Characteristics and early diagnosis of Motor Neuron Disease and associated COVID-19 outcomes in 67 million individuals in England: a comparative study of phenotyping approaches using AI and clinical guidelines

Publication date: Jul 01, 2025

Early diagnosis of motor neuron disease (MND) is key for improving outcomes but remains challenging despite the existence of clinical guidance. The COVID-19 pandemic further exposed gaps in timely diagnosis and highlighted increased mortality risk and concerns around vaccine effectiveness in this vulnerable group. Yet, population-level evidence to inform or improve diagnostic strategies is lacking. In this study, we analysed linked health data from 67 million individuals in England to compare clinical guideilnes and AI-derived phenotyping approaches for early MND detection, and to assess COVID-19-related mortality and vaccination effects in people with MND. We used nationwide linked health records from 67 million individuals in England to identify MND cohorts during 2014-2019 and 2020-2021, estimating period prevalence and patient characteristics. Four phenotyping approaches were implemented: the MND Association Red Flag List and three AI-derived models-knowledge graph, GPT-4, and real-world data-driven machine learning. These were used to develop models for diagnosing MND and predicting onset 1, 3, and 5 years prior to coded diagnosis. Prediction methods included logistic regression, random forest, support vector machines, and recurrent neural networks. Model performance was evaluated using AUROC, F1 score, sensitivity, specificity, and PPV. COVID-19-related mortality among individuals with MND was assessed using Kaplan-Meier survival analysis. Of 67,270,015 individuals, from 1st January 2014 to 31st December 2019, we identified 12,240 people with coded MND diagnosis. For MND screening, the MNDA guideline, though designed for diagnosis, showed poor to moderate discrimination (AUROC: 0.62-0.63) while combining the guideline with AI derived ones (the ensemble) could improve it significantly (0.66-0.68). For the diagnosing task, the guideline had good discrimination (0.70), but lower overall performance (F1 score: 0.44) and the real-world data driven approach (hypothesis free) achieved much better results (AUROC: 0.78; F1: 0.55). As for COVID-19 mortality risk, compared to matched controls, MND patients had elevated risk (HR=2.97, 95% CI: 1.97-4.48) in wave 1, while fully vaccinated individuals in wave 2 demonstrated non-statistically significant higher risk (HR=1.27, 95% CI: 0.69-2.34). This population scale study showed the MNDA guideline did not show very effective power in either screening or diagnosing MND probably due to the missing of predictive phenotypes available in routine care. The hypothesis-free phenotyping approach, applying AI on real-world datasets for deriving predictive phenotypes, demonstrated a great utility by identifying 13 novel phenotypes from 7 ICD-10 chapters that had significant effects in predicting MND. Our findings indicate that while MND patients face substantially elevated COVID-19 mortality risk, full vaccination appears to effectively reduce this excess risk to levels comparable with the general population

PDF

Semantics

Download Document