Citations

ABSTRACT As artificial intelligence (AI) continues to rapidly reshape health care, there is a critical need for clear frameworks for clinicians to ensure ethical, equitable, and effective integration and use of AI in patient care. Key integrations of AI include enhancing health communications, patient engagement, clinicians' training, pharmaceutical advertising, clinical decision‐making, and automation of clinical operations and workflow. However, there are growing concerns related to regulatory gaps, the spread of misinformation, security threats, patient and data privacy leaks, and widening health disparities gaps. These concerns are exacerbated by limited institutional infrastructure and limited AI literacy of clinicians and patients. Recent policy developments reflect efforts to guide responsible AI development and use. While progress has been made, the lack of standardized human oversight remains a critical gap, particularly as policies may not fully consider the challenges and complexities at institutional, societal, technical, and individual levels. Thus, herein, the Drug Information Practice and Research Network (DI PRN) of the American College of Clinical Pharmacy (ACCP) : (1) explores current and emerging multifaceted challenges, a call to action, and opportunities of AI integration in health care including examination of the regulatory, ethical, operational, and health equity implications; and (2) provides practical recommendations for responsible use of AI through DI PRN‐developed example case‐based approaches and best practices infographics to enhance AI literacy for diverse learners including clinicians, trainees, and patients. This DI PRN opinion paper highlights the importance of proactive governance frameworks and equips and empowers diverse learners with practical AI literacy tools to confidently engage with AI technologies in patient care.

Abstract Aims Endogenous endophthalmitis (EE) is a rarely reported complication of infective endocarditis (IE). In an international series, we sought to determine the clinical and microbiological profile, treatment, and outcome of patients presenting with IE-related EE. Methods and results Cases recorded from 2014 to 2023 in nine centres in Europe and the United States were collected. Results were compared to a matched control group. Conclusion Sixty-six patients with EE were reported, mean age of 65.2 ± 14.9 years, 71% (n = 47) male. Blood cultures were positive in 97% (64 cases) of patients, with a predominance of streptococci (46%, n = 30). As compared with the control group (n = 264), the EE group presented with more frequent diabetes (35% vs. 21%, P = 0.02), history of cirrhosis (9% vs. 3%, P = 0.04), glomerulonephritis (15% vs. 0.4%, P < 0.001), embolism before admission (92% vs. 55%, P < 0.001), and Janeway lesions (9% vs. 1%, P = 0.002). Streptococcal infection (46% vs. 26%, P = 0.001) was more frequent and Enterococcal infection (0% vs. 18%, P < 0.001) less frequent in the EE group. The main ocular symptoms were a decrease in visual acuity (96%), red eye (55%), and ocular pain (55%). Treatment of EE consisted of intravitreal antibiotic injection in 55 (83%) patients and vitrectomy in 17 (26%). Improvement of visual acuity was observed in 36 (55%) patients. Conclusion EE is a serious complication of IE with severe residual vision impairment. Patients with IE should be evaluated for ocular complications, since early detection of EE is crucial to prevent delays in management and to preserve visual function.

Background/Objectives: Cesarean section (C-section) is a common surgical procedure associated with an increased risk of 30-day postpartum hospital readmissions. This study utilized machine learning (ML) to predict readmissions using a nationwide database. Methods: A retrospective analysis of the National Surgical Quality Improvement Project (2012–2022) included 54,593 patients who underwent C-sections. Random Forests (RF) and Extreme Gradient Boosting (XGBoost) models were developed and compared to logistic regression (LR) using demographic, preoperative, and perioperative data. Results: Of the cohort, 1306 (2.39%) patients were readmitted. Readmitted patients had higher rates of being of African American race (17.99% vs. 9.83%), diabetes (11.03% vs. 8.19%), and hypertension (11.49% vs. 4.68%) (p < 0.001). RF achieved the highest performance (AUC = 0.737, sensitivity = 72.03%, specificity: 61.33%), and a preoperative-only RF model achieved a sensitivity of 83.14%. Key predictors included age, BMI, operative time, white blood cell count, and hematocrit. Conclusions: ML effectively predicts C-section readmissions, supporting early identification and interventions to improve patient outcomes and reduce healthcare costs.

