Doctors and their patients often don’t have the information they need on the relative effectiveness of different treatments. Clinical trials provide invaluable data but can’t and don’t cover the myriad of individual circumstances in the real world of patients. As part of the Affordable Care Act, a number of hospitals, research centers, clinics, insurers and patient groups are working to create a massive database of medical records – stripped of personally identifiable data. The idea is to allow scientists to study the relative effectiveness of any number of different drugs, devices and treatment plans, but questions about privacy persist. Please join us to talk about big data and medicine.


  • Dr. Joe Selby Executive director, Patient-Centered Outcomes Research Institute family physician, clinical epidemiologist, and health services researcher
  • Ceci Connolly Managing director and leader of PwC's Health Research Institute; co-author of "Landmark: The Inside Story of America's New Health Care Law and What It Means for Us All."
  • Dr. Eugene Rich Senior fellow and director, Center on Health Care Effectiveness, Mathematica Policy Research
  • Dr Deborah Peel Practicing adult psychiatrist founder and chair, Patient Privacy Rights


  • 10:06:54

    MS. SUSAN PAGEThanks for joining us. I'm Susan Page of USA Today sitting in for Diane Rehm. Diane is visiting station KRCU in Cape Girardeau, Mo. As part of the Affordable Care Act, some medical providers are in the process of pooling patient data to be used for research. Proponents say it will provide critical information to doctors and their patients on the relative effectiveness of different treatments, devices and drugs.

  • 10:07:20

    MS. SUSAN PAGEBut there are privacy issues to be addressed. Joining me in the studio to talk about the project: Dr. Joe Selby of the Patient-Centered Outcomes Research Institute, Ceci Connolly of PWC's Health Research Institute, and Dr. Eugene Rich of Mathematica Policy Research. Welcome.

  • 10:07:37

    DR. JOE SELBYThank you.

  • 10:07:37

    MS. CECI CONNOLLYThank you.

  • 10:07:38

    DR. EUGENE RICHThank you.

  • 10:07:39

    PAGEAnd joining us by phone from Moscow, Dr. Deborah Peel of Patient Privacy Rights. Thanks for being with us.

  • 10:07:47

    DR. DEBORAH PEELHi. Thanks for calling.

  • 10:07:49

    PAGEAnd we welcome our listeners to join our conversation as well. You can call our toll-free number, 1-800-433-8850, send us an email at, or find us on Facebook or Twitter. Well, Joe Selby, start by just explaining to us what this database is going to be.

  • 10:08:10

    SELBYThanks, Susan. PCORnet is a national research network that's supported by PCORI, the Patient-Centered Outcomes Research Institute, that involves researchers, patients, clinicians and healthcare delivery systems and that will conduct needed research, research that couldn't be conducted in any other way.

  • 10:08:30

    SELBYIt takes advantage of the remarkable amount of clinical information that's being placed into electronic health records every day and is presently untapped, as far as research is concerned, and of the remarkable energy of patients who want this information used to answer their questions and those of loved ones.

  • 10:08:50

    PAGESo it aggregates data from how many patients?

  • 10:08:54

    SELBYWe believe that, at the end of 18 months, we'll have data on somewhere between 15 and 20 million patients.

  • 10:09:00

    PAGEFifteen to 20 million Americans. And what kind of data would it be?

  • 10:09:05

    SELBYIt would include both data that comes from electronic health records and other computerized data in healthcare delivery systems, the old claims data, for instance, and, for the first time, substantial amounts of data provided directly by patients. As I said, patients are actively involved in this network and patient-reported information is a crucial kind of data that haven't been available...

  • 10:09:31

    PAGESo it would be things like your blood pressure or what pills you were taking or whether you had a heart attack.

  • 10:09:38

    SELBYExactly. And also how you are feeling and what symptoms you're having.

  • 10:09:42

    PAGEAnd so, Dr. Gene Rich, how might a doctor use this? How might this database be tapped?

  • 10:09:46

    RICHWell, I mean, there certainly -- the first way these data could be used would be to generate a much larger volume of more timely published research. And this could be research that now would be much more salient to individual clinical circumstances, to particular patient populations, to clinical settings that might not otherwise be involved in a traditional randomized controlled trial.

  • 10:10:17

    RICHSo now clinicians would be provided with, you know, the kinds of research they really always wanted to have, but often didn't have available. There's also the possibility in the future that we might learn how to use these kinds of data systems for even more sort of real-time analysis, but I would say that's an idea that's sort of out long in the future.

  • 10:10:41

    PAGEReal-time analysis. Like, a doctor has a patient. I want to figure out what the best drug to use for this condition is. Let me tap into this database and figure out what worked.

  • 10:10:49

    RICHRight. I mean, as Joe knows, there are a lot of issues involved in making sure that that would provide the kind of reliable information that you would want that kind of search to have. But that is sort of a long-term vision of what a network like this might be able to provide, a real-time information to clinicians they could share with patients.

