Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Are marginal structural models used for repeated/panel data?

    I've been studying this paper from the Stata journal by Fewell et al on implementing MSM (https://ageconsearch.umn.edu/bitstre...art_st0075.pdf). I've only found reference material for marginal structural models in the context of time-to-event analyses. Most published studies in epidemiology and social sciences seem to apply it in Cox/pooled logistic models.

    What about panel / longitudinal repeated data: Can the principals of MSMs be applied to such mixed models or GEE, and are there any examples of it in the literature?

  • #2
    You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output and sample data using dataex.

    This is not a technique many of us use so you'll probably have to look back at the underlying literature and/or or contact the authors.

    Comment


    • #3
      Marginal structural models are needed when you have repeated measurements with time-varying confounding. With such data GEE and mixed models will be biased. MSM and other g-methods are commonly used in epidemiology and biostatistics. Hernan/Robins Causal Inference is a textbook on these model
      https://www.hsph.harvard.edu/miguel-...nference-book/

      Comment

      Working...
      X