Israel Medical Association Journal

Background: Identifying drug–drug interactions (DDIs) in dermatology can be cumbersome and time-consuming using traditional methods.

Objectives: To explore the potential of ChatGPT-4o, an artificial intelligence (AI)-based chatbot, to streamline the identification process.

Methods: ChatGPT-4o was tasked with assessing DDIs among commonly prescribed dermatological medications. The accuracy and reliability of the chatbot's outputs were compared against established references for 43 interactions.

Results: ChatGPT-4o successfully identified all evaluated interactions. It accurately described the interaction effects in 42 cases, with only one example of misdescription.

Conclusions: The findings highlight the potential of ChatGPT to serve as an effective and efficient alternative for identifying and understanding DDIs in dermatology, despite one noted error that emphasizes the need for ongoing verification against trusted references. Further research is needed to validate its use across a broader range of medications and clinical scenarios.

Background: Biological agents for anti-tumor necrosis factor-α therapy have revolutionized treatments for autoimmune diseases; however, approximately 20% of rheumatology and 40% of gastroenterology patients do not respond to the therapy, or they show reduced drug efficacy because of anti-drug antibody (ADA) formation.

Objectives: To evaluate laboratory tools for individual monitoring of infliximab therapy and the relationship between ADA and infliximab serum levels, ADA and clinical response, and ADA and autoantibodies.

Methods: Our study comprised patients treated with infliximab and affected by selected rheumatology and gastroenterology diseases. Sera were analyzed for infliximab, total-anti-drug antibodies (Total-ADA), and free-anti-drug antibodies (Free-ADA) serum levels and for the detection of specific autoantibodies.

Results: We analyzed 73 patients. Total-ADA were detected in 26 rheumatology and 21 gastroenterology patients. Serum infliximab levels were significantly lower in Total-ADA positive patients (P = 0.01 for rheumatology group, P = 0.02 for gastroenterology group). A lack of response was observed in 7 rheumatology and 15 gastroenterology samples. Total-ADA serum levels were statistically significantly higher in patients with treatment failure in both groups (P = 0.01 and P = 0.001, respectively). There was no significant association between the presence of Total-ADA and other autoantibodies. Free-ADA were detected in only 27 rheumatology patients. Results showed a significant correlation with clinical outcome (P = 0.006).

Conclusions: The correlation with clinical response suggests that the presence of ADA could interfere with efficacy of therapy. The tests for monitoring therapy may be an important tool to assist clinicians in early detection and prevention of therapy failure.

Objective: To assess the effect of establishing a new dedicated acute stroke unit within a department of neurology on indicators of process of care and outcome of acute stroke in a routine clinical setting in Israel.

Methods: Stroke patients admitted to the Sheba Medical Center during the period March 2001 to June 2002 were included in a prospective study according to selection criteria. Data on demographics, risk factors, co-morbidities and stroke severity were collected. Indicators of process of care and outcome were assessed at hospital discharge and 30 days follow-up. Comparison between outcome variables by hospitalization ward was conducted using logistic regression analysis adjusting for confounders.

Results: Of 616 acute stroke patients (mean age 70 years, 61% men, 84% ischemic stroke), 353 (57%) were admitted to general wards and 263 (43%) to the stroke unit. Diagnostic procedures were performed more often, and infection rate was lower in the setting of the stroke unit. Poor outcome (modified Rankin scale ≥3 or death) was present less often in patients managed in the stroke unit both at hospital discharge (adjusted odds ratio 0.5, 95% confidence interval 0.3–0.8) and at 30 day follow-up (adjusted OR[1] 0.6, 95%CI[2] 0.3–0.9). A Functional Independence Measure score ≤90 or death at 30 day follow-up was less frequent among patients managed in the stroke unit than in general wards (adjusted OR 0.5, 95%CI 0.2–0.8).

Conclusions: Improved outcomes and higher adherence to guidelines were observed in patients treated in a stroke unit within a department of neurology. The results suggest that patients with acute stroke should have access to treatment in a dedicated stroke unit.