  • 10:11:14

    SELBYAnd, in fact, if you can imagine, Susan, right now, we make decisions about what treatments to recommend that are based on the average results from relatively small groups of people who've been involved in clinical trials. If you can imagine being able to draw on the data of millions of persons to get much closer to the answer of what would work best for me based on what has worked best for patients like me in the past.

  • 10:11:39

    PAGESo, Ceci, this sounds incredible. On the other hand, it sounds just like what we have happening every time we shop on Amazon, this kind of use of big data.

  • 10:11:49

    CONNOLLYWell, Susan, I think you're hitting on a very important point here, which is the promise of big data has really only been a promise in healthcare for a very long time. If you think back to the dawn of electronic medical records, this is what we were told, finally having all of your information in that one place, even in the cloud, we're going to learn so much. And I think what's exciting is it appears that we are finally nearing capitalizing on big data for our own health and our community health.

  • 10:12:22

    PAGENow, Deborah Peel, you've had some concerns about privacy when it comes, for instance, to the use of electronic records. Tell us what your concerns have been.

  • 10:12:33

    PEELWell, there are a lot of concerns. The first thing is, the public and your listeners and maybe some of the other speakers, I believe, do not really understand how little control patients have over any of their data today. I mean, patients control virtually none of the many, many users of data. And it was interesting to me, for you to say -- I think it was Joe that was talking about the future promise of real-time analysis.

  • 10:13:11

    PEELWell, you know, if you were me, you would be studying what commercial products are already claiming to do right now, which is real-time -- actually there's a product from Mayo called Bedside Analytics. It's claiming to do exactly what you say will take quite a lot of time with your database. This data is already being used by many others without patient consent.

  • 10:13:40

    PEELAnd the big problem, I would say, about research in general is the public has never endorsed and does not endorse the use of their data without being asked first. They very much support research, but projects where people are not specifically asked run a huge danger of driving patients away from healthcare and away from wanting to participate in electronic systems.

  • 10:14:08

    PAGEAnd would it drive -- when you say it would drive them away from healthcare, is it because they're concerned that they'll be identified with their -- people will be able to tell this is this particular person's medical history?

  • 10:14:19

    PEELIt's not just identification of data, which, by the way, is very easy. You know, all the researchers say that the data is de-identified, but if you talk to computer scientists, there's really no question it's easily re-identifiable because there's so many databases out there that are public that can be, if you will, electronically cross-walked with health data to re-identify it. And so, you know, everyone that uses and sells data right now -- like IMS Health also claims to have the largest research database of 500 million people that includes many of the things that PCORI database includes.

  • 10:15:02

    PEELIt includes electronic health records, prescription records. It includes all payer claims database and social media to create longitudinal profiles of patients that they sell to 5,000 clients. And in addition to that, their statement to the SEC to sell stock says that, in fact, that they buy, sell and trade our health data, all of our health data, with 100,000 data suppliers that cover 780,000 live daily data feeds. And so...

  • 10:15:41

    PAGESo let me give Joe Selby a chance to respond, first to the question about identification.

  • 10:15:46

    PEELSure, sure.

  • 10:15:47

    PAGEIs it easy, as Deborah Peel said, to identify particular patients and their medical history through this database?

  • 10:15:55

    SELBYIt's not easy, but it is doable. I think the key points here are that the data exists already. We are not calling for pulling these data out of the healthcare systems and sending them somewhere to another large database. We're calling for them to stay in the networks, in the individual networks and to be protected with the same security that protects the data already.

  • 10:16:16

    SELBYSo first of all, I completely agree with Dr. Peel that privacy has to be held in the highest regard as we begin to take advantage of the potential of these data. We cannot make this an either/or. It can't be either privacy or research. It has to be a version of both and that's why we've put patients into patient organizations, patient networks, individual patients into every part of PCORnet so that as we're devising the privacy and security policies within each network, patients, as well as clinicians and healthcare systems, are involved.

  • 10:16:56

    PAGEEugene Rich.

  • 10:16:58

    RICHYeah, at Mathematica, we had the chance to evaluate the Recovery Act Investment and compare the effective research, which was over $1 billion and nearly 100 projects on data infrastructure. And among the lessons that we learned from those projects was the importance of dealing with the patient confidentiality and data security issues up front. A lot of those projects in the Recovery Act built on previous investments like the Sentinel Network, the HMO Research Network, projects that ARC had undertaken.

  • 10:17:35

    RICHAnd there was an entire initiative called the Electronic Data Methods Forum that continues where the investigators supported by these different projects have been exploring different ways to make sure that these data could be used for research, but also were confidential. And there are some quite sophisticated techniques that these researchers have come up with.

  • 10:17:58

    PAGESo, Ceci, do you think patients should be concerned about their privacy in this new database?