Objectives: Current standard of care after a First-Time patellar instability event is non-operative management, with exceptions for patients with osteochondral fracture or loose body requiring immediate surgical intervention, patients with significant pathoanatomy, or contralateral patellar instability. For recurrent patellar instability, there is controversy as to when isolated medial patellofemoral ligament reconstruction (MPFL-R) is enough. The purpose of this study is to determine (1) the comparative outcomes of isolated MPFL-R after first-time patellar instability (FTPI) and recurrent patellar instability (RPI) and (2) the impact of osteochondral injury on outcomes after surgical treatment of FTPI. Methods: A prospective, multicenter cohort study (JUPITER: Justifying Patellar Instability Treatment by Results) database was queried for patients who underwent patellar stabilization surgery (PSS) between December 2016 and September 2022. Patients were included if they underwent a medial patellofemoral ligament reconstruction (MPFL-R) for FTPI or RPI. Those with less than 2 years of clinical follow-up, revision PSS, or concomitant bony procedures at the time of MPFL-R were excluded. Independent t-tests were performed to compare means for continuous variables, and chi-squared tests were performed to compare frequency for categorical variables, between groups. Results: A total of 739 patients met inclusion criteria with a mean age of 15.6 ± 3.3 years, and 59.9% female sex. FTPI occurred in 216 (27.2%) patients, and of these 104 (48.1%) sustained an operative chondral or osteochondral lesion. Those with FTPI had a significantly greater trochlear depth (2.9 mm vs. 2.6 mm; P = .048), but a significantly lower rate of pathological trochlear bump > 5 mm (52.3% vs. 67.8%; P = .008) than those with RPI (Table 1). Among the FTPI cohort, those with treatment of a chondral lesion were less likely to have patella alta (Caton-Deschamps Index > 1.3) (14.9% vs. 41.3%, P = .006), and less frequently had a mild or severe J-sign (25.0% vs. 42.7%, p = .016) than those without a chondral intervention (Table 2). Postoperatively, there were no differences with rates of recurrent instability (8.3% vs. 12.3%; P = .130) or return to sport (90.4% vs. 94.5%; P = .182) between FTPI and RPI, respectively (Table 3). There were no differences with rates of recurrent instability (7.6% vs. 9.1%; P = .807) or return to sport (89.1% vs. 91.7%; P = .755) between those with or without a chondral intervention, respectively (Table 4). Conclusions: This study demonstrates no difference in rates of recurrent post-operative instability between FTPI and RPI with rates consistent with the current literature. MPFL reconstruction remains as a durable surgical treatment option for patients presenting with either FTPI or RPI with relatively low rates of recurrent instability. In this patient population, patella alta was found to be protective of osteochondral injury which is also consistent with existing literature. Continued, long-term investigation is needed to understand outcomes for First-Time patellar instability and to better define the indications for surgical interventions in this population.

BACKGROUND Lennox-Gastaut syndrome (LGS) is a severe developmental and epileptic encephalopathy characterized by multiple seizure types, intellectual disability, and distinctive EEG findings. Deep brain stimulation of the centromedian nucleus (CM-DBS) is an emerging therapy for LGS, but pediatric experience remains limited. OBSERVATIONS The authors report a single-center experience with CM-DBS in 6 children and adolescents (ages 12–18 years) with LGS. One patient experienced infection-related device removal after 2 months. The remaining 5 received sustained stimulation for 2.5–5 years. All 5 showed caregiver-reported improvements in seizure burden and alertness, with corroborating Clinical Global Impression–Improvement scores. Three patients exhibited marked seizure reduction following targeted reprogramming toward a previously identified optimal target of stimulation in the anterolateral CM. Pre- and post-DBS scalp EEG recordings were available in 1 patient and showed a reduced burden of interictal discharges. LESSONS CM-DBS is a promising treatment for pediatric, medically refractory LGS when resective approaches are unsuitable. Benefits were sustained over years. Lead localization and direction of stimulation appear important to optimize clinical benefit. These findings support the feasibility and safety of pediatric CM-DBS and highlight the need for prospective trials incorporating EEG-based outcomes and patient-centered measures including comorbidities and quality of life. Early, network-targeted neuromodulation may improve long-term outcomes. https://thejns.org/doi/10.3171/CASE25478