Explore the vast range of titles available.

Large group collaborative teaching approaches are rapidly gaining popularity in undergraduate medical education. The case-based collaborative Learning (CBCL) pedagogy was instituted for pre-clerkship teaching at Harvard Medical School in 2015 with subsequent implementation at other medical schools. CBCL emphasizes inductive reasoning, integrates basic and clinical sciences, stimulates curiosity, and fosters teamwork. Given the ongoing educational evolution, guidance on designing and facilitating collaborative learning sessions, such as CBCL may benefit faculty in their instructional design efforts. This perspective article describes strategies to create effective collaborative sessions using CBCL as an example. We reviewed the literature and summarized ten years of experience in CBCL teaching through the lens of contemporary theories of teaching and learning. The recommendations are organized into three main domains: Instructional Design, Facilitation, and Professional Transformation, each aligned with the theoretical principles of CBCL. The recommendations provide a conceptual model to assist faculty in designing engaging and effective class materials and support students’ professional transformation during collaborative learning sessions.

Sexual dysfunction is common in men and women with neurological diseases. Medications used in neurology can cause sexual dysfunction independently of the disease process and this may adversely affect patients’ quality of life. This review focuses on medications commonly prescribed to neurological patients that may contribute to altered sexual function, and discusses how they may differ in men and women.

Symptoms related to bladder, bowel and sexual dysfunction are common in patients with neurological diseases and are associated with significant morbidity and impacts on quality of life. However, compared with other neurological problems, they are often less visible, under-ascertained and under-treated. This review aims to provide neurologists with an evidence-based, structured approach to evaluating bladder, bowel and sexual dysfunction in patients with neurological disorders. As with any other presenting concern, this requires a structured approach to history, examination and the judicious use of investigations.

Herpes simplex virus (HSV) encephalitis can induce an autoimmune encephalitis mediated by autoantibodies against the N-methyl-D-aspartate receptor (NMDAR). Post-HSV NMDAR encephalitis and de novo NMDAR encephalitis have been more commonly described in children and young adults. We describe the case of a 67-year-old woman with post-HSV NMDAR encephalitis and review the relevant literature. Clinical, serological, neurophysiological, and imaging evaluations were undertaken in the evaluation of this patient. A literature review was performed. Nearly 2 months after a typical course of HSV encephalitis confirmed by HSV polymerase chain reaction studies from the spinal fluid and treated with intravenous acyclovir, a 67-year-old woman suffered neurological deterioration. There was no evidence of active HSV infection, but NMDAR antibodies were found in her serum and spinal fluid. The patient improved after initiation of immunosuppressive therapy. All patients who experience new or recurrent neurological symptoms following recovery from HSV encephalitis should be evaluated for post-infectious autoimmune encephalitis, including NMDAR encephalitis.

Background Bright white light therapy (LT) can improve fatigue in several disease states but has not been studied in multiple sclerosis (MS). Objective To determine whether controlled home-based LT is feasible, tolerable, and well-adhered to in MS-associated fatigue. Methods A randomized, controlled trial of twice-daily 1-h bright white LT (BWLT) (10,000 lx, active arm) versus dim red LT (DRLT) (< 300 lx, control arm) was performed. Adults with MS-associated fatigue were enrolled for 10 weeks: 2-week baseline, 4-week intervention, 4-week washout. Results 41 participants were enrolled; 35 were randomized (average age 42 years, 80% female; BWLT n  = 20; DRLT n  = 15). 31 were in the intention to treat analysis. The average duration of LT sessions was similar between groups (BWLT 60.9 min, DRLT 61.5 min, p  = 0.70). The most commonly reported adverse event was headache. There were no events that led to discontinuation. Baseline fatigue was severe in both arms (each 53/63 points on the Fatigue Severity Scale (FSS), p  = 0.92). FSS was lower following BWLT (FSS 45.8 post-LT, p  = 0.04; 44.9 post-washout, p  = 0.02 intra-group compared to baseline FSS) and DRLT (FSS 46.7 post-LT, p  = 0.03; 43.9 post-washout, p  = 0.002 intragroup compared to baseline FSS). There was no difference between BWLT and DRLT groups in the magnitude of reduction of FSS scores ( p  = 0.81 after LT; p  = 0.77 after washout for between group comparisons). Similarly, MS quality of life metrics improved in both arms but were not significantly different between groups after LT ( p  = 0.22) or washout. Conclusions LT is safe, feasible, and well-tolerated in people with MS-associated fatigue. Improvement in both light spectra likely indicates a strong placebo effect for the DRLT group.

The most common systemic infective cause reported in association with RESLES is influenza, but other viruses (rotavirus, measles, human herpesvirus 6, Epstein-Barr virus, varicella zoster virus, mumps and adenovirus) and bacteria (Salmonella enteritidis, Escherichia coli and Legionella pneumophila) have also been reported. 1 Neurologists should be aware of the differential diagnosis of symmetrical, diffusion-restricting splenial lesions in order to search for an underlying cause, or, if the cause is in the past, to reassure the patient and other providers that the lesion is usually benign and reversible.