  • 10:18:03

    CONNOLLYWell, I think, as Joe and Deborah and the others have all pointed out, there are certainly going to be risks and we hear about data security breaches in any number of industries, not just healthcare. So it continues to be a modern-day challenge of the world in which we live. I can tell you, Susan, when we do a lot of surveys, we are seeing a growing number of patients that are very willing to share information if it does something like further research.

  • 10:18:33

    CONNOLLYAnd so they really appreciate those tradeoffs.

  • 10:18:36

    PAGEWe're going to take a short break. When we come back, we'll continue our conversation and our phones lines are open. You can call our toll-free number, 1-800-433--8850. Stay with us.

  • 10:20:01

    PAGEWelcome back. I'm Susan Page of USA Today sitting in for Diane Rehm. And joining me in the studio, Ceci Connolly. She's managing director and leader of PwC's Health Research Institute. She is the co-author of the book "Landmark: The Inside Story of America's New Healthcare Law and What It Means For All of Us." (sic) And Dr. Eugene Rich, senior fellow and director of the Center on Healthcare Effectiveness at Mathematica Policy Research.

  • 10:20:27

    PAGEAnd Dr. Joe Selby, he's executive director of the Patient-Centered Outcomes Research Institute. He's a family physician and a health services researcher. And joining us by phone from Moscow, Dr. Deborah Peel. She's an adult psychiatrist. She's founder and chair of a group called Patient Privacy Rights. So in talking about this new database and how it might be used, Joe Selby, how would it be different from and perhaps more valuable than the clinical trials that we're all familiar with?

  • 10:20:59

    SELBYA couple critical differences. First of all, clinical trials are typically conducted in volunteers. And that already limits our ability to be confident that the results apply to people in real world practice. Both the patients are different and the clinical settings where the clinical trials are conducted are different.

  • 10:21:17

    SELBYWe need -- we want to know what works in real-world practice. So that's the first thing. These data come from real-world practice in general and they are not selected. They are not so selected. They include people with all of the other illnesses and the symptoms and social circumstances that patients face every day. They're not selected.

  • 10:21:37

    SELBYThe second is they're much larger. If we want to know what works best for whom, we need much larger studies, much bigger amounts of data to look in patient subgroups. I will hasten to say though, Susan, that a number of these studies, even in these large settings, will still need to be randomized trials. And they will still certainly, in those cases, need patient to consent.

  • 10:22:05

    PAGESo there are advantages to clinical trials that you wouldn't have in this database.

  • 10:22:09

    SELBYClinical trials are thought of as the most like a laboratory experiment. You control everything. You get a precise answer. The problem is the precise answer doesn't necessarily generalize very well to the real world.

  • 10:22:20


  • 10:22:20

    CONNOLLYWell, and also clinical trials and FDA approval focus on safety and efficacy, two very important thresholds. But when you get out in the real world and you're a patient with an illness, you care about more than that. You care about your quality of life. When we talk to people with an illness they don't say I want safety and efficacy. They say, I would like to walk my daughter down the aisle for her wedding in six months and is this medication going to enable me to do that.

  • 10:22:46

    CONNOLLYOr you think about, the drug may not be tested in many 80-year-olds but if you're the 80-year-old who's then prescribed that medication, what information is there to guide you and your physician as to whether or not it's going to be a good fit for you?

  • 10:23:00


  • 10:23:01

    PEELThis is...

  • 10:23:02

    PAGEYes, Deborah, go ahead.

  • 10:23:05

    PEELOK. All right. As a practicing physician and as someone who represents 10.3 million people, I really have to jump in to comment on some of the things that have been said. First of all, if I understand what Joe said, PCORI does have patient advisors but the individual patients themselves are not asked if they wish to participate specifically in research projects. And that's a very important right that patients have had in this country. It's a very important part of research, which is respect for patient autonomy and decision making.

  • 10:23:43

    PEELThe other thing is that there are many polls about people wanting to be altruistic about research. That's very true. But the most detailed neutral poll that I'm aware of was conducted by Prof. Alan Weston with Harris Interactive for the Institute of Medicine. And it looked at all kinds of populations of people, vulnerable people, elderly people, men, women, education levels.

  • 10:24:12

    PEELAnd only -- and they looked in depth at how people felt about the use of their electronic health information without consent, without permission, specific permission. And only 1 percent were willing to have their data used for research without permission.

  • 10:24:33

    PEELOnly 19 percent would agree if they were told that their data was de-identified and an institutional review board looked at whether it was a valuable project. So these numbers are very important because, as you all know, many polls are conducted by people who want a particular outcome. And this one was not. (unintelligible) ...

  • 10:24:56

    PAGEAll right. Let me ask Joe Selby about your first point on whether patients whose data is going to be shared in this aggregated system. Do they have a chance to give their permission or there a chance to say no, I don't want to be part of that?

