Read our blog yet?

TrackREX: Track and reconcile non-CRF data including biosamples, labs or ECG with CRF-derived datasets. Reconciliation is automated with an intuitive and streamlined approach that is system-independent.  

​

TrackREX is a re-engineered solution that is based on a previously released PK (pharmacokinetic) reconciliation tool. It comes with numerous enhancements and has evolved into a client-focused solution whereby only those functions that are required are included. The main focus of the tool is on tracking and reconciling of data, however, additional functionality can be added throughout the life-cycle of a study.

​

TrackREX takes a typical approach to reconciling data from different sources (see below).  Apples & Oranges personnel set up the tool for each study and provide clients with the output, with a 100% datapoints 'scan' and then reconciled with automatic findings presented in clearly defined outputs.  The overall process allows for a much more refined experience for end-users:

​

1) Data inputs from any data source and with no restriction by EDC platform or data type (accepts RAW/SDTM).

2) The 'matching process' (reconciliation) is performed automatically with any mismatches identified.

3) The final output shows the end-user discrepancies and any potential errors with suggestions for resolution. An added bonus is that TrackREX will retain all user actions and any associated comments for inclusion in future report outputs (do double-data review!).

​

​

​

​

​

​

​

​

​

​

​

​

 

​

​

​​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​​​​​​

TrackREX functionality

​

The tool (and output) is pre-programmed

Typically PK/vendor reconciliation is performed by using SAS to import EDC and vendor datasets and programmatically compare data fields that are captured in both datasets. This requires a SAS programmer and DM to collaborate with agreed specifications. To arrive at an effective solution can prove expensive and time consuming; a process repeated per study instead of a single department-wide solution. At the end of the exercise, the result is still a flat file that needs manual review, often requiring double-data review at subsequent review cycles (after all, how do we know old data didn’t change?).  Our Apples and Oranges personnel need only to ‘feed’ both the protocol’s time and events schedule (T&E) and raw data into the tool in the format recognised by the tool’s own mapping table.  This provides a re-usable, familiar, output to clients across multiples studies, yet is tailored per-protocol.  Thus, TrackREX is not in need of re-programming for each and every study.  

​

Flexible data sources

As a stand-alone program utilising Excel VBA to produce outputs, the TrackREX solution will work with any EDC/CRF or lab platform and can feed in both raw and/or SDTM datasets once available in .xlsx or .csv format. It is possible to add further systems into the equation such as IWRS/IVRS and CTMS platforms which will enable further functionality (see below).

​

Feed in ancillary data such as vendor and other sponsor spreadsheets

Often studies will have ancillary data that can be useful to be incorporated into the reconciliation review. This could be sponsor feedback (pertaining to certain samples or patients) or perhaps input from a lab vendor (e.g. updates on samples that may not yet have been received or processed).  As an additional add-on service, this type of ancillary data can all be incorporated into the TrackREX output if they are associated with key variables that can be programmatically mapped to a sample. By having reconciliation and ancillary data under one tool, this makes the process a far more efficient solution. 

​

Adjustable review mode allowing users to add/remove columns as per personal preference

Reconciliation sheets typically are large, cumbersome to navigate and review comprehensively.  Our tool is pre-built with the ability to have columns hidden, on show and re-ordered based on client request (or ‘specification’).

​

Easy to understand discrepant data issues

Thousands of hours are typically wasted during any clinical trial by individuals manually comparing data points. Whilst this can simply be done programmatically, our solution takes it a few steps further. Data anomalies or discrepancies are not only highlighted, but suggested resolutions or humanised language are also displayed (examples: ​"The sample was drawn at 03-May-19 13:49 and falls outside of expected time window (expected between 03-May-19 13:50 & 03-May-19 14:10.  Please check.“; " The Accession Number 1029860 has been used for another subject.  Please check.”). It can also suggest where the action needs to be made such as at site or lab vendor.