Background Basilar artery occlusion can cause locked-in syndrome, which is characterized by quadriplegia, anarthria, and limited communication via eye movements. Here, we describe an uncommon stroke syndrome associated with endovascular recanalization of the top of the basilar artery: “reverse locked-in syndrome.” Methods We report the case of a patient with atypical neurological deficits caused by acute ischemic stroke of the midbrain tegmentum. We perform neuroanatomic localization of the patient’s infarcts by mapping the magnetic resonance imaging (MRI) data onto a brainstem atlas. Results A 61-year-old man presented with acute coma and quadriplegia due to top of the basilar artery occlusion. He underwent emergent endovascular thrombectomy, with successful recanalization of the basilar artery at 4 h and 43 min post-ictus. The patient regained consciousness and purposeful movement in all four extremities, but the post-procedure neurological examination demonstrated bilateral ptosis with complete pupillary and oculomotor paralysis. MRI revealed infarction of the bilateral oculomotor nuclei in the midbrain tegmentum. At 9-month follow-up, he had anisocoria and dysconjugate gaze, but was living at home and required minimal assistance in performing all activities of daily living. Conclusions Since the patient’s deficits were the exact opposite of those described in locked-in syndrome, we propose the term “reverse locked-in syndrome” to describe this neurological entity characterized by bilateral ptosis, non-reactive pupils, and ophthalmoplegia with preservation of consciousness and extremity motor function.

Large language models (LLMs) are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm Assessment for Risk in Medicine), a benchmark using 100 real primary care-to-specialist consultation cases to measure frequency and severity of harm from LLM-generated medical recommendations. NOHARM covers 10 specialties, with 12,747 expert annotations for 4,249 clinical management options. Across 31 LLMs, potential for severe harm from LLM recommendations occurs in up to 22.2% (95% CI 21.6-22.8%) of cases, with harm of omission accounting for 76.6% (95% CI 76.4-76.8%) of errors. Safety performance is only moderately correlated (r = 0.61-0.64) with existing AI and medical knowledge benchmarks. The best models outperform generalist physicians on safety (mean difference 9.7%, 95% CI 7.0-12.5%), and a diverse multi-agent approach improves safety compared to solo models (mean difference 8.0%, 95% CI 4.0-12.1%). Therefore, despite strong performance on existing evaluations, widely used AI models can produce severely harmful medical advice at nontrivial rates, underscoring clinical safety as a distinct performance dimension necessitating explicit measurement.

Large language models (LLMs) and medical AI tools are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm Assessment for Risk in Medicine), a 1,100-task benchmark of primary care-to-specialist consultation cases to measure the frequency and severity of potentially harmful errors from LLM-generated medical consultation recommendations. NOHARM covers 10 specialties, with 12,747 expert annotations for 4,249 clinical management options. Across 20 notable LLMs and 4 widely used retrieval-augmented generation (RAG) clinical AI tools, direct application of recommendations carried potential for severe harm in up to 24.6% of cases, with errors of omission accounting for more than 80% of severe errors. Harm potential was not uniform across systems, with clinical AI tools outperforming generalist LLMs, and multi-agent AI teaming further improving performance in generalist models. In a randomized study of 101 U.S.-licensed generalist physicians, AI assistance improved physician performance compared to conventional resources. However, AI-assisted physicians frequently omitted valuable AI-generated recommendations and still scored lower than many AI systems alone. Had those recommendations been incorporated, combined human-AI responses would have outperformed both the human and AI system as used, suggesting complementary strengths and unrealized potential in human-AI teaming. Collectively, these results show that despite strong performance on medical knowledge benchmarks, widely used AI tools can produce medical consultation advice with the potential for severe harm, and highlight the need for explicit measurement of clinical safety. The benchmark and leaderboard are publicly available to support ongoing evaluation and improvement of AI systems used for clinical care.

Large language models (LLMs) are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm Assessment for Risk in Medicine), a benchmark using 100 real primary care-to-specialist consultation cases to measure frequency and severity of harm from LLM-generated medical recommendations. NOHARM covers 10 specialties, with 12,747 expert annotations for 4,249 clinical management options. Across 31 LLMs, potential for severe harm from LLM recommendations occurs in up to 22.2% (95% CI 21.6-22.8%) of cases, with harm of omission accounting for 76.6% (95% CI 76.4-76.8%) of errors. Safety performance is only moderately correlated (r = 0.61-0.64) with existing AI and medical knowledge benchmarks. The best models outperform generalist physicians on safety (mean difference 9.7%, 95% CI 7.0-12.5%), and a diverse multi-agent approach improves safety compared to solo models (mean difference 8.0%, 95% CI 4.0-12.1%). Therefore, despite strong performance on existing evaluations, widely used AI models can produce severely harmful medical advice at nontrivial rates, underscoring clinical safety as a distinct performance dimension necessitating explicit measurement.