[This Transcript is Unedited]
Doubletree Hilton Hotel
8727 Colesville Road
Silver Spring, MD 20910
Agenda Item: Summary of Previous Day's Testimony
DR. TANG: Good morning. Welcome back to the second day of the NCVHS Quality Subcommittee meeting on Measures that Matter to Consumers. We had a very productive and long day yesterday with a lot of good information. Thanks to Matt for putting that all together. This morning we are setting up to put together the skeleton of our letter of recommendations to the secretary. Blackford and I were thinking about this meeting coming out with findings, conclusions, and recommendations. We can fill in some of the gaps. Susan will have her own notes once she has full access to the transcript, and will fill in some of the details.
The high level findings, conclusions, and recommendations are what we're after for this morning. Matt has kindly agreed to be our scribe in terms of capturing things as we go. Matt, could you please make that a slightly bigger so we all can see it at the other end of the table? Why don't we have some general comments first and then we'll get into these findings, conclusions, and recommendations.
DR. FITZMAURICE: What I heard was that consumers are not using quality measures very much. In fact, they don't know where to go to get them. That's what I took away from Christine Bechtel and some of the others. Yet, there is some very intricate statistical work being done with quality measures. It's the gap between the quality measure and what one as a consumer needs. They need to make choices about providers and it doesn't get down to the provider level on many cases.
DR. MIDDLETON: I share Mike's sentiment and the thought comes to mind for the paper. I think we could perhaps echo Don Berwick, again, and say consumers are flying blind. It's not only the docs.
DR. COHEN: I felt that two underlying themes emerged for me, I thought it was a phenomenal presentations. One is that we need to give consumers more actionable information so they can make better decisions. The other piece that I hadn't focused on is, we need to work on providers, too, so that they get it and they have incentives to do the things that consumers really want and really need.
The third theme is that we need to connect the providers to the consumers in a way that the interaction is powerful and meaningful to each.
Finally, there is a lot out there. Part of this is that there is not a lot of rocket science that needs to be done. The information exists – it is just not packaged or presented - by and large the information exists, or techniques to collect the information exist, but there hasn't been any incentive to collect them and share them in meaningful ways.
DR. MIDDLETON: I was struck, just to follow Bruce's thought, I think the plethora of information that is available in non-standard forms or non-comparable forms. I was very impressed with the Consumer Protection Disclosure Project, if I got the acronym right, which really laid out ten very concrete steps or criteria, that we might emulate. We might suggest standardization around an MDS of some kind, a minimal data set, or minimal measure set, an MMS, that is very usable to the consumer. To your point on accessibility, wasn't it amazing that the mobile app was the most useful thing to Christine Bechtel?
DR. HORNBROOK: Two things, first, there is a considerable amount of work to be done to teach physicians how to use functional health task measures in their own practices. They have to understand the nature of the instruments being used. They have to understand the properties of each item of the scales. They have to be coached in how to interpret patterns of scores within a patient and within a patient across time and across their panels. It takes a while for that to happen. It's going to be customized to the kinds of patients that doctors see because they need to understand that when he looks at patient “x” or “y”, how does that show up in the questionnaire or does it not show up?
I've talked with David Feeney a lot. He's been doing this with various practices. He has some experience where physicians will eventually learn how to use, in this case, the Health Utilities Index, and then when he pulls it out because the research project is over, they call him back in and say that they want to use this, continue to use it, as clinical. It was in an advanced cardiac practice with transplants.
In any case, it is clear that when we go to the Permanente Medical Group, they see this stuff as a lot of extra cost and a lot of extra work. The physicians can't be bothered. The burden of proof is on us to show that physicians can do a better job if they actually have objective measures. You can make a case to the medical group management because they want to look at how all the physicians perform in functional terms. You can get some of the leaders to see that they need an objective measure of output, which a functional status measure gives you, especially if it's preference weighted. You can then compare across your physicians, their modules, the clinics, the departments, the disease experts of disease “y”, and looking at cost per quality, if you have some way of asking about relative productivity contributions to the output of the overall healthcare system.
That is, in itself, a whole set of technologies that economists fall into very quickly, but physicians just shudder - it bounces back off, because they don't want to be seen as economists and they pride themselves on the art of medicine, not necessarily the science of medicine. All the informatics revolution essentially is showing how much of the cottage industry of medicine is falling away because we can now view all the things that happen that we've never been able to view before, except by paying lots of money to abstract the chart.