  • 10:25:11

    SELBYWell, let me just repeat what I said a minute ago that if this is a study that is changing therapy in some way, if it's a study that is, for example, a randomized trial, individual patient consent is an essential part of that. If it's a study that requires a million individuals -- the electronic health record data of a million individuals, a number that would render seeking and obtaining informed consent from every person impossible.

  • 10:25:41

    SELBYIf it's a choice between not being able to do an important piece of research and having to contact -- try to contact a million people, first of all, you wouldn't be able to contact 30 to 40 percent of them.

  • 10:25:54

    PEELYes, actually you could.

  • 10:25:55

    SELBYYou just -- you can't even with multiple phone calls, with emails, with all the ways we have. A lot of people just don't find the time to respond. I don't respond -- I hate to admit it -- when people call me about research sometimes. I try to.

  • 10:26:09

    PAGEBut if this is coming from your medical provider, couldn't the medical provider, at the point you sign up for care or when you come in for a checkup, you know, require you to...

  • 10:26:19

    SELBYWell, this is...

  • 10:26:19

    PAGE...look at a form and sign up or not?

  • 10:26:21

    SELBY...this is exactly what needs to change and why we're happy to have patients, as well as the healthcare delivery systems, involved in PCORI net. We need a dialogue, a national dialogue about the topic we're talking about today. I hope a lot of people are listening. But we need to talk about precisely this point, how do we take advantage of the data to get answers to questions when it's not -- when the risks are extraordinarily low and it is not possible to talk to people one by one. I think that's the question on the table.

  • 10:26:50

    PAGEEugene Rich.

  • 10:26:51

    RICHYes. And I think another advantage of the PCORI network and the opportunities for learning here are that we're now going to be having a system that connects clinicians and their offices with a robust research network, a mechanism whereby it can be easy to remind the physician, hey, there's a protocol that this patient would fit into.

  • 10:27:17

    RICHOr here's an opportunity for the patient to agree to participate in ongoing data collection so that there really is -- there are a variety of mechanisms that can be tested out with a lot of input from patient advisors regarding what would be the most comfortable mechanism for pulling together that patient information.

  • 10:27:42

    PAGECeci, Deborah Peel quoted a survey on behalf of the Institute of Medicine by Harris Interactive that showed very low rates of willingness to share -- people to share their medical data. Is that consistent with the survey data you've looked at?

  • 10:27:57

    CONNOLLYWell, if I understood that -- and I've not looked at that particular study, but if I understood, it was around the notion of if you've been asked, you know, given some sort of consent. And I think that that can be very important in terms of public attitudes. There is anxiety about things being done with your personal information and you're not aware of it. Of course, as you point out, Amazon is a good example of that happening to us now regularly in society.

  • 10:28:25

    CONNOLLYSo we're in a very rapidly changing world environment today and we're all trying to kind of catch up with how we're going to navigate this brave new future, if you will. I do think, Susan, though that you have an excellent idea, which is I know when I go to my physician I'm signing a consent for various other bits of information that's going to various people. I think one thing for this country and healthcare providers and patients to consider is a broader encompassing consent for the purposes of research.

  • 10:28:57

    PAGEHere's an email we've gotten from Thomas who writes us from Heartland, Mich. He says "The idea is a good one for research but is another example of what is the problem when people lose trust in the federal government. I, for one, would not trust them to keep it private even when I told them I wanted to keep it private." Let's go to the -- Joe Selby, did you want to respond to that? Yes.

  • 10:29:17

    SELBYWell, I get that and we at PCORI get that. And that's precisely why this data model is a distributed data model. The data stay at the health systems. And what centrally are the results of analyses, no possible chance in the world of learning something about an individual.

  • 10:29:33

    PAGELet's go to Paul who's calling us from Raleigh, N.C. Paul, hi, you're on the air.

  • 10:29:38

    PAULHi. My question is just fairly quick, but it's a little longer term. So in 20 years, what does this database do to medicine? That's really it, and I can take the question off the air.

  • 10:29:49

    PAGEOK. Paul, thanks very much for your question. Who on the panel would like this one? In 20 years, how will medicine look different because of this database?

  • 10:29:58

    SELBYIt -- first of all, probably the most important thing is that the quality of care will go up. The quality of decisions and choices will go up. And the decisions will be much more personalized. They'll be much more tailored to the clinical condition and the preferences of individual patients. And patients and clinicians should become better partners. I think the other thing is, clinicians will become, once again, a bit more of the scientist. They'll have great confidence in looking at data.

  • 10:30:27

    PAGECould the government use this in ways -- I mean, are insurance companies in terms of what procedures they're going to pay for, what procedures they'll approved based on which ones this database shows work or doesn't work?

  • 10:30:41

    CONNOLLYWell, I certainly think that there is the potential for -- you know, when you talk about comparative effectiveness research, one goal is to figure out what medication, what therapy, what treatment, what option works best for which individuals. And even though PCORI and the ACA are prohibited from considering cost in their deliberations, insurance companies certainly are not.

