Precision medicine goes global: How collaboration can redesign healthcare
Sharing health information to inform precision medicine initiatives is taking data sharing to a whole new level—and crossing international datelines.
Sharing health information to inform precision medicine initiatives is taking data sharing to a whole new level—and crossing international datelines.
Spurred on by value-based care payment models and the need to manage populations of patients, payers and providers have become more adept at sharing data within their own systems and with other organizations in their region.
The U.S. government has pushed to improve interoperability and allow for wider dissemination of genomic and other data through programs such as the Precision Medicine Initiative and Sync for Science for good reason: Developing optimal medical treatments and drugs holds great promise for improving population health.
But precision medicine is a data-hungry discipline. And managing all that data—getting accurate and actionable information to the right place at the right time and using it in an appropriate way—is a heavy lift. “That's where the bulk of the effort is going to be on the clinical side,” says Brad Ptasienski, a director in consulting firm West Monroe Partners’ technology practice. “There's going to be big data management and master data issue that arises from all this data [and] getting the data in the right place at the right time will be a challenge.”
Creating a data fabric
Researchers need as much standardized data and as many sample data sets as possible to discover what's possible, he adds. “What kind of insights can they drive? What sets of variables seem to correlate? They're going to be the ones that are trying to figure out just how valuable this information is.” And on the clinical side, applying data-driven “eureka moments” from the research side and applying it in clinical medicine “is going to be the crux of the work effort.”
And data privacy, of course, cuts across everything. Genomic data is “the ultimate PHI … it's everything about you,” Ptasienski says.
“There's a data fabric that has to get created in the ecosystem,” says Munzoor Shaikh, a director in the healthcare and life sciences practice at West Monroe. “So imagine all these disparate EMRs and systems and labs: That's what exists today. There's no common place for them to talk.”
A national database might not be the solution, he says, especially as data-sharing spreads across the globe. Rather, he envisions domain-specific data hubs, including one focused on precision medicine. “It’s a bit like blockchain in the sense that not one person owns it,” he says. “It's owned, essentially, by the overall ecosystem.”
Building an ARC
A network of academic medical centers in Israel, the US, Canada and Europe have created just such a hub. The ARC Innovation Center was founded at Israel’s largest hospital, Sheba Medical Center in Ramat Gan—participants include Intermountain Healthcare in Utah, Stanford Healthcare in California and Ottawa Hospital in Canada.
ARC, which stands for accelerate, redesign and collaborate and includes both a virtual and physical workspace, focuses on using AI, big data and genomics to drive precision medicine initiatives. The goal is to “change healthcare completely within 10 years,” says Eyal Zimlichman, MD, Sheba’s chief innovation officer.
Redesign and collaboration are critical, Zimlichman says. “Healthcare has so many problems. You need to destroy what we have and build it from scratch. No one academic medical center, no matter how strong, would be able to do that alone.”
One of the reasons healthcare hasn’t realized the full potential of precision medicine is that it has focused too much on genomics, Zimlichman contends. While genomic data is one layer, ARC also collects a broad spectrum of data that includes proteomic, metabolomic and microbiomic data, patient-generated data from wearables and sensors—even Internet browsing history, which can be a predictor of depression.
“Accessing all of these elements together will get us to where we want to be,” Zimlichman says. “And this is the main aim of precision medicine: being able to predict, being able to prevent, being able to treat precisely [as opposed to] the trial and error that we do today.”
Gathering data is one thing—sharing it in an accurate, secure and meaningful way is another.
“There needs to be a common platform,” Zimlichman says. “Data from different sources is like different languages. And even if you do get into one location, whether it's a cloud or whether it's an onsite server … you cannot just start working on them together.”
ARC uses a cloud platform called MDClone, which creates “synthetic data” from original data sets. The result is data that’s statistically equivalent to the original but contains no individual’s personal health data.
“Now we are able to take all of our malignant melanoma patients from Sheba, all of the patients from Intermountain and all of the patients from Ottawa and join them together into one source of data,” Zimlichman says. “And our abilities to do the calculations on the big data are obviously much improved.”
Going beyond borders
Canada’s Ottawa Hospital, a 1,200-bed academic health science center, participates in the ARC initiative with an eye toward providing high-quality care to its patients and better understanding the population it serves.
