Nov 7, 2014

Wendy: With the vast amounts of data that we have available to us now, we're able to better support decision making at the point of care in clinics and in hospitals. The biggest question is how do we take advantage of that data, how do we collect it, how do we integrate it, and how do we present it to people in a way that helps them make those decisions? That's what's coming next on The Scope.

Announcer: These are the conversations happening inside healthcare that are going to transform health care. The Health Care Insider is on The Scope.

Wendy: My name is Wendy Chapman and I'm the chair of the Department of Biomedical Informatics at the University of Utah. And today we're very lucky to have a visitor from Mayo Clinic, Dr. Jyoti Pathak, and he is an associate professor of informatics at Mayo.

Dr. Pathak: Thank you. First of all thank you for the invitation Wendy, it's been a fun trip here to Utah. So, my role at Mayo Clinic in the Center for Science of Health Care Delivery, which is a new center that was formed about, roughly, a year and a half ago, is to lead the Clinical Informatics program, where we are dealing with a lot challenges and issues in terms of integrating clinical data with genomic information, with data we are collecting from different environmental sources to facilitate better decision making, better patient based outcomes research and hopefully clinical interventions that would save lives and improve the patient outcomes.

Wendy: Tell us about some of the projects that you're involved in and how the integration of that data can really improve outcomes for patients.

Dr. Pathak: One of the projects that I am involved in is a project funded by the National Institute of Health, it's called Electronic Medical Records and Genomics. And the primary focus of that particular project is how we could extract clinical data from our electronic health record systems at Mayo and combine that with genomic information that is extracted from human DNA, human plasma. And to identify associations between different diseases and genes, between different responses to drugs and genes, drug toxicities and genes, so that we are able to identify those patients ahead of time and, again, intervene early to prevent a bad outcome.
So, in particular, for example, we are doing a study where we are looking at a drug called Plavix or Clopidogrel and trying to identify genetic mutations that might predispose someone as a poor metabolizer. In other words, the drug is not necessarily acting completely and we are trying to identify those patients who are at the risk of being poor metabolizers so that either we could regulate the drug mechanism or we could regulate the drug behavior or in many cases switch the drug itself.

Wendy: What are the types of data you need to integrate to answer those kind of questions?

Dr. Pathak: So, we deal with a variety of information, a lot of our data actually does come from the electronic health record systems, so that comprises of demographic information such as gender, age, and the like. It requires information about different medication history that a particular patient would have experienced, it requires information about different diagnoses and different labs and procedures that were done, and more increasingly it also requires information about your genetic profile. So we have systems and we have procedures in place that allow us to sequence the human genome and extract the relevant genotypes of interest which we call as Actionable Genotypes, in other words, we are in a position to integrate such information within the Mayo's electronic health record system so that we could act on that information.

Wendy: So, in building a digital infrastructure what are the challenges you face in collecting that data and in integrating it and presenting it to people for decision making?

Dr. Pathak: In general I think that there are many, many challenges but here are the top three that come to my mind. So the first challenge is really around data representation because the moment we are talking about combining clinical data that is typically within our electronic health record systems with genomic information that is not typically in electronic health record systems, there are several challenges in how you actually represent that information, what kind of nomenclature is used to semantically identify that data and that's a huge problem that we at Mayo are trying to solve and many others are trying to address in the academic community.
The second challenge that I personally think is very important is really around the volume of this information. The moment we are thinking about doing whole genome sequencing of our human genomes the vast amount of information that we are extracting is unprecedented. And that requires a new suite of technologies that requires a new way of processing that information that is just coming and just getting developed in the market. And those technologies have not necessarily been highly available in the health care community. So we are looking at systems that are, for example, used in other industries such as the airline industry or the finance industry where they are using systems for processing large volumes of information or big data systems and we are trying to incorporate that within Mayo.
And I guess the third challenge really is around security and privacy, because at this time the moment we are starting to talk about our genetic information it becomes very personal in nature and that requires a very different way of how you store the data, that requires a very different way of how you protect that information. And particularly if you are trying to exchange data between different institutions or different healthcare organizations, and such security protocols are either getting developed or in many cases undeveloped. And again that's a challenge that I see becoming more and more prevalent as this information becomes readily available across the United States Healthcare System.

Wendy: Have you had any success in your work sharing data with different institutions?

Dr. Pathak: So data sharing, particularly medical data sharing has always been a very thorny topic. We have had , I would say, reasonably good success, again, within the merged consortium, the Electronic and Medical Records and Genomics Consortium, a lot of it was hinging upon establishing what we call data use agreements. So they were pre-agreements that were signed between multiple different institutions who agreed to share de-identified electronic health record data and genotype data across the consortium members.
Now, for those institutions who are not part of the consortium members I think the problem still exists and again I think there is a role for the technology to play here in terms of enabling secure data exchange, there's a role for our institution review boards to play here to embrace the idea of data sharing and I think in general there is a role for the whole research community to play in terms of enabling and encouraging data sharing because it is going to eventually have a shared benefit across our entire research enterprise.

Wendy: Here in Utah we've had a lot of success getting people in the community to share their health data as well as their genetic data to help us discover genes that cause certain diseases, and I wonder if you've had similar experiences and what your hope for the future is in regard to that?

Dr. Pathak: My hope really is that I think we are moving into a whole new world of patient centeredness. So, we are going to increasingly see involvement of patients in research activities, in scientific activities, throughout the entire spectrum where they're inherently part of the system and hence are increasingly having the mechanisms and the willingness to share information; whether that be family history information, whether that be genealogy information, whether that be clinical health information or whether that be information about your daily activities.
I think with the advances in our information technology systems, with the advances in the whole science of informatics and I think increasing awareness in the community that we really need patient powered research or patient centered research. I suspect we are going to see increasing momentum in data sharing.

Wendy: What's the role for our listeners?

Dr. Pathak: I think we are entering an age where a lot of our digital health information is actually in your pocket with the new iPhones and the new Android phones they are going to monitor your heartbeat, they are going to monitor your blood pressure 24/7, and I think it's becoming an interesting opportunity for computer scientist for informaticians for bioethicists and the like to develop new frameworks, develop new systems that not only make it feasible to share information easily but more importantly to share information in a much more secured way.
I suspect that's where there's a lot of hesitation, in terms of how this information would be shared, and in many cases how this information would be misused working throughout the system, in terms of creating policies, creating governance models, that enable better protection and better security around this data is really what is needed. But I can certainly see a trend where us as individuals are increasingly going to share more and more information in an anonymized way that is eventually going to benefit us and our families and our friends.

Announcer: The ScopeRadio.com is University of Utah Health Sciences Radio. If you like what you heard be sure to get our latest content by following us on Facebook, just click on the Facebook icon at TheScopeRadio.com.

For Patients