  • 10:31:09

    PAGEAnd so they could look at this and say, this only works this...

  • 10:31:13 this particular group.

  • 10:31:15

    PAGEAnd you've all talked about being able to customize treatment. Do you mean by looking at -- would this database enable you to look at someone's genetic makeup and try to tailor a treatment to people who had certain characteristics?

  • 10:31:31

    RICHThe gentleman from Raleigh asked about in 20 years. And I think probably in about that interval, genetic information on a large-scale basis will be a part of the electronic medical record, and it will figure into decision making for a lot of conditions.

  • 10:31:45

    PAGEI'm Susan Page, and you're listening to "The Diane Rehm Show." We're taking your calls and reading your emails. Join us at 1-800-433-8850. Deborah Peel, let me put that question to you. This database is in the works. It's supposed to be up and operational in September 2015. Looking 20 years down the road, how would you see the practice of medicine as being different because of it?

  • 10:32:08

    PEELI think if we continue to not ask individuals, I think we're going to put everything in jeopardy in terms of getting these great answers to the research that all of us want. Let me give you a recent example of what's gone on in education. So Google was offering free electronic record systems to schools that collected massive amounts of data. And they wanted to be able to tailor lessons to each child and so forth. But they did not get the parents' permission. They did not get the students' permission. And the data is all to be shared with numerous third parties. It killed the project.

  • 10:32:50

    PEELNot asking people for consent today is ridiculous. Everyone can be contacted by email. We get alerts from our banks all the time. People are very familiar with this. So technology allows exactly what we're talking about where each person can be given choices. Yes. I myself broadly want to contribute to any comparative effectiveness research there is. Or, no, I only want to be involved in research on breast cancer or diabetes or whatever. People should be able to decide broadly or narrowly which research they want to participate in.

  • 10:33:30

    PEELThis is one of the key tenets of research, that it's voluntary and that people's autonomy and rights to make decisions are respected. And we know that the public is very altruistic. If you ask them they will do it. And I would really hate to see these kinds of things continue to go on with people not being asked. People feel very violated.

  • 10:33:53

    PAGEDeborah peel, if they -- Joe Selby indicated a willingness to talk about this process. Clearly, this is just in the works now, not fully operational. If there was a process of approval beforehand, is this a network that you think would make some sense and improve medicine? If they ask permission before people's data was shared? Deborah?

  • 10:34:20

    PEELYeah, of course it would. Absolutely. No. With permission. we would get all of the data that's needed to make these very improvements, plus we'd get trust with our patients, trust with us.

  • 10:34:32

    PAGERight. Interesting. Eugene Rich.

  • 10:34:34

    RICHYes, so I think this network, building out this concept over the next 20 years, I'm hopeful that it will mean that both clinicians and patients will be much more used to using real salient evidence to guide decisions. Unfortunately, you know, so often now the available research isn't relevant to that setting or that circumstance.

  • 10:35:00

    RICHAnd so clinicians are used to sort of in many cases not practicing evidence-based medicine, and patients are not used to being engaged in what the evidence means for them. Twenty years from now, this can be an entirely different set of conversations where we have much more powerful evidence and much better tools for engaging patients in the uncertainties that remain.

  • 10:35:24

    CONNOLLYAnd, Susan, this is so critically important right now because if you stop and think about the larger trends in healthcare today, individual consumers and families are having to make many more of the healthcare decisions themselves. They're responsible for paying more and more of the cost. And so as we go forward with more of that responsibility on the individuals, we want to arm them with good information.

  • 10:35:49

    PAGECould consumers access...

  • 10:35:50

    PEELThey want to be armed.

  • 10:35:52


  • 10:35:52

    PEELThey want to be armed, but they want to be asked. I mean, the idea that this...

  • 10:35:57


  • 10:35:57

    PEEL...could be imposed on people is really not a good idea. It really -- individuals want to be asked and they want to help. And I would point out too that by design patients have been cut out of the electronic health system today. You probably know that we were supposed to be able to get downloaded copies of our records since the HIPAA privacy rule went into effect. But most of us can't do that even today, whatever this is...

  • 10:36:23

    PAGEAnd Ceci, let me ask you.

  • 10:36:25

    PEEL...12 years later. We can't even communicate directly with our doctors in most cases. Patients and doctors have really been sidelined by the data holders and the large institutions. And we need to get doctors and patients back at the center of this.

  • 10:36:41


  • 10:36:42

    SELBYFirst of all, that's our mantra, so I think we're in complete agreement there. PCORI net is made up of 11 networks that spring from healthcare delivery systems, but 18 networks that actually are patient organizations, online patient communities, patient advocacy organizations.

  • 10:37:00

    SELBYAnd they are the people from whom we're hearing with most energy and demand that we use these data and in fact that we look into some of the currently really burdensome practices in terms of excessive informed consents. So I really agree, and I think PCORI net is a good forum for centering the research on patients and clinicians.