Then, on the other side, of course, is convincing patients that these instruments are in their best interest and answer the questions honestly. In many cases, there are lots of examples, starting with the OASIS and also with some of the Medicare questionnaires that Medicare is not mandating, you can get response rates of 95 to 98 percent, if you allow some other some other surrogate to answer the question, too, in the case of people who can't answer it. If they understand that this isn't going to affect their benefits or their care, it's going to the doctor, they're talking to their doctor, they're talking to their health plan, you don't get a market research type response rate, you get people who are actually doing their healthcare.
I'm on a project now doing questionnaires across 60,000 dual-eligible's in Kaiser nationally, it is part of this special needs program where you get an extra bunch of money to do a lot of extra things for these patients, but, of course, the money you get isn't necessarily anywhere near where their needs are. They are filling out this short health status questionnaire at very high rates since it goes in to making a difference to their care.
You can find special populations where this functional health status assessment is done routinely, well, and it makes a difference. You have to then move it down into the regular population, including kids. All I can say is, how do we convince the mainstream who are mostly healthy, and chronically ill patients, that this functional status assessment is in their best interest and is a way to make the health system accountable for what they're getting and what they're paying for. That's the burden that we face in this room of trying to put a little bit of light into the consumer side of things, saying that we need to ask for this kind of stuff. We need to be thinking of routine health questionnaires as a necessary element of high quality care, not just something that some surveyor is putting on top of me.
DR. MIDDLETON: It's a vital sign.
DR. HORNBROOK: As a vital sign, yes.
DR. WARREN: I had a couple of responses from thinking about this last night. The first point I want to make is some response to something that Mark said in that we need to get patients to fill these questionnaires out accurately. It's the word “accurately” that is in question. Thinking about some of the folks I know, they would fill them out accurately according to their perception of health. The other key word is “perception”.
A lot of people perceive that they can do things and they have certain functions, but if they were assessed objectively by a healthcare professional, that assessment would not match. I think that leads us to a lot of the problems that we see in why health professionals don't always trust patient-generated information, when they should trust it because it is what the patient generated. It is what their perception, preference, their lifestyle, has said is real for them, even if it is only real at the moment of filling that out.
Even the opportunity to take that moment and to create that cognitive dissonance between the patient's perception and the reality as viewed by others, is a teachable moment for them with healthcare. I think we need to start looking at that.
The other thing that I realized is that right now within health profession's education, all of the certification bodies are re-looking at their criteria to accredit schools. There is a nice sweet spot to get into the dialogues of the different organizations that accredit to talk about the importance of functional assessment, quality, et cetera. They're already looking at quality, but it's more the traditional view of quality, it's not the quality that we heard about yesterday, of a more consumer-driven and what people would like to have and things like that.
The last one I had is always one that I get concerned about and that is the money. If we continue to pay people in 15-minute increments and we continue not to pay for wellness for functional status assessment, et cetera, nothing is going to change. So we will continue on as we go. Somewhere along the line as we do all this we consider how we got to where we were with reimbursement and where the sweet points are for changing that whole strategy.
DR. CARR: Sorry I missed yesterday. I read the Powerpoints and they were very interesting. From my own experience, I would say that a key thing is the understanding that receiving these questionnaires requires a response and the response may be a fundamental change. To illustrate that in one of our hospitals that is the nicest place you would ever want to be, St. Anne's Hospital, they are very embracing and very kind, they greet you at the door, everything is very patient centered. The results were coming back not representative of what they thought they had done. They did a drill down and they found that there was a difference between what the patient said and what the spouse said, not the children or cousins, but the spouse.
With that discordance, they went back to figure out why that is. When they looked at how they organized their day, all of the education, all of the planning, all of the meetings that talked about next steps, occurred in the middle of the day when the spouse wasn't there. They realized that they had excluded the spouse, and therefore when you ask if you got discharge instructions, no, did you understand this.
The result was that the entire hospital looked at everything they did and found one way in each department that they could incorporate a spouse into whether it was going to MRI, tea time, or meeting with the doctor. It profoundly moved those scores. I think that as we do all of this there is an obligation to commit, that once you hear it, you are going to do something about it. What you're going to do is going to move you out of physician-clinician centricity and really embrace the patient and the family.
