Stimpson, Georgia; (2024) Incorporating data insights and clinical knowledge in statistical methodologies: Modelling disease trajectories in children with Duchenne Muscular Dystrophy and Spinal Muscular Atrophy. Doctoral thesis (Ph.D), UCL (University College London).

Text Thesis Final anon.pdf - Accepted Version Access restricted to UCL open access staff until 1 October 2025. Download (103MB)

Abstract

Duchenne Muscular Dystrophy (DMD) and Spinal Muscular Atrophy (SMA) are two heteroge- neous, rare, neuromuscular conditions affecting children. In these diseases, observational datasets containing high-dimensional longitudinal data exist, however, informative visiting and structural missingness processes can limit their utility. For my PhD, I have also developed centiles-for-age for several of the motor function outcomes in DMD, using methods adapted from the growth lit- erature. Additionally, I have focused on using structured modelling techniques to understand the highly heterogenous rates of disease progression. For DMD patients this has included analysing the previously unexplored and complex relationship between steroids, which delay loss of ambula- tion, their growth-related side effects, and loss of ambulation. I used a structured joint modelling approach to model both longitudinal outcomes (growth) and a survival outcome (age at loss of ambulation). In untreated SMA patients, I have investigated the Revised Hammersmith Scale (RHS) and the utility of untreated data given the changing phenotypes of SMA. The final project is based around identifying predictors of sitting in Nusinersen-treated SMA patients. Using sur- vival tree-derived methodologies, I have identified key factors at baseline which define subgroups of patients that are more likely to achieve sitting. However, due to the nature of systematic data collection in rare diseases, and in SMA in general, I have also investigated methods to understand the structured missingness in the data. As the therapies for SMA are still evolving, I hope that this research can be used as a foundation to build a treatment decision algorithm, identifying patients who are most likely to sit or who are responding well to treatment, so that they can be offered the most effective treatment.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item