This month’s news from across the specialty

February 1, 2024

In this monthly column, members of the DermWorld Editorial Advisory Workgroup identify exciting news from across the specialty.  

ChatGPT is a readily accessible and widely available artificial intelligence tool mainly used by physicians for administrative tasks and creating educational materials. However, you might want to experiment with it as a clinical decision support tool as well.

An article in JAMA Network Open showed that ChatGPT-4 may improve physicians’ diagnostic accuracy in interpreting test results (2023;6(12):e2347075). In the study, medical cases along with test results were presented to ChatGPT-4 and 550 health care practitioners. The chatbot was more accurate than physicians in assessing the likelihood of diagnoses in all five cases where test results came back negative. When diagnostic tests in the clinical scenarios were positive, ChatGPT was equal to doctors in assessing risk.

The researchers noted that many physicians struggle with estimating pre-test and post-test probabilities of disease, a diagnostic tool known as probabilistic reasoning. Lack of skill in this area can lead to overestimating risk and may lead to over-treatment, additional tests, and over-medicating. However, probability is where chatbots, such as ChatGPT-4, excel and may be helpful as tools in weighing the significance of test results in the context of a clinical exam. The article reveals a future where we have the option to improve medical test interpretation by consulting with chatbots. This article is also exciting as it shows a simple way that dermatologists can experiment with integrating artificial intelligence into clinical practices right now without the need for expensive labs or specialized training.

A recent systematic review highlighted in JAAD identified all prospective randomized clinical trials (RCTs) on PubMed, Embase, and Cochrane Central Register of Controlled Trials between Jan. 1, 2005, and Aug. 26, 2020, that focused on psoriasis, lichen planus, alopecia areata, seborrheic dermatitis, vitiligo, acne/acneiform eruptions, eczema, and atopic dermatitis. A total of 9,518 articles were identified, and 1,504 manuscripts were included in the final analysis. The results demonstrated a slight increase in the reporting of race and ethnicity in dermatologic RCTs. However, most racial minorities such as African Americans and Hispanics continue to be significantly underrepresented. As of 2019, a substantial amount (40.5%) of trials in the United States still did not report the race or ethnicity of the enrolled participants.

Some of the most common reasons for low enrollment of minorities in RCTs include disparities in socioeconomic status, limited access to specialty centers that conduct clinical trials, lack of trust in health care facilities, fear of being exploited, and inability to meet the time or financial demands required of clinical trials.

The January issue of JAAD focused on the issue of underrepresentation in clinical trials, highlighting the success of the 2023 Skin of Color Society Meeting the Challenge Summit. The summit helped secure commitments from the NIH, the FDA, and pharma that appropriate diversity would be required for future clinical trials. JAAD requires authors to consider appropriately diverse populations to be enrolled in clinical trials and make their enrollment plan available to reviewers.

In concordance with the American Heart Association (AHA), the American Dental Association (ADA), and the American Academy of Orthopaedic Surgeons (AAOS), the 2008 JAAD advisory statement on antibiotic prophylaxis in dermatologic surgery definitively recommends prophylaxis only in patients at high risk of infectious endocarditis or hematogenous total joint infection who are having dermatologic surgery: (1) on a site perforating oral mucosa, or (2) on an infected surgical site.

The ADA and AAOS have said that risk/benefit and cost/effectiveness ratios fail to justify routine antibiotic prophylaxis for patients other than those listed above, and that physicians tempted to utilize prophylactic antibiotics in other patients must weigh any perceived benefit against potential side effects and the global development of antibiotic resistance (10.14219/jada.archive.2003.0289). Despite this, two recent surveys of MMS surgeons have revealed that nearly 75% routinely prescribe prophylactic antibiotics to their patients (10.1097/DSS.0000000000002676; 10.36849/JDD.2020.4695).

To investigate the role of perioperative empiric antibiotic therapy (PEAT) on the risk of surgical site infections (SSI) within 30 days of having MMS, one study enrolled 321,863 deidentified MMS patients — split into a cohort who received PEAT on the day of MMS (22%), and those who did not. Of the enrolled patients, 0.12% were diagnosed with SSI within one month of MMS. The risk of SSI was 0.11% in patients who received PEAT and 0.13% in patients who did not. The absolute risk reduction for PEAT in MMS was 0.01%; this was not statistically significant.

As dermatologic surgeons, we must take into consideration the data which support limited benefit of most PEAT. We also must weigh that limited benefit against the risks of the prescribed medications and the risk of burgeoning antibiotic resistance. To assist us in responsible prescribing, we can endeavor to follow specialty-specific guidelines on the topic.

Recently, I had a patient with a germline loss-of-function mutation in the interleukin-36 receptor antagonist gene present to my clinic with a rapidly progressive generalized pustular psoriasis (GPP) flare. He had been doing great for years with an interleukin-17 inhibitor but lost drug access due to his health insurance changing and his prior dermatologist retiring. This unfortunate situation permitted me the opportunity to prescribe spesolimab (Spevigo), an interleukin-36 receptor antagonist that was approved by the FDA in 2022 for GPP flares in adults.

In the phase 2 trial leading to its approval (N Engl J Med. 2021;385:2431-2440), patients with a GPP flare were randomly assigned in a 2:1 ratio to receive a single 900-mg intravenous dose of spesolimab or placebo. Patients in the spesolimab group were significantly more likely to have complete clearance of pustular lesions at one week than those in the placebo group (54% vs. 6%, p<0.001). However, serious adverse events and infections were more common in patients who received spesolimab, including reports of drug reaction with eosinophilia and systemic symptoms and drug-induced hepatic injury. My patient had an immediate response to the drug with rapid clearance of all pustules within days and fortunately no adverse effects.

Additional DermWorld Resources