Indirect treatment comparison accounting for the differently timed scheduled assessments  

Challenge

Some time-to-event outcomes are interval-censored: they are only known to occur within an interval (i.e., between the current and last visit when the patient was known to be event-free)

  • If interval-censored events are treated as happening at the time of the visit, then the risk of the outcome may be underestimated due to the systematic delay in recording

  • In the context of indirect treatment comparisons (ITCs), there may be bias in favour of the study which has longer time intervals between scheduled assessments 

Request

Perform an ITC accounting for the differently timed scheduled assessments between the index and comparator trials

Response

  • Assessed various potential methods to offer potential solutions to this problem
  • Recommended using Assessment Schedule Matching (Kapetanakis et al. 2019)1  to reduce the potential bias by adjusting the time-to-event survival function of the index treatment
  • Developed programs to apply this novel methodology to the client data 

Systematic reviews and evidence synthesis        Statistical analysis

 1 Kapetanakis, V., Prawitz, T., Schlichting, M. et al. (2019) Assessment-schedule matching in unanchored indirect treatment comparisons of progression-free survival in cancer studies. PharmacoEconomics. ISSN 1170-7690.