalized anxiety, panic attacks, claustrophobia, or motion sickness, which are arguably exactly the patients most likely to benefit from an anxiolytic intervention, so generalizability to your highest-anxiety patients is genuinely uncertain. They also flag practical barriers: cybersickness risk in a subset of patients, and upfront device cost and staff training time for a consumer headset system.

Practically, I'd file this as interesting but not yet practice-changing. The anxiety reduction was statistically significant and the absolute two-point drop on a six-item scale is plausibly clinically meaningful for an individual patient's procedural experience, but this is feasibility-level pilot data without a control group, in a population that screened out the patients most likely to need help. It's a reasonable thing to consider piloting in your own practice if you have the equipment and bandwidth, but it's not something the evidence yet supports mandating or expecting dramatic effect from at scale.

That wraps the October issue. Four pieces with a common thread, really — each one is chipping away at a gap between what we assume in practice and what the data actually show, whether that's debulk histology, referral pathways, operating room ergonomics, or patient experience. Thanks for listening, and I'll see you next month.