UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Modelling disease activity in juvenile dermatomyositis: A Bayesian approach

van Dijkhuizen, EP; Deakin, CT; Wedderburn, LR; De Iorio, M; (2017) Modelling disease activity in juvenile dermatomyositis: A Bayesian approach. Statistical Methods in Medical Research 10.1177/0962280217713233. (In press). Green open access

[thumbnail of van-Dijkhuizen_Modelling_disease_activity_AAM.pdf]
Preview
Text
van-Dijkhuizen_Modelling_disease_activity_AAM.pdf - Accepted Version

Download (236kB) | Preview

Abstract

Juvenile dermatomyositis is the most common form of the juvenile idiopathic inflammatory myopathies characterised by muscle and skin inflammation, leading to symmetric proximal muscle weakness and cutaneous symptoms. It has a fluctuating course and varying prognosis. In a Bayesian framework, we develop a joint model for four longitudinal outcomes, which accounts for within individual variability as well as inter-individual variability. Correlations among the outcome variables are introduced through a subject-specific random effect. Moreover, we exploit an approach similar to a hurdle model to account for excess of a specific outcome in the response. Clinical markers and symptoms are used as covariates in a regression set-up. Data from an ongoing observational cohort study are available, providing information on 340 subjects, who contributed 2725 clinical visits. The model shows good performance and yields efficient estimations of model parameters, as well as accurate predictions of the disease activity parameters, corresponding well to observed clinical patterns over time. The posterior distribution of the by-subject random intercepts shows a substantial correlation between two of the outcome variables. A subset of clinical markers and symptoms are identified as associated with disease activity. These findings have the potential to influence clinical practice as they can be used to stratify patients according to their prognosis and guide treatment decisions, as well as contribute to on-going research about the most relevant outcome markers for patients affected by juvenile dermatomyositis.

Type: Article
Title: Modelling disease activity in juvenile dermatomyositis: A Bayesian approach
Location: England
Open access status: An open access version is available from UCL Discovery
DOI: 10.1177/0962280217713233
Publisher version: https://doi.org/10.1177/0962280217713233
Language: English
Additional information: This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions.
Keywords: Longitudinal Data, Markov chain Monte Carlo, Stochastic Search Variable Selection, mixed model, juvenile dermatomyositis
UCL classification: UCL
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > UCL GOS Institute of Child Health
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > UCL GOS Institute of Child Health > Infection, Immunity and Inflammation Dept
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences > Dept of Statistical Science
URI: https://discovery.ucl.ac.uk/id/eprint/1561247
Downloads since deposit
218Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item