Welcome back to the journal review. This episode covers the December twenty twenty-five issue of JAMA Dermatology, and we've got three pieces to get through — a randomized trial on postoperative wound dressings, a retrospective cohort study on skin cancer surveillance in transplant recipients, and the accompanying editorial that puts that cohort study into clinical context. Let's get into it.

First up is a randomized clinical trial out of Indiana University comparing hydrocolloid dressing against daily petroleum ointment for scar appearance after excisional surgery. The clinical problem here is pretty familiar to all of us — hydrocolloid dressings have a long track record in chronic wound care and in split-thickness skin graft donor sites, where they've actually been shown to cut healing time by around forty percent compared to standard care. The proposed mechanism is the usual moist-wound-healing story: the dressing forms a gel with wound exudate, keeps the site occluded but permeable, and theoretically speeds epithelialization while reducing inflammation. The appeal for our patients is obvious — one application that stays put for a week, versus daily cleaning and reapplication of ointment. But despite all that chronic-wound literature, nobody had actually run a randomized trial testing hydrocolloid against petroleum ointment specifically after dermatologic excisions with a standard linear closure. That's the gap this trial fills.

Methodologically, this is an investigator-blinded, individually randomized trial, single center, enrolling adults undergoing standard excision or Mohs surgery followed by a linear bilayered repair — they excluded flap and graft repairs, anyone using topical chemotherapy on the site, adhesive allergies, hair-bearing sites, prior hydrocolloid users, and anyone who couldn't complete the survey. Of four hundred forty-four patients screened, one hundred forty-six were randomized, split roughly evenly between the two arms. The blinding here is worth pausing on because it's a nice example of pragmatic trial design around an inherently unblindable intervention — you obviously can't hide a hydrocolloid dressing from the patient wearing it, so they blinded the surgeon instead, by having the nurse place wound care instructions in the room only after the surgeon left, and then had three separate blinded Mohs surgeons rate photographs, averaging their scores to reduce single-rater bias. The primary outcome was patient-reported scar appearance on a modified, image-based Visual Analog Scale, chosen deliberately because it didn't require an in-person visit — the authors note many of their patients commute more than an hour, so a photo-based scale maximized retention and also let them use out-of-state blinded evaluators. Sample size was powered using an effect size pulled from their own prior retrospective study, landing on a minimum of about one hundred patients to detect a difference with reasonable power.

On to results. At day seven, patients rated hydrocolloid-covered scars meaningfully higher than petroleum-covered ones — about seven point four versus six point six on that ten-point scale — and this early difference was statistically significant. But by day thirty and day ninety, that gap essentially closed; the confidence intervals straddled zero, meaning no real difference remained. Averaged across all three time points, hydrocolloid did edge out petroleum by about a quarter of a point on a ten-point scale, and technically that reached statistical significance, but let's be honest — a quarter of a point is not something any of us would notice on a photograph, so this is a textbook case of statistically significant but clinically trivial. Surgeon ratings essentially mirrored the patient ratings and didn't meaningfully differentiate the two arms either. On safety, the hydrocolloid group trended toward more complications across the board — postoperative bleeding in about one in five patients versus under one in ten with petroleum, wound dehiscence in about six percent versus zero, and surgical site pain in about one in five versus roughly one in eight — but none of these differences reached statistical significance, and nobody in either group needed postoperative antibiotics. Where hydrocolloid clearly won was patient experience: roughly nine in ten patients found it convenient versus under half in the petroleum group, and about three-quarters found it comfortable versus roughly half with petroleum — both differences statistically significant and clinically substantial.

The authors' own conclusion is measured: hydrocolloid is a suitable alternative that yields comparable scar outcomes and comparable complication rates, so the choice should hinge on cost and patient preference rather than assuming one is superior. The honest limitations here are worth flagging — this is a single center, patients obviously couldn't be blinded given the visible dressing, there was an imbalance in diabetes prevalence favoring more diabetics in the hydrocolloid arm which could confound wound healing outcomes, and the complication analysis, while reassuring in direction, was underpowered — those wide confidence intervals around bleeding and dehiscence mean a true difference can't be excluded with this sample size.

So what do you actually do with this. I'd call this practice-supporting rather than practice-changing. It gives you permission to offer hydrocolloid as a legitimate one-and-done option for the linear closure patient who has poor mobility, lives far away, or simply hates daily wound care — the comfort and convenience data are real and meaningful. But I would not walk away thinking it's risk-free; the trend toward more bleeding and dehiscence, even without statistical significance, means I'd still think twice for patients on anticoagulants or with other bleeding risk factors, and I'd want a larger safety study before fully equating the two.

Now let's move to the retrospective cohort study on skin cancer surveillance in solid organ transplant recipients, and then its companion editorial, since these two really need to be discussed together.

The background problem is one we all live with — transplant recipients carry a dramatically elevated skin cancer risk relative to the general population, on the order of twenty to sixty-five fold for cutaneous squamous cell carcinoma, seven to ten fold for basal cell carcinoma, and one and a half to three fold for melanoma, and these cancers behave more aggressively, with roughly nine-fold higher cancer-specific mortality. The trouble is that annual screening adherence in this population is poor, hampered by access, geography, insurance, and awareness. Back in twenty nineteen, a tool called SUNTRAC — the Skin and Ultraviolet Neoplasia Transplant Risk Assessment Calculator — was developed to stratify transplant recipients into low, medium, high, and very high risk categories using race and ethnicity as a proxy for phototype, pretransplant skin cancer history, age at transplant, sex, and organ type, with reported five-year skin cancer incidences of about one percent, six percent, fifteen percent, and forty-five percent across those four tiers respectively. What hadn't been tested was whether actually deploying this calculator as a real-world surveillance program changes screening behavior, cancer detection, and resource use. That's the gap this study addresses.