“Even though we have a lot of data and we’ve helped make it accessible, we don't even have all the data for the population of patients we look after,” says Alan J. Forster, MD, the organization’s vice president of innovation and quality “People talk about big data, but we're not really a big data shop right now because we don't have all of the possible data that you could have of people that could tell you a story about their health.”
Collaborating with organizations “around the world who are trying to accomplish similar goals in a different way does help accelerate your understanding as to what works and what doesn't work,” Forster says. “And so I think that's a pretty important part of this. As an industry, we tend to think within our borders … Solutions may come faster if we look across borders. Different health systems’ views can help us reconcile what's working and not working.”
Bigger data sample sizes helps, too, in a district that includes just about 1.3 million people. “If you're trying to predict things and you don't have a huge sample size, you can have quite imprecise estimates of what you're trying to predict,” Forster says. “And then you don't have a real ability to validate. So that becomes a problem if you're trying to develop solutions.”
The cloud platform makes it easy for researchers to ask clinical questions without the aid of a data scientist to develop queries or understand the data, he adds. “The opportunity for information to be stored in the cloud where people can connect to it virtually, with almost no risk from a privacy perspective … unlocks our ability to understand the impact of our drugs, it helps us understand the value created from some of our procedures. It helps us understand the value created by healthcare. To me, that is where innovation will start to really accelerate.”
Creating incentives
“The challenge is how do we get people to participate in multiple hubs? What's in it for them?” West Monroe’s Shaikh says. “Without a reward mechanism, you're never going to have people playing and sharing.”
Forster agrees. “It can't all be upside and no downside. Right now the risk is not really shared. People can walk away from providing care in almost any system. In fact, at times walking away from a service can actually be profitable. If you really want to make this work, you have to think of the health of the population as a whole and you have to have penalties and incentives aligned with the goal of achieving population health,” he says.
“The people who can really move this agenda forward in the US are the big insurance agents. They have the power to bring people together and make it happen.”
When that piece falls into place, patients will benefit. “For so long in healthcare, we've been fixated on why we shouldn't do things,” Forster says. Data data-sharing “unlocks our ability to understand the impact of our drugs, it helps us understand the value created from some of our procedures. It helps us understand the value created by healthcare. And so, to me, that is where innovation will start to really accelerate.”
Spurred on by value-based care payment models and the need to manage populations of patients, payers and providers have become more adept at sharing data within their own systems and with other organizations in their region.
The U.S. government has pushed to improve interoperability and allow for wider dissemination of genomic and other data through programs such as the Precision Medicine Initiative and Sync for Science for good reason: Developing optimal medical treatments and drugs holds great promise for improving population health.
But precision medicine is a data-hungry discipline. And managing all that data—getting accurate and actionable information to the right place at the right time and using it in an appropriate way—is a heavy lift. “That's where the bulk of the effort is going to be on the clinical side,” says Brad Ptasienski, a director in consulting firm West Monroe Partners’ technology practice. “There's going to be big data management and master data issue that arises from all this data [and] getting the data in the right place at the right time will be a challenge.”
Creating a data fabric
Researchers need as much standardized data and as many sample data sets as possible to discover what's possible, he adds. “What kind of insights can they drive? What sets of variables seem to correlate? They're going to be the ones that are trying to figure out just how valuable this information is.” And on the clinical side, applying data-driven “eureka moments” from the research side and applying it in clinical medicine “is going to be the crux of the work effort.”
And data privacy, of course, cuts across everything. Genomic data is “the ultimate PHI … it's everything about you,” Ptasienski says.
“There's a data fabric that has to get created in the ecosystem,” says Munzoor Shaikh, a director in the healthcare and life sciences practice at West Monroe. “So imagine all these disparate EMRs and systems and labs: That's what exists today. There's no common place for them to talk.”
A national database might not be the solution, he says, especially as data-sharing spreads across the globe. Rather, he envisions domain-specific data hubs, including one focused on precision medicine. “It’s a bit like blockchain in the sense that not one person owns it,” he says. “It's owned, essentially, by the overall ecosystem.”
Building an ARC
A network of academic medical centers in Israel, the US, Canada and Europe have created just such a hub. The ARC Innovation Center was founded at Israel’s largest hospital, Sheba Medical Center in Ramat Gan—participants include Intermountain Healthcare in Utah, Stanford Healthcare in California and Ottawa Hospital in Canada.