  • 10:37:20

    PAGEWe're going to take another short break. When we come back, we're going to go straight to the phones and take some of your calls and questions. Stay with us.

  • 10:40:01

    PAGELet's go to the phones and take some questions and comments from our callers. We'll first go to Rock Hill, S.C., and talk to James. James, thanks for holding on.

  • 10:40:09

    JAMESThank you. My question was -- and the direction the conversation's going really dovetails into it. Just the amount of trust right now between patients and medical care providers, I'm wondering if this data aggregation also includes any information about the cost of service and, I guess, any quality of service claims with the providers, so that it would help patients, both from a research side, but also to be more empowered consumers.

  • 10:40:38

    PAGEWhat an interesting point. Let me ask our panel.

  • 10:40:41

    CONNOLLYWell, I have to say, James, I'm so glad you're asking about that because, not necessarily PCORnet, but we saw a major announcement and release of data last week by the federal government of payments to physicians for Medicare. And it is this sort of cost transparency that so many consumers, patients, are desperately wanting to see.

  • 10:41:05

    CONNOLLYBecause what we are identifying is wide variation in our country on cost and medical outcomes, with not a lot of good explanation for it, I believe. And I'm going to let Joe jump in here as the expert. I think that claims data can be a part of this new proposed network, as well. But to the larger point, we are finally getting information out on what care costs in this country. And I think it's going to be a very valuable discussion.

  • 10:41:36

    PAGESo before we leave the point of the Medicare payment data that came out last week, it'd been 30 years since this data had been publicly released. I think it was only released because the Wall Street Journal really pursued it.

  • 10:41:46


  • 10:41:46

    PAGESo kudos to the Wall Street Journal on that. What kinds of things could we learn from that data, from the Medicare data that was released?

  • 10:41:53

    CONNOLLYWell, for instance, you can take a look at a map of the United States and you can see if there are places that our outliers, in terms of doing lots more of a certain procedure or charging a whole lot more for a certain procedure. And maybe they're not -- they're patients aren't doing any better, but they're charging two, five, 10 times as much money. You can also look for patterns and, again, match it up with electronic medical record data and begin to really figure out what works in healthcare.

  • 10:42:22

    PAGESo, Joe Selby, is that something a patient could do with this new database, when it's up and running?

  • 10:42:32

    SELBYWell, PCORI interprets its authorizing legislation as staying away from studies of costs and cost effectiveness. So we are in the business of focusing on clinical outcomes. And comparing the clinical outcomes, which include things, like how am I feeling and what can I do, and am I having symptoms. We don't support with PCORI funds, the collection of cost data.

  • 10:43:01

    SELBYI think, though, that this convergence of patients, clinicians and systems in PCORnet could be -- I mean, we've got to bring these various stakeholders together around the whole of healthcare delivery, and certainly -- just yesterday, there was another large meeting here in D.C. about transparency with respect to costs, across the board in commercial plans. So I think PCORnet could create a forum for folks to talk about that.

  • 10:43:28

    PAGESo you look at whether it works, but not how much it costs.

  • 10:43:31

    SELBYWe -- yes. And one of the reasons we do that is precisely what Ceci said, that costs vary so much that in a research setting it's kind of tough to make sense of the costs. We look at resource utilization or, you know, visit rates or surgery rates, but we don't look at the costs associated with them.

  • 10:43:49

    PAGEEugene Rich?

  • 10:43:51

    RICHYes. And as Joe said, once we have the evidence regarding what intervention works best for which kinds of patients, as this process of cost and price transparency plays out over this next 20 years as we've been talking about and probably in that case, much sooner than 20 years, now clinicians and patients will be able to look together at both the evidence and the cost implications for that patient and their family.

  • 10:44:23

    RICHAnd really make that a patient centered decision, where the patient's values are up front. Clearly, our current system now, clinicians don't really have the relevant evidence and they don't have the tools to present that to the patients. And our system of pricing and costs have made it very difficult to have useful point-of-care conversation about that.

  • 10:44:53

    PAGELet's talk to John. He's calling us from Boston, Mass. Hi, John.

  • 10:44:56

    JOHNHi. My question was about the usefulness of the data. One of the underlying assumptions of the discussion is that the data will be helpful. But that's only true if the data are collected with a protocol that makes them uniform across providers. It also assumes that diagnoses that clinicians make are always accurate. So I was wondering how the data collection would deal with these problems of standardization and diagnosis.

  • 10:45:25

    PAGEYeah, great question, John. Eugene Rich?

  • 10:45:28

    RICHYes. Well, this is one of the issues that has been a challenge right along, as researchers have been working on what could be done with these kinds of data. And there were, again, in the recovery act investment in data infrastructure, a lot of explorations of how to deal with this problem. And I think over the years we've already identified opportunities where this kind of marriage of electronic health record data to, say, claims data and other administrative data, can be very powerful.

