Community
Hackathons
Spring Hackathon: Making Replications Count
4th - 6th May 2026.
Organised by FORRT (Framework for Open and Reproducible Research Training) (GitHub). I was part of a team developing a tool to automatically extract data from published replication studies. This will support and expand the FORRT Library of Replication Attempts (FLoRA), improving access to structured evidence on replication.
Clubs
NHS model reuse club
Coming soon!
I will be helping to set up and support this club, with a focus on sharing best practices and encouraging reuse of simulation models and analytical code across NHS and academic settings.
University of Exeter St Luke’s coding club
March - July 2025.
I supported sessions at the coding club. This included helping participants during activities, and coordinating practical aspects such as organising food when the main organiser was unavailable.
Conferences
RSECon26
9th - 11th September 2026.
The UK’s annual conference for the Research Software Engineering community, focused on reproducible, sustainable, and high-quality research software.
South West Analytics & Infrastructure in Healthcare (SWAIH) 2026
8th July 2026.
A regional conference bringing together healthcare analysts, data scientists, and infrastructure specialists to share applied work across NHS and academia.
RPySoc 2025
13th - 14th November 2025.
The NHS-R / NHS.pycom open-source conference, centred on Python, R, and collaborative data science in healthcare.
Simulation Workshop 2025 (SW25)
31st March - 2nd April 2025.
Conference focused on simulation modelling methods and applications across industry, defence, and healthcare.
European Health Economics Association (EuHEA)
5th - 8th July 2022.
An international conference for health economists to present research on economic evaluation, policy, and health system performance.
I shared two presentations at this conference:
Objectives The National Institute of Health and Care Excellence, which gives guidance on health/social care interventions to be provided by the UK National Health Service, has recently stated that preference-based-measures (PBMs) of wellbeing may be used in cost-effectiveness analyses, in addition to generic and condition-specific (CS) PBMs of health-related quality of life. We aimed to assess the comparative responsiveness of wellbeing, generic and CS PBMs in the context of multiple sclerosis (MS).
Methods The MS Register (MSR) is a cohort, representative of the UK population of people with MS. In September 2019, March 2020 and September 2020, MSR respondents completed wellbeing PBMs– the ICEpop CAPability measure for Adults (ICECAP-A) and Adult Social Care Outcomes Toolkit (ASCOT); CS PBMs– the MS Impact Scale Eight Dimensions (MSIS-8D) and MSIS-8D-Patient (MSIS-8D-P); and a generic PBM- the EQ-5D-3L. They also completed the Fatigue Severity Scale, MS Walking Scale-12, Hospital Anxiety and Depression Scale, and a bespoke measure of ‘illness events’ that may have occurred between time-points e.g. a relapse, a new symptom, leaving employment due to MS. Responsiveness of the PBMs was compared in relation to changes in fatigue, mobility, anxiety and depression between time-points, and by the occurrence of illness events. Standardised response means and Cohen’s D were calculated, and t-tests used to assess change. Multivariable pooled OLS regression analyses were undertaken for each PBM with standard errors clustered by respondent. Variables were standardised to enable comparison.
Results 1,742 individuals gave responses for at least two consecutive time-points. Overall, CS PBMs were the most responsive, followed by ICECAP-A, ASCOT and lastly the EQ-5D. For example, coefficients (95% CIs) in relation to changes in anxiety were: MSIS-8D -0.253 (-0.304,-0.202); MSIS-8D-P -0.248 (-0.300,-0.197); ICECAP-A -0.195 (-0.243,-0.147); ASCOT -0.170 (-0.217,-0.122); EQ-5D -0.162 (-0.219,-0.105). The PBMs were more responsive to changes in anxiety and depression than to changes in mobility or fatigue. They showed little responsiveness, and no consistent pattern, in relation to illness events.
Discussion Our results support the use of CS and wellbeing PBMs alongside the EQ-5D, to maximise the potential of detecting treatment benefits in cost-effectiveness analyses of treatments for MS. The lack of responsiveness to illness events may reflect limitations in the content or design of the illness events measure, however this was informed by a systematic search of qualitative literature and devised with a MS patient involvement group. Alternatively, the finding may be indicative of limited responsiveness of the PBMs to key illness events in the lives of people with MS.
Objectives The EQ-5D is a generic, preference-based-measure (PBM) of health-related quality of life, recommended by the National Institute of Health and Care Excellence in the UK as a source of health state values (HSVs) for the calculation of quality-adjusted life-years (QALYs). Concerns have been raised, however, about the sensitivity of the EQ-5D to differences and changes in the health-related quality of life of people with multiple sclerosis (MS). This has led to the development of MS-specific PBMs, such as the MS Impact Scale Eight Dimensions (MSIS-8D) and the MSIS-8D-Patient (MSIS-8D-P), which are based on public and patient preferences respectively. We aimed to compare the discriminative validity of the EQ-5D, MSIS-8D and MSIS-8D-P across a range of demographic and clinical characteristics, using data from a large UK cohort of people with MS.
Methods UK MS Register responses from 2011 to 2019 were analysed descriptively and using multivariable linear regression to investigate the strength of independent relationships between demographic/clinical characteristics and HSVs on each of the PBMs. Robust regression was used as residuals did not meet assumptions of normality/homoscedasticity. Continuous variables were centred.
Results There were 14,385 respondents. Mean age was 55.3 (11.4) years and 72.8% were female. Mean disease severity score on the Expanded Disability Status Scale (EDSS) was 5.1 (2.0). Mean EQ-5D, MSIS-8D and MSIS-8D-P values were 0.562 (0.308), 0.603 (0.18) and 0.643 (0.173) respectively. Compared to EQ-5D values, mean MSIS-8D and MSIS-8D-P values were generally higher and their absolute differences between groups by demographic/clinical characteristics were smaller. Regression analysis indicated that HSVs were significantly negatively associated with disease severity (EDSS) (p<0.001) and significantly positively associated with age (p<0.001) on all three measures.
Discussion This study provides a new source of generic and MS-specific HSVs, based on a recent, representative sample of people with MS in the UK, which can be used in economic evaluations. Many cost effectiveness analyses of treatments for MS still rely on a source of EQ-5D values collected in 2005. The HSVs reported here will better reflect current treatment practices, which have altered the profile of the UK MS population. The smaller differences between subgroups of people with MS that were identified by the MS-specific PBMs are likely due to the narrower range of HSVs for the MSIS-8D (0.079 to 0.882) and MSIS-8D-P (0.138 to 0.893), compared to the EQ-5D (-0.594 to 1). This is a common characteristic of PBMs other than the EQ-5D, and is an important factor to consider when comparing cost-effectiveness results based on different sources of HSVs.
