Data integration for ISS/ISE

Integration of data from a number of clinical trials for an Integrated Summary of Safety (ISS) and Efficacy (ISE) requires careful planning and includes the following planning steps [see ICH M4]:

  • Assess the analysis and reporting requirement for the ISS/ISE
  • Consider these requirements against pre-existing study level analysis and reporting
  • Determine what data types need integrating across studies and at what level (SDTM/ADaM) the integration should occur at.

Assess the analysis and reporting requirement for the ISS/ISE

Before starting integrating data across studies it is key to have a clear understanding of what questions the ISS/ISE is trying to answer and how these requirements differ to those already answered by the individual studies.

It is helpful to also consider the potential questions of the rapid response part of the submission and approval process. Considering possible questions at an early stage allows the potential for related derivations to be discussed and agreed before the time-pressured phase of rapid response. Obvious rapid response questions will usually end up being covered in the main ISS/ISE analysis. Lower priority questions may not be included in the main ISS/ISE analysis, but preparing for these during the data integration set up would still be beneficial.

The analysis required to answer these questions will drive what data types need to be combined and the derivations that need to be applied to the study data.

Analysis requirements against pre-existing study level analysis

Do current or planned study level analyses provide sufficient information to answer ISS/ISE requirements? To be considered sufficient, outputs from individual studies need to be able to provide a complete response to the regulatory questions without the need for the reader to combine tables. The regulator may want to interrogate the source data for any ISS/ISE table or figure, so this should be clearly identified and easily accessible.

Often derivation algorithms (e.g. for efficacy endpoints) evolve throughout the phases of studies that are included in a submission. If this is the case, the earlier studies will need re-deriving so that consistent algorithms are applied across all studies.

Consideration must also be given to consistency of coding (e.g. medications and adverse events) across the studies. Often later studies are ongoing and the coding dictionary versions are changing as new version come out. As the project gets close to database lock date for the final studies, it is usual to consider freezing the version of the coding dictionaries used in those studies. This allows sufficient time to recode the earlier completed studies to the agreed version.

Controlled terminology used in the component studies needs to be considered as well. Controlled terminology covers items like parameter codes and unit descriptions. These need to be standardised in the combined datasets to match the current applicable external standard and to be standardised across the studies.

How to combine and create ISS/ISE datasets

From the previous steps there should be a good idea of what datatypes will contribute and how the study level algorithms differ to the required.

Only the study level data used in the production of the necessary ISS/ISE analysis are required to be included in the combined datasets. Care should be taken though as ISS/ISE analysis requirements can continue to change and the potential for rapid response question on these data types may arise later in the study reporting process.

Combine study datasets at the SDTM or ADaM level? If the algorithms required at the ISS/ISE level are different to what exists in most of the individual studies it may make sense to combine at the SDTM level and rederive the endpoints to create the ISS/ISE ADaMs once at the ISS/ISE level.

From Phastar’s experience, it is often the case that required algorithms have been implemented in later studies only.

Rederiving ISS/ISE ADaMs from combined SDTMs has the added benefit that the algorithms used at the ISS/ISE level are then independent from the individual study level algorithms. This allows the ability to more quickly change ISS/ISE algorithms as often study reporting teams may be different to the ISS/ISE team.

More often than not with ISS/ISE the combining of data is at the SDTM level.

Combining at the ADaM level and not having ISS/ISE SDTMs is feasible but tends to be in cases where there are few later phase studies being combined. Combining at the ADaM level has the benefit that combined SDTMs are not created and no CRT documentation is required for them. Often in these cases the ISS/ISE Case Report Tabulation (CRT) documentation is a lot simpler as the derivations are documented at the study level.

It is important to carefully consider all benefits and issues when deciding which level to integrate study data into ISS/ISE. It can be problematic to swap approach later in the ISS/ISE reporting cycle. Therefore, it is recommended to be conservative when making the decision to combine at the SDTM or ADaM level.

Phastar has considerable experience in the planning for and management of the reporting work for the data integration in ISS/ISE. Every ISS/ISE is different and we are expert at tailoring the ISS/ISE reporting to meet client and regulator requirements in the most efficient manner.