  • 10:46:00

    RICHOne example I'll give. A few years ago, as folks probably are aware, robotic surgery became a technique for hopefully improving outcomes from prostate cancer surgery. Now, if you're a health services researcher working with claims like me, you'd love to be able to look at that, but the way coding was done for robotic surgery, you couldn't actually tell which patients had the robotic surgery or not.

  • 10:46:27

    RICHAnd the surgical intervention process is one that is in constant evolution, so that a claim for a sort of surgical technique may not tell you whether the latest version of that technique was used. But that's the sort of the thing that is very easily and consistently detected through electronic health records.

  • 10:46:48

    RICHThere are other examples, though, where we're still trying to figure out certain kinds of confidential patient information that many -- for many good reasons clinicians don't always even record in the record. And in a circumstance like that might be where we actually need a formal protocol in order to collect the data.

  • 10:47:07

    PAGEWhat kind of data would someone not want to put in their record or clinician might choose not to put in?

  • 10:47:11

    RICHOh, let's say for example, that I'm, as a general internist, weighing which treatment I should -- how I should approach this patient's diabetes and hypertension. And I recognize that one of their problems is I think they're also a heavy alcohol user. And I'm trying to work with them on that. But it may be that because of the particular setting I'm in the patients are very uncomfortable with me having that in the record.

  • 10:47:37

    RICHSo I may -- there are a variety of techniques I might use to remind me that that's a consideration. But to the taxonomists who are analyzing the electronic health records, you can't tell. And as I've just described, that could even vary from clinical setting to clinical setting, where in some clinical settings that's recorded, in other clinical settings that's not. And that would create problems for the kinds of observational studies we're talking about. But those are all issues that PCORnet will allow people to study and explore.

  • 10:48:11

    PAGEYou know, we've gotten a lot of emails to Deborah's point about who owns your medical data. Here's an email from Melissa, who writes, "The question is who will profit from this. And I don't mean medically, I mean dollars. It seems too many companies are profiting from individual data all the time, so why shouldn't patient data be considered personal property and those patients be paid for access to their data?"

  • 10:48:33

    PAGEAnd a similar question that we've gotten from another listener, Eric, who writes, "I'd like to hear Joe Selby comment on who owns the data, what ownership of that data allow. What concerns me is the discoveries resulting from analysis of data ultimately will be commercialized by pharmaceutical and medical device companies, like what has happened in the past." And I wonder, Joe, if you could respond to that. And I'd then like to hear, also, Deborah have a chance to respond. But, Joe, you first.

  • 10:48:57

    SELBYWell, I completely believe that in the end patients own their data. Now, how we operationalize that and deliver care at the same time and maintain the quality of care and improve it and conduct research is a challenge. I want to point again to those 18 patient-powered networks. They very clearly understand that they own the data and they make the rules within each network that are comfortable to them about how it's used.

  • 10:49:27

    SELBYSo that's the bottom line. I think we have a system today where the data reside under the control of the healthcare systems. And we have to ask ourselves, how do we take advantage of it? But, for example, there is this initiative called the Blue Button Initiative, which will eventually allow patients to get their own, to extract all of their own data from claims and electronic health records. And we are hopeful that the patient-powered networks will be able to use that function to both get their data's individuals (sic) and to include it in their network's database.

  • 10:50:02

    PAGEDeborah Peel, who would you respond to these listeners who are raising concerns about who owns their medical data?

  • 10:50:10

    PEELWell, that is a very important question. And our bi-partisan coalition has over 50 organizations in it. We represent 10.3 million people. That's one of the issues that we struggled with because apparently, according to the lawyers, "property rights model" wouldn't work well for medicine. The ownership of data is too difficult a legal question to answer or to push anything through Congress that says we definitely own our data.

  • 10:50:41

    PEELHowever, morally and ethically, it's very hard to say that it belongs to anyone other than the patient. It's true doctors have to keep records, but they're supposed to be data stewards on our behalf. So it's a difficult question, but we have always had in this country very strong rights to control our data. Whoever, you know, whoever wants to believe they own it. And I can tell you that most of the institutions that hold our data believe it's their data. They believe it's an asset of the corporation.

  • 10:51:16

    PEELAnd that's why companies like IMS Health, and the 100,000 health data suppliers and so on believe that they can sell and trade our data, but it's, you know, it should belong to us. And technology could absolutely enable us to control it and to manage it and to share it with whom we like very easily. The problems that we're talking about today really aren't technology problems.

  • 10:51:41

    PEELLike Blue Button, it's not hard at all to create the ability to download instantly your data and get it. We just haven't been given it because the larger institutions want to control this data. They actually don't want patients to have their data.

  • 10:52:01

    PAGEI'm Susan Page, and you're listening to "The Diane Rehm Show." Let's go back to the phones and we'll talk to Marilyn, who's calling us from Tuftonboro, N.H. Marilyn, thanks for being with us.

