EVIDENCE-BASED BAYESIAN MODELING TO PREDICT SURVIVAL IN PATIENTS WITH SPORADIC INCLUSION BODY MYOSITIS (SIBM)
Author(s)
Capkun G1, Schmidt J2, Ghosh S3, Sharma H3, De Vera A1, Demin I1, Risson V1, Amzal B4, Wiecek W5, Obadia T4
1Novartis Pharma AG, Basel, Switzerland, 2Department of Neurology, University Medical Center Göttingen, Göttingen, Germany, 3Novartis, Hyderabad, India, 4LASER ANALYTICA, London, UK, 5LASER Analytica, London, UK
OBJECTIVES: sIBM is a rare muscle disease characterized by dysphagia, atrophy and weakness of proximal and distal muscles, typically manifesting in ≥ 50s. Limited data has been published on the disease and how mortality is associated with disease progression and severity. Our objective was to predict mortality in sIBM patients based on six disease-related outcomes that may influence patient’s life span. In addition, our aim is to select the model providing best prediction of mortality against existing literature. METHODS: Literature review provided data on association between outcomes related to manifestation of sIBM and mortality in elderly people and patients with neurodegenerative diseases. Demographic and disease-related outcomes data from sIBM patients were collected at annual meetings of the Myositis Association. Gender- and age-adjusted survival curves from WHO served as reference survival rate. Simulation of mortality in 2000 virtual cohorts was performed using a Weibull parametric model with combinations of up to six disease-related outcomes potentially accounting for additional mortality risk in sIBM patients: 6-minute walking distance (6MWD), BMI, dysphagia, falling, aspiration pneumonia, being wheelchair-bound. Results were compared with the mortality rates observed in two published sIBM cohorts, adjusted for demographics. RESULTS: The simulation could exactly replicate the observed 12-year mortality rate of 72% from the multi-country study published by Cox et al., when using dysphagia and 6-minute-walking-distance as predictors. For the second publication of the sIBM cohort in France by Benveniste et al., the observed 2-year mortality rate (18%) was comparable to the % baseline mortality without any predictors. CONCLUSIONS: Our results suggest that long-term survival in sIBM patients can be predicted with age, gender, 6MWD and dysphagia as predictors, though with a possible country effect. Further validation of our findings is required through natural history studies in sIBM , especially regarding the impact of the disease on the short-term survival.
Conference/Value in Health Info
2017-05, ISPOR 2017, Boston, MA, USA
Value in Health, Vol. 20, No. 5 (May 2017)
Code
PSY29
Topic
Clinical Outcomes
Topic Subcategory
Relating Intermediate to Long-term Outcomes
Disease
Rare and Orphan Diseases, Systemic Disorders/Conditions