​

Apply data cuts by date, visit, treatment arm, etc

Most clinical trials will have numerous analyses or data review points during the study. Trying to determine which samples need to be reconciled around cut off dates or cohorts can be complex. Providing the raw data is available, other attributes such as cohort, treatment arms can be displayed in the final output, together with of course dates that may fall outside of that snapshot. With output available in the most familiar of all programs study personnel will always demand (Excel!), this is ideal for studies with multiple deliverables!

​

​No need to re-review data

Actions are stored for future review cycles (i.e. the user will not have to re-review data that has not changed). A fully reviewed and commented upon output will feed back into TrackREX and automatically merge with new or changed raw data.  Therefore issues resolved from prior data review and without data changes will retain both comments and statuses.  This is one of the biggest benefits of our solution in that users do not have to start each review from scratch or indeed have to review comments made during a previous review cycle if the issue itself is considered closed.

​

Permanent issues

Each study will have data issues that will never be resolved. These are typically identified during reconciliation and will include incorrect sampling times, samples not taken or lost samples which may be need to be recorded as protocol violations. The tool therefore allows for these permanent issues to be recorded and be exported for review by the PK scientist or statistician during the conduct of the study.​​

​

No more manual review with multiple open spreadsheets

During the lifetime of  a typical study, TrackREX can save hundreds of hours per study team!

Data reviews will typically consult numerous spreadsheets as part of their data reconciliation efforts: 

• EDC extracts (Drug administration, Sample ID/reference point (if included in CRF), visit information

• Lab vendor

• Sample Data Requisition listings from site/lab

• Other ancillary data from vendor or sponsor

​

These data extracts are typically 'programmed' in a haphazard fashion with rudimentary Excel functions including VLOOKUP and manual filtering and at worst are reviewed entirely manually. Part of this review process may not only be repeated by the same reviewer multiple times during the study, but will likely also be  undertaken by other functions. This overlap and duplication of effort has a tremendous impact not only on resourcing but may also have ramifications on the study results themselves. As the usual review of sample data is via manual method, the very activity is often performed seldom or too late in the study when issues are caught too late. At this stage, endemic issues are too late to rectify and missing samples are identified too late to be actioned.  Based on A&O’s experience, it is not uncommon for a study close to lock to have ~5% error rates when it comes to reconciliation.

​

Additional potential

At A&O, we listen to what our clients need.  In addition to comparing vendor to CRF data for incorrect labelling, timing, missing vendor data or indeed missing or incorrect CRF entries, given so much information is already in a single tool there is the potential to explore the data in other ways.  This may include critical KPI collection - see below!

​

Status metrics / site performance KPIs

Comprehensive metrics can be made available that provide ​study specific insights or measurable KPIs against portfolio criteria.  Were specific sites struggling to perform and require additional training?  Was sample shipment well coordinated and timely?  Was the protocol ‘too expecting’ with a sample schedule too difficult to adhere to in the clinical setting?  Has the CRO missed a lot of data discrepancies, and should the task itself be brought back in-house?

​

​​

​

TrackREX vs Laboratory Information Management System (LIMS)

​

Whilst LIMS have long been regarded as ultimate solution for tracking and reconciling lab data, they often fall short of being able to do this effectively. Typical issues associated with a LIMS:

​

• Difficult to implement and maintain

• Costly to purchase and service (overly complex, steep learning curve)

• System is slow (often due to associated amounts of data from various inter-connected systems)

• Resistance by adopters (Different parties along chain of custody use multiple incompatible systems)

• Outputs may not be in the program study personnel want (Excel!)

​

Overall result: ‘Back to square one’:

• Manual tracking of Samples are tracked at multiple points along the chain of custody with numerous tracking sheets being referenced

• Clerical (human) error prone recording (e.g. no scanning of barcode labels)

• No automated tracking!

• Samples are lost, inaccurate labelling = lost data!

​

LIMS, with their associated issues and lack of full adoption means that there is often a shortfall when it comes to dealing with both the tracking of samples and reconciliation of the associated data. TrackREX will never replace a LIMS,  but will function independently of all systems and plug any of the real-world shortfalls which are typically tracking and reconciliation of sample data.