DR. COHEN: Just to echo that, one of the profound themes yesterday was we need to be listening to the patient more. Anything that we design needs to be tested, verified, confirmed, developed, from the patient's point of view. A lot of times, we think we know how to make these decisions when we are missing something that is so routine and obvious, the way people receive information, what activates and motivates them to change their behaviors, and how they understand what's going on. That's a fundamental, almost culture shift in the way that we do business, the system does business.
DR. TANG: Let me pick up from Justine's anecdote and suggest one way to respond. It's because not every support person can be there at all times, but technology could be such that a patient could be broadcasting. Think of it like a blog, let's call it a blog-- and keep people informed, which I think is what this spouse was saying, one, didn't know, and two, didn't feel part of it. That's just an example. I'm not trying to make the solution, but it's saying how can you respond to something and make things better without saying that we have to have meetings where everybody is always there which isn't practical.
To follow up on what both Mark and Judy were saying, I think that towards the end of your comments Mark, you're saying that we have to capture and present information that is realistic for humans to digest. We can't blame either the physicians, saying, oh, you should be able to interpret all these scores and code in the way that the person was doing with ICF just with training - it's not possible. I think the onus is on us to figure out how-- and yet, we heard that there isn't a general functional status measure that is meaningful to people. You have to get it specific.
How do we standardize even the specific so that it looks the same way in each specific case? That's an example of a challenge for us, for the clinicians, and the same thing on the patient side, you also mentioned in your example that there was just a few short questions. We have to do that for all the things. In a specific way, we need to be able to capture information about the patients' functional status in their specific area and present it in a way that the clinician can consume it just the way he or she does. Do you see what I'm saying? That's our side – pointing to our data side. We can't blame the patient for not submitting or the doctors for not consuming the stuff when it is not consumable.
DR. MIDDLETON: Just on this thread, I worry a little bit though, Paul. I think we have to acknowledge that what we heard yesterday is that there is a whole bunch of different stuff that the patient is actually interested in. The title of our meeting is “Measures that Matter”, and we really had the consumer in mind.
Perhaps there is are three circles that have the provider-centric measures of their care, the consumer-centric measures, and then sort of population or societal measures. Maybe we need a new graphic.
MS. GREENBERG: It is like the MHI Report.
DR. MIDDLETON: It does harken back to that very same vision and we need to understand which is which, and where they are. I think one of our challenges is going to be to acknowledge the provider-centric challenges in all of the ways that you just did, frankly, but we need to radically promote the notion of these measures that matter to consumers. It was Dave Lansky and Joyce Debow, the basics, staying healthy, living with illness, getting better, changing needs, major performance demands consistent across all the clinical practice, patient experience. Joyce talked about outcomes and episodes, shared decision quality, physician performance, and cost. That's a different class of stuff than we usually measure the physicians by, ourselves by. Let's not forget. I think we have leading examples from across the country where this is happening, and if we could borrow and promote that to the national level, I think we will do justice to the cause.
MS. GREENBERG: I agree with – as usual – with what all of you have said. Building on what you said, Paul, I think there is a conversation that is not taking place, clearly. We heard that there is a lot of information out there. It's not standardized, that's for sure. In fact, I meant to ask Heather Smith, and I didn't get a chance to, about the three setting-specific tools that CMS is using, that we have known for years now and in fact, if there was legislation, I thought, requiring them to harmonize those three, OASIS, the MDS, and the other one. It is a home health, it is nursing home and rehab.
As far as I know, that still hasn't happened. Even within the Medicare population where they get at some level all the data, you have three different tools here, and we know that the same person goes through all of those. In fact, hopefully, they will. They will be in the hospital, they will go to the nursing home for rehab, they will get home health at home. That's the way it is supposed to go if you're making some progress. The lack of standards is really severe.
Also, I absolutely agree that it's not a question of who is at fault or who is right or wrong, in the sense of some of these tools and measures have been very well studied. It's that you need all of these different perspectives and then they need to be able to communicate with each other because in functioning, you do need the clinician. There are clinicians whose whole training is to assess functioning. You don't throw that out the window. That plays a role.
