Neurology

Stepped-wedge trial: teleneurology altered referrals but did not reduce secondary-care use

In 983 adults seen in 41 rural German GP practices, teleconsultations reduced management solely in general practice (38.3% vs 50.7%). Referrals to neurologists fell, but referrals to other specialties (11.9% vs 4.7%) and hospitals (12.8% vs 2.6%) increased. No differences were observed in quality of life or health status at three months follow-up. Unblinded GPs and lower adherence to recommendations during intervention limit interpretation. 1

Systematic review: intranasal stem cells in humans—feasible and safe but biased evidence

Nineteen records (7 published studies; 104 total participants reported) examined intranasal stem cells for neurological and respiratory conditions. Reported outcomes suggest feasibility and acceptable safety, but no study achieved low overall risk of bias. Small samples, heterogeneous conditions and products, and confirmed publication bias preclude efficacy conclusions. Authors conclude that adequately powered, quadruple-blinded, placebo-controlled RCTs are required before clinical translation. 2

Scoping review: AI for gait and balance shows prognostic promise but lacks validation

Eighteen studies (2009–2025) applied AI to gait and balance in stroke, Parkinson disease, and MS, mostly for prognostic purposes (72%). Machine learning dominated methods, but none performed prospective or external validation on independent datasets. Studies were heterogeneous and concentrated in stroke populations, limiting generalizability. The field is promising but currently methodologically immature for routine clinical use. 3