Discussion: View Thread

  • 1.  Estimating a relative treatment effect

    Posted 07-22-2016 11:33

    Does anyone know of a method to get an estimate of hazard ratios (HR) of Treatment A to Treatment B from literature, when we have estimates of Treatment A to control from one study and Treatment B to the same control in a separate study?

    The 2 studies were run by the same investigator and have the same design. The responses for the control group in the 2 studies were almost identical (1.29 / 1000 days in Study 1 and 1.3 /1000 days in Study 2), which gives me confidence in combining the data from the 2 studies.  I have estimates of the HR compared to control from both studies, 95% CIs for the estimates, and sample sizes. 

    Can I combine the data to get an estimate and 95% CI for HR of A/B?

    Thank you. 


     

    ------------------------------
    Shelley-Ann Walters
    Biostatistical Specialist
    3M Health Care Business

    ------------------------------


  • 2.  RE: Estimating a relative treatment effect

    Posted 07-22-2016 12:56

    Health economists (and others)  like methods of "indirect comparison" 

    I'm generally skeptical of methods that compare groups that weren't randomized..

    https://www.ispor.org/workpaper/interpreting-indirect-treatment-comparison-and-network-meta-analysis-studies-for-decision-making.pdf

    ------------------------------
    Chris Barker, Ph.D.
    Consultant and
    Adjunct Associate Professor of Biostatistics


    ---
    "In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
    -Steve Lacy



  • 3.  RE: Estimating a relative treatment effect

    Posted 07-23-2016 20:49

    That ISPOR Task Force Report on Indirect Treatment Comparisons referenced by Dr. Barker? It has a "Part 2" companion report, located at:

    http://www.sciencedirect.com/science/article/pii/S1098301511013283

    Also, here is the PubMed abstract of an application of network meta-analysis to comparing overall survival between different RCTs in the melanoma context: 

    http://www.ncbi.nlm.nih.gov/pubmed/23024154 

    ------------------------------
    Eric Siegel, MS
    Research Associate
    Department of Biostatistics
    Univ. Arkansas Medical Sciences