At the same time, the person who is or is not functioning is the one that knows how it is impacting on his or her life. You need both. One isn't right and one isn't wrong. They need to inform each other. But there needs to be some kind of conversation when there is a disconnect. That's where we get to the incentives. There is really no incentive, as I heard it, in the system for this communication to take place. It doesn't always have to be the physician, it could be another clinician. It could be electronic tools. There are a lot of different ways. The way people were talking past each other was phenomenal. That's what I heard most, well-meaning people talking past each other.
DR. COHEN: So how do we make sense of all this? I remember when we were discussing about what we were going to do these two days and we came up with the theme, measures that matter. Our incentive was considering some use cases. Essentially, a doctor tells me my PSA has doubled, or it's time to enroll in a new health plan, or I am in a plan, how do I choose my doctor, or my cholesterol is elevated, what do I do?
I'm thinking, given what people were saying yesterday about understanding narratives, if we could lay out some use cases from all perspectives in terms of what information, what standardized information the consumer should be getting, what the employer should be doing, what the plan and insurer should be providing, what the physician or providers should be communicating. Here are the tools, here is the information, here is the way to provide this information to deal with these use cases that would evolve into a story about what kind of information is available, how to communicate it, and what all of the participating partners need.
DR. TANG: Building on your case, Marjorie, in terms of all the instruments, one of the things that Heather talked about was her computer-assisted-- a way to branch it. So you can see how some tools are built for specific assessments and you can have it all in the computer and then you can get the appropriate one as one way of “harmonizing”. A lot of people think that we need to combine them all. Maybe you don't, and you just have to apply the right instrument to the right problem.
DR. HORNBROOK: Paul, there are two dimensions of computer assistance. John Ware uses it to improve reliability. He asks one question on each of the domains and depending on how you answer, he knows he has to drill down further to figure out where you are inside that larger scale, but you have a scale of responses. Of course, he knows from item-response theory, which of those response categories are really broad and which ones are narrow psychometrically. That is one dimension of computer assistance.
The other dimension of computer assistance is what I would like to think of as function versus symptoms. If you ask a prostate cancer survivor about incontinence, if you ask the question right, you will get a number of how many times there is an accident, or how many pads he uses during the day, or whatever. You still don't know how that impacts his ability to go to work and ability to do things. We always face that bouncing back and forth between the clinical reality and the functional impact. John Ware would say that the SF-36 or any of his questionnaires-- eventually that symptom becomes severe enough to affect his ability to go to work, ability to walk around the block. The functional patterns will pick up a clinical problem. When you ask patients, they want their headache fixed, the pain fixed, the range of motion fixed. Sometimes, they don't see the relevance of this high-level questionnaire. When dealing with their physician, has a very specific symptom. If you go into your urologist and ask about the fact that you have to get up three times a night to go to the bathroom, they will give you a questionnaire that he uses inside the office to decide whether you are worth treating or if you're not quite there yet, because it's not bothering you enough. Some doctors might use clinical symptom indexes to figure out, and every office in Kaiser has the ten item pain scale, the Wisconsin Pain Scale, in it, so that they can write down a number that allows for some degree of comparability with pain across doctors.
DR. TANG: I think you are right on in terms of the objective measure versus does it interfere with your quality of life. We need both. It's interesting, maybe what we need to do is - one of our potential recommendations related to the design of these things. In the paper, you couldn't do that. In an electronic world, you could. It's so much more useful. It answers the right questions for both parties.
The nice thing is that there is a win-win here because what doctor doesn't want to know that their patient feels like they're getting better? This addresses Judy's question about whether it is the dichotomy between what the clinician observes and what the patient feels. What the patient feels is always what matters. What clinician wouldn't like, whatever they do, to know that the patient is feeling better? That's a win. It goes back to Mark's thing, there are two kinds of perspectives, but if there is a way to measure, which we couldn't do in the paper world, and report it so that it could be consumed, that would be wonderful.
MR. QUINN: I think that is really important is the way that people talk about their health is in terms of functional status and their daily living. The way that clinicians talk about health is in a completely different way. To have this Rosetta Stone between the two, as a way of translating back and forth is really powerful.
Even on less clinical measures, such as the satisfaction one, the way the Press