With numerous technologies beginning to disrupt the ways in which data are collected and generated, Sarah Fal speaks to Terry Katz of Merck Animal Health, who offers his views on the changing landscape of data management.
Sarah Fal: Can you give me a little more of an explanation on what eSource is and what technologies fall under it?
Terry Katz: eSource refers to the original capture of data electronically per the EMA and FDA, both of whom have formal guidelines (FDA regulation – 21 CRF 11 and the FDA Guidance on Electronic Source Data in Clinical Investigations, and the EMA’s Reflection Paper on Expectations for Electronic Source Data and Data Transcribed to Electronic Data Collection Tools in Clinical Trials).
The main aim is to shift industry technology to eSource, which means we need innovation in this area. Currently, there are already providers offering great solutions such as Target Health. Many big pharmaceutical companies tend to be reluctant to this change due to their conservative nature. My team has been innovative with using tablets and laptops to collect clinical data at the point-of-contact, which adds the benefit of immediate checks for data quality within expected ranges.
SF: How do you predict big pharma will adapt to eSource?
TK: It will all come down to a willingness for change from the top. However, there are some other trending areas in the clinical data area in relation to Fast Healthcare Interoperability Resource (FHIR).
This is a tool that has been developed over the past four to connect electronic health records (EHR) and to sponsor databases. Here, you can extract Electronic Medical Record Data and Electronic Health Records. Crucially, FHIR was made to connect the 4000+ EHR systems under HL7, and is additionally, it can be used to connect to CDISC EDC.
SF: How is this helpful to the industry?
TK: You can transfer data from one system to another in a smooth and effective way. In the U.S., as opposed to the U.K., there are so many formats for collecting data making integration more difficult. However, it’s still needed. We want to be able to pull health records directly and put them onto other systems without manual transcribing. This gains efficiency and enhances data integrity.
SF: When it comes to data formatting, what are the pressing concerns data managers face?
TK: With Central Laboratory Systems, you can now create XML, which is extensible mark-up language required for the submission process. Here you can refer lab data into a standard machine readable tool like EHR to be readily available to the clinician. Data is in XML already, but EHR does not read this yet.
All patient records could then pbe updated seamlessly as it’s all in one language. XML is needed for FDA submission, though packaged-XML is a challenge for some agency systems. It is a complex combination of many XML tables into a single file. It’s good for extract, but it’s a big challenge for some FDA groups where we have to unpackage, merge, and then submit.
SF: Oh I see, other than FDA submission, what are some of the other purposes of the data you acquire?
TK: Once data is integrated it goes to statisticians who provide analysis, which then flows to the R&D team. In the future, I envision that artificial intelligence will do some of the analysis as it is already being deployed in disease diagnosis.
Head of Data Management and Statistics
Merck Animal Health