Plain language summary (192/200):
Researchers have recently started using a new method, called
“clone-censor-weighting” (CCW), to compare the effect of starting drug
or medical treatments during specific time windows (e.g., between 0 and
30 days or between 30 to 90 days after a heart attack). Using CCW to
make these comparisons can be very complicated, however, both in terms
of actual implementation and in terms of the actual question researchers
are answering. We walk readers through a fairly simple application of
the CCW method to compare the effect of starting aspirin or clopidogrel
within 30 days, within 90 days, or between 30 and 90 days of heart
attack within fake data originally created by the Centers for Medicare
and Medicaid services. We also provide statistical code for readers to
duplicate our analyses. In addition to removing bias from measured
variables that impacted whether people were removed from our study, we
also created figures to represent when people started treatment within
each of the hypothetical treatment regimens being studied. While using
CCW can be challenging and it answers a deceptively complicated research
question, it is very useful for comparing different windows in which
people might start treatment.