Potential issue compute.treatment.episodes and dosage.change.means.new.treatment.episode parameter

[SebastianKN1998]

Hi,

Unlike medication.change.means.new.treatment.episode, where if the type of medication supplied changes, the new medication type is considered a new treatment episode (regardless of whether there is overlap between medication supply). dosage.change.means.new.treatment.episode parameter does not appear to start a new treatment episode?

From the documentation it appears like it should start at new treatment episode. Dosage.change.means.new.treatment.episode parameter “Logical, should a change in dosage automatically start a new treatment episode?”, but to me it appears like that is not what happens.

It looks like it does not consider the dosage change as a new treatment episode if there is an overlap of medication supply between the “old” and “new” dosage.

See example where it creates a new treatment episode when the medication type changes, but not if the dosage changes.

For now I rename my medication type each time the dosage changes to achieve the same results with the medication.change.means.new.treatment.episode parameter.

library(data.table)
library(AdhereR)


test_dose <- data.table(
  id3 = c(1, 1, 1),
  eksd = as.Date(c("2023-01-01", "2023-02-01", "2023-04-01")),  # gap before dose change
  dd2 = c(10, 10, 5),
  duration = c(30, 30, 30),
  med.type = c("apixaban", "apixaban", "apixaban")
)
test_med <-data.table(
  id3 = c(1, 1, 1),
  eksd = as.Date(c("2023-01-01", "2023-02-01", "2023-04-01")),  # gap before dose change
  dd2 = c(10, 10, 5),
  duration = c(30, 30, 30),
  med.type = c("apixaban", "apixaban", "drugB")
)
eps <- compute.treatment.episodes(
  data = test_dose,
  ID.colname = "id3",
  event.date.colname = "eksd",
  event.duration.colname = "duration",
  event.daily.dose.colname = "dd2",
  medication.class.colname = "med.type",
  consider.dosage.change = T,
  dosage.change.means.new.treatment.episode = TRUE,
  medication.change.means.new.treatment.episode = TRUE,
  maximum.permissible.gap = 90,
  suppress.warnings = FALSE,
  return.data.table = TRUE
)
print(eps)






eps2 <- compute.treatment.episodes(
  data = test_med,
  ID.colname = "id3",
  event.date.colname = "eksd",
  event.duration.colname = "duration",
  event.daily.dose.colname = "dd2",
  medication.class.colname = "med.type",
  consider.dosage.change = T,
  dosage.change.means.new.treatment.episode = TRUE,
  medication.change.means.new.treatment.episode = TRUE,
  maximum.permissible.gap = 90,
  suppress.warnings = FALSE,
  return.data.table = TRUE
)
eps2

Thanks for the nice package.

Kind regards,
Sebastian