Novel statistics to handle rare diseases and small sample sizes using Bayesian techniques: Application to Multiple Sclerosis (MS) and Systemic Lupus Erythematosus (SLE) in pregnancy

Study type

• Study type:
  Non-interventional study
• Scope of the study:
  Assessment of risk minimisation measure implementation or effectiveness

  Disease epidemiology

  Drug utilisation

Non-interventional study

• Non-interventional study design:
  Case-control

  Cohort

Study drug and medical condition

• Medical condition to be studied:
  Multiple sclerosis

Population studied

• Age groups:
  Adolescents (12 to < 18 years)

  Adults (18 to < 46 years)

  Adults (46 to < 65 years)
• Special population of interest:
  Pregnant women
• Estimated number of subjects:
  5989000

Study design details

• Main study objective:
  The ultimate aim of the study is to use novel methods of analysis concerning the synthesis of information from disparate data sources to provide data about medications that are used by women suffering from rare diseases such as SLE or MS, on their safety in pregnancy to the mother and child.
• Outcomes:
  • Major congenital malformations -, Maternal : - Preeclampsia -Gestational diabetes Pregnancy : • Still birth • Spontaneous abortions • Live births • Elective terminations • Ectopic pregnancy Birth/Neonatal : • SGA • LGA • Malformations split by organ system • Pre-term birth • Neonatal stroke • Neonatal death • Neonatal infections Children : - Death - Infections
• Data analysis plan:
  -Algorithm Identification: Validity judged by consistency of estimated prevalence rates within registries across time and by comparison with published prevalence rates. -Utilisation study: Prevalence of MS and SLE drug utilisation compared between pregnant women and matched non-pregnant women using conditional logistic regression. -Safety study : Medication safety analysed by comparing women with SLE or MS who are exposed to a specific medication with women with SLE or MS who are unexposed . In addition, comparaison to women without either MS or SLE to estimate the association of each disease with adverse pregnancy outcomes. Unadjusted logistic regression models, then a range of adjusted models will be run to account for confounders and effect modifiers. Analytic results from each data sources will be combined using novel meta-analytic techniques to overcome any issues of small numbers occurring. -EUROmediCAT study : Case-malformed analyses(logistic regression)