ARC, which stands for accelerate, redesign and collaborate and includes both a virtual and physical workspace, focuses on using AI, big data and genomics to drive precision medicine initiatives. The goal is to “change healthcare completely within 10 years,” says Eyal Zimlichman, MD, Sheba’s chief innovation officer.
Redesign and collaboration are critical, Zimlichman says. “Healthcare has so many problems. You need to destroy what we have and build it from scratch. No one academic medical center, no matter how strong, would be able to do that alone.”
One of the reasons healthcare hasn’t realized the full potential of precision medicine is that it has focused too much on genomics, Zimlichman contends. While genomic data is one layer, ARC also collects a broad spectrum of data that includes proteomic, metabolomic and microbiomic data, patient-generated data from wearables and sensors—even Internet browsing history, which can be a predictor of depression.
“Accessing all of these elements together will get us to where we want to be,” Zimlichman says. “And this is the main aim of precision medicine: being able to predict, being able to prevent, being able to treat precisely [as opposed to] the trial and error that we do today.”
Gathering data is one thing—sharing it in an accurate, secure and meaningful way is another.
“There needs to be a common platform,” Zimlichman says. “Data from different sources is like different languages. And even if you do get into one location, whether it's a cloud or whether it's an onsite server … you cannot just start working on them together.”
ARC uses a cloud platform called MDClone, which creates “synthetic data” from original data sets. The result is data that’s statistically equivalent to the original but contains no individual’s personal health data.
“Now we are able to take all of our malignant melanoma patients from Sheba, all of the patients from Intermountain and all of the patients from Ottawa and join them together into one source of data,” Zimlichman says. “And our abilities to do the calculations on the big data are obviously much improved.”
Going beyond borders
Canada’s Ottawa Hospital, a 1,200-bed academic health science center, participates in the ARC initiative with an eye toward providing high-quality care to its patients and better understanding the population it serves.
“Even though we have a lot of data and we’ve helped make it accessible, we don't even have all the data for the population of patients we look after,” says Alan J. Forster, MD, the organization’s vice president of innovation and quality “People talk about big data, but we're not really a big data shop right now because we don't have all of the possible data that you could have of people that could tell you a story about their health.”
Collaborating with organizations “around the world who are trying to accomplish similar goals in a different way does help accelerate your understanding as to what works and what doesn't work,” Forster says. “And so I think that's a pretty important part of this. As an industry, we tend to think within our borders … Solutions may come faster if we look across borders. Different health systems’ views can help us reconcile what's working and not working.”
Bigger data sample sizes helps, too, in a district that includes just about 1.3 million people. “If you're trying to predict things and you don't have a huge sample size, you can have quite imprecise estimates of what you're trying to predict,” Forster says. “And then you don't have a real ability to validate. So that becomes a problem if you're trying to develop solutions.”
The cloud platform makes it easy for researchers to ask clinical questions without the aid of a data scientist to develop queries or understand the data, he adds. “The opportunity for information to be stored in the cloud where people can connect to it virtually, with almost no risk from a privacy perspective … unlocks our ability to understand the impact of our drugs, it helps us understand the value created from some of our procedures. It helps us understand the value created by healthcare. To me, that is where innovation will start to really accelerate.”
Creating incentives
“The challenge is how do we get people to participate in multiple hubs? What's in it for them?” West Monroe’s Shaikh says. “Without a reward mechanism, you're never going to have people playing and sharing.”
Forster agrees. “It can't all be upside and no downside. Right now the risk is not really shared. People can walk away from providing care in almost any system. In fact, at times walking away from a service can actually be profitable. If you really want to make this work, you have to think of the health of the population as a whole and you have to have penalties and incentives aligned with the goal of achieving population health,” he says.
“The people who can really move this agenda forward in the US are the big insurance agents. They have the power to bring people together and make it happen.”
When that piece falls into place, patients will benefit. “For so long in healthcare, we've been fixated on why we shouldn't do things,” Forster says. Data data-sharing “unlocks our ability to understand the impact of our drugs, it helps us understand the value created from some of our procedures. It helps us understand the value created by healthcare. And so, to me, that is where innovation will start to really accelerate.”
More for you
Loading data for hdm_tax_topic #better-outcomes...