7 Case Studies Previous The Children’s Hospital, UC Denver On Going Johns Hopkins University NorthShore University HealthSystem
8 My Concerns Using IT to provide a solution that does not have an institution level policy Getting adequate representation from stakeholders Developing a consensus Transparent process
9 Comments from survey… These comments reflect only the personal perspective of anonymous authors Probably not the perspective of their institutions Certainly not mine – I don’t have personal consensus yet
10 Comments … The EMR should NOT be part of the eSource material that is considered under the FDA proposed guidance. We have already done a few studies that would sweep selected labs from the EMR source systems, transform the MRN to the study ID, and send it daily to the study sponsor. I would not consider the Cerner lab system that was the source of the HL-7 feed to be covered by the FDA for research, but I would consider the programs that did the transformation and shipping to be covered. The CTSA's are making a big push to use the EMR data for research. Both for finding participants and for collecting clinical data on them.
11 The data elements identified as PHI in the HIPAA regulations are ROUTINELY found in the EMR so the first sentence is incorrect. HIPAA states that when these data elements are moved from the clinical care context into the research context, then additional oversight protections are triggered for these PHI data elements. So the SAME data element in a clinical care context has different regulatory oversight than when that exact same piece of data is used in a research context.
12 What principle trumps the other? the argument that withholding any research data from clinical practitioners represents a patient safety risk …or… the argument that it is unethical to use data collected under the terms and protections specified in the IRB, and by the sponsor or perhaps required under a NIH Certificate of Confidentiality, in a clinical care context. Especially when doing so could lead to disclosures to insurance companies, with the negative consequence of rendering a patient uninsurable forever.
13 Additional Considerations My concern is the fact that some research data should not be discoverable to 3rd party payers. What if any legal/regulatory restrictions exist? I believe this is a wrong ‘question’. I think that the EMR infrastructure can be used to store and manage research data, but for the most part such data should be kept logically separate and inaccessible from EMR.
14 Here we have concern about confidentiality of research data once it is in the EMR. In a multi- institution research environment, data sharing is still a huge obstacle (including de-identified data), how do we overcome that barrier? New forms for Breaches of Confidentiality?
15 What are the IRB implications of having data that is collected for research, but that is relevant to standard clinical care and captured in the EHR? And if visible by clinicians not listed on the IRB Protocol? Are there any effective ways to facilitate the research community engaging with the clinical practice community to encourage the collection of data in clinical operations? So that collection will best support research while not being burdensome to clinical workflow!
16 Patient care data is insufficiently contextualized to understand the conditions under which it was collected. Even a measure as simple as a serum calcium can't be used for research purposes if the specimen was collected on persons in fundamentally different states -- if the states aren't accessible in the same retrieval as the serum calcium. And discharge diagnoses are worthless!
17 Next Steps… There are several threads within the CTSA landscape How do we keep each group informed of the others’ progress? Different institutions may chose very different polices/paths Does it matter as CTSA partners?
Continuing the discussion There will be an Affinity Group created on the CTSA wiki Contact me with the subject line: EMR Research Data email@example.com 18