2- Not all participating institutions had policies limiting the activities of the detailers, this might contribute to the differences in the study results.
detailers, which or detailers; this
3- number of responded- responses
with more surveyed and responded doctors- this bit is unclear to me- what do you mean?
4- follow up followed
This can reduce as many non- respondent as possible, which can affect the reliability of statistical estimates, possibly leading to interpretation bias. (does this read better?)