This is a retrospective cohort study run through Kaiser Permanente Northern California, an integrated system covering more than four and a half million members — and that setting is actually central to why this design makes sense. You can't randomize an entire health system's screening protocol, and a retrospective design lets them leverage years of existing registry and electronic health record data across a large, diverse population rather than prospectively recruiting a comparably sized cohort, which would take far too long for a rare-ish outcome like transplant-associated skin cancer. They identified just over two thousand adult solid organ transplant recipients from twenty sixteen through twenty twenty-three and matched each one to twenty non-recipients on sex, race and ethnicity, and treating facility — a sensible way to isolate the effect of transplant status itself while holding demographic risk factors constant. They then split the transplant cohort into a pre-implementation period, twenty sixteen through twenty twenty-one, and a post-implementation period starting in twenty twenty-two, when their modified program — called KP-SUNTRAC — went live. The modification worth noting is that they added a fifth risk category: any low or medium risk patient who develops a posttransplant skin cancer gets reclassified as very high risk for ongoing surveillance, since the original SUNTRAC tool only predicted first-cancer risk at the time of transplant and didn't address what to do afterward. They tracked first detected skin cancer through cancer registry data and diagnosis codes, confirmed by dermatologist chart review, and they measured resource utilization as the sum of dermatology and non-dermatology encounters plus pathology specimens, using a fixed twenty-seven month window after transplant so that pre- and post-implementation groups had equal follow-up time for fair comparison.

The headline numbers: transplant recipients had roughly an eight-fold increased risk of developing skin cancer compared to matched non-recipients, which is both highly significant statistically and, obviously, enormous clinically — nothing new conceptually, but a strong confirmation in a large modern cohort. After the surveillance program launched, the risk of a first detected skin cancer was about two and a half times higher in the post-implementation period than before — and this is the finding that matters most, because it tells you the program worked as intended: better screening found more cancer, presumably earlier. Screening rates themselves roughly doubled in the high-risk group and roughly doubled to a bit more than doubled in the very-high-risk group after implementation, both statistically significant improvements. And critically, healthcare resource utilization did not significantly increase system-wide after rollout — meaning they found more cancer in the patients who mattered most without overwhelming the system.

The editorial that follows, written by Doctor Jambusaria-Pahlajani — who was actually involved in developing the original SUNTRAC tool — fills in some texture worth knowing. She notes the cohort broke down as about fifteen hundred patients pre-implementation and roughly five hundred eighty post-implementation, and highlights a striking detail from the study: in the post-implementation period, ninety-three percent of high-risk patients and effectively all — one hundred percent — of very-high-risk patients developed at least one skin cancer. That's an extraordinarily high yield for a targeted screening population, which is really the whole argument for risk-stratified surveillance — you're not screening broadly and hoping to catch a few cases, you're screening a population where the disease is nearly universal. She also points out that screening rates in the low- and medium-risk groups actually decreased after implementation, which reflects appropriate deprioritization of resources away from patients unlikely to benefit, rather than a failure of the program. And she notes that resource utilization — visits and biopsies — was actually numerically lower in the high and very-high-risk groups post-implementation compared to pre, though that particular comparison didn't reach statistical significance.

On limitations, both the original authors and the editorialist are candid: it's a retrospective design with the usual issues of missing data and loss to follow-up when patients leave the health system, and it comes from a single, community-based, multispecialty integrated system, which may limit generalizability — their outreach model, using phone calls and patient portal messaging from dermatology clinics, assumes an infrastructure that smaller or less resourced practices may not have. The editorial also flags that the extension of SUNTRAC to reclassify post-transplant skin cancer patients as very-high-risk hasn't been independently validated, though it's a reasonable extrapolation given separate data showing about half of transplant patients with one skin cancer go on to develop another. And SUNTRAC itself doesn't capture immunosuppression regimen, UV exposure, or Fitzpatrick skin type directly, which could shift individual risk estimates somewhat, even though the tool has held up in multiple external validation cohorts.

Practically, here's the takeaway for those of us doing surgical oncology in this population: this doesn't change how you manage an individual transplant patient's skin cancer once it's in front of you, but it's a strong, real-world validation that risk-stratified surveillance — rather than blanket annual screening for every transplant recipient — is a viable, resource-efficient model that actually increases detection where it counts. If you're involved in program design, a multidisciplinary transplant-dermatology workflow, or advocating within your institution for how transplant patients get triaged into dermatology, this is genuinely useful ammunition. It supports moving away from one-size-fits-all annual screening toward a tiered approach where your highest-risk patients get seen every six months to a year and your lowest-risk patients are followed on a much longer horizon, freeing up capacity precisely for the patients in whom, per this data, skin cancer is nearly a certainty.

That wraps our three articles for this issue. To summarize quickly: hydrocolloid dressing looks like a reasonable, patient-preferred alternative to petroleum ointment with equivalent long-term scar outcomes, though keep an eye on the unproven-but-real trend toward bleeding and dehiscence in higher-risk patients. And the transplant surveillance work makes a solid case that SUNTRAC-based risk stratification, now tested in real-world practice, can meaningfully sharpen where we direct screening resources in this high-risk population without overwhelming the system. Thanks for listening, and we'll see you next month.