  • 10:52:11

    MARILYNThank you. Good morning.

  • 10:52:12

    PAGEGood morning.

  • 10:52:14

    MARILYNI want to share some information with the listeners and the panel. And I know that this works. If you have something that you do not want to be known by everybody, you want it be confidential, and it's something that has to be on your record that the doctors and nurses have to have access to that information, what you can do is you can always request that that information be written down in Latin.

  • 10:52:57

    PAGENow, is that true, Dr. Selby? Can you request the information be written down in Latin? Because I think that is a great idea, and Latin teachers across the country, I think, would embrace this as an employment project for themselves.

  • 10:53:12

    SELBYWell, we'll have to check first with current medical school curricula to see if Latin is still taught. I -- it's -- I can't help smiling. And I don't have much more of an answer than that.

  • 10:53:23

    PAGEMarilyn, perhaps that's something to think about down the road. Let's go to Tom, calling us from Baltimore. Hi, Tom.

  • 10:53:29

    TOMYes. The subject you guys are discussing is very important. My comment today is that I read in the newspaper -- at the same time I read about this -- what you guys are discussing, about this medical database -- that our attorney general of the United States wants $15 million to expand research into psychological backgrounds to provide to local law enforcement. Is there a connection between what the attorney general wants for his database and what you guys are talking about?

  • 10:53:54

    PAGETom, let's ask Joe Selby. Is there any connection with that?

  • 10:53:57

    SELBYPlain and simply, no. None at all.

  • 10:54:00

    PAGEHere's a question. Here's an email question that we've gotten from Jim, who writes, "When," on the whole issue of asking for consent, "How much will asking for consent skew the data? Would it still be a representative sample?" If you go to asking consent, Eugene Rich, would it risk making the information less reliable?

  • 10:54:18

    RICHWell, I think it depends on the nature in which the consent request is structured. As Joe's pointed out, traditional randomized control trials always suffer from the problem that the patients enrolled are not representative of the patient population. Pragmatic trials that could be conducted in typical clinical settings with more efficient consent processes could be better at that regard. But there still would be an effect on the process.

  • 10:54:55

    RICHI think that there are other mechanisms that we've discussed where the patients could prospectively give consent for the use of the data, with certain safeguards and certain limitations. And each of these steps in the process would make, I think, the database more representative of the typical patients that were being treated.

  • 10:55:15

    PAGEHere's an email from Beverly. She writes, "I'm a registered nurse, and I had both a brain tumor and ovarian cancer. In both instances, I had the opportunity to consent to being part of research studies. I considered both times as a way of paying back the medical advances I was benefitting from, in a way, that was my way of thanking all the people who'd been part of the research I was currently benefitting from."

  • 10:55:36

    PAGEIn the brain tumor community, there are many people who seek to be involved in the latest studies in hopes of getting the most effective treatment. Very much along the lines of what you were saying, Ceci, that you see in attitudes by some people.

  • 10:55:47

    CONNOLLYYes, absolutely. And I certainly commend Beverly for everything that she's been through and given to further the research. And that is very much a common theme in our surveys and focus groups with citizens.

  • 10:56:00

    PAGEAnd I wonder, also, if attitudes are different between older people and younger people. Because I know on many of these things that involved privacy, you see some real changes with younger people.

  • 10:56:12

    CONNOLLYSure. There are definitely some generational differences. You think about certain generations of little bit younger than me, who are willing to post pictures of rashes on Facebook and have a discussion about what everybody thinks that rash might be or something like that. I don't know that I would be inclined to go that route myself. On the other hand, my mother is on email and texting all the time.

  • 10:56:35

    PAGECeci Connolly has joined us this hour, also Dr. Joe Selby, Dr. Eugene Rich and Dr. Deborah Peel. Thank you all for being with us this hour on "The Diane Rehm Show."

  • 10:56:45

    CONNOLLYYou're welcome.

  • 10:56:47

    PAGEI'm Susan Page of USA Today sitting in for Diane Rehm. Thanks for listening.

  • 10:56:50

    ANNOUNCER"The Diane Rehm Show" is produced by Sandra Pinkard, Denise Couture, Susan Casey Nabors, Rebecca Kaufman, Lisa Dunn, Danielle Knight, and Alison Brody. The engineer is Erin Stamper. Natalie Yuravlivker answers the phones. Visit for audio archives, transcripts and podcasts. Call 202-885-1200 for more information. Our email address is, and we're on Facebook and Twitter. This program comes to you from American University in Washington, D.C. This is NPR.

Topics + Tags


comments powered by Disqus
Most Recent Shows

Revisiting The Decision To Drop The Bomb

Thursday, May 18 2023As President Biden's visit to Hiroshima dredges up memories of World War II, Diane talks to historian Evan Thomas about his new book, "Road to Surrender," the story of America's decision to drop the atomic bomb.