The Application Of Biostatistics In Treatment Effect Evaluation Of New Hypolycemic Medicine

Diabetes is an increasingly prevalent chronic disease that seriously affects people worldwide.The number of patients is predicted to rise to 0.55 billion by the year 2030.The research of new hypoglycemic drug and how to assess the efficacy scientifically are crucial to patients’ health.The main objects of this thesis is to investigate the efficacy of new drug in patients with type 2 diabetes mellitus(T2DM).Efficacy evaluation includes 2 aspects,one is the evaluation of the treatment effect at the end of the trial,and the other one is the treatment effect over time and whether there is significant differences between patients with different characters.Clinical trial is widely conducted to assess the efficacy of treatment.In Chapter 1,I introduce steps and methods in clinical biostatistics.From a lot of paper,I introduce current challenges and latest research progress in efficacy evaluation of new hypoglycemic drug.In Chapter 2,I define steps and methods to evaluate the imbalance in a randomized clinical trials with two(treatment and control)groups.And then though MCMC methods to simulate how the randomized imbalance influence the evaluation of treatment effect in different imbalanced score.In the evaluation of treatment effect in primary endpoints,model selection is a crucial issue.In this thesis,consider the randomized information,I compare 4 kinds of models in different scenarios and show how to choose appropriate model.And I derivate in which kind of scenarios,even if we don’t choose the most appropriate model,we will get asymptotic estimation of treatment effect.Then use MCMC to simulate the trials indifferent scenarios.The models includes linear,quadratic and models with change continuous baseline to categorical by a cut-off value.In Chapter 3,I briefly introduce the trials and data of this new drug,then consider the imbalance of the trail and use the most appropriate model to evaluate the treatment effect.The conclusion is that after 24 weeks,the mean HbA1 c change from baseline of patients with T2 DM is-0.52% compared with those who don’t take the drug.In Chapter 4,consider the treatment effect over time and the differences between patients,I use Longitudinal data Hierarchy Linear Model to analysis repeated measure data of patients.There are some conclusions.The efficacy of the new drug is significant and the mean HbA1 c change from baseline is-0.187% every 6 weeks.There is no significant differences between patients with different gender,age and baseline HbA1 c.The rate of HbA1 c change from baseline of patients with HbA1 c more than 8.5% is more than patients with HbA1 c less than 8.5%.In Chapter 5,I summarize the conclusions and prospect the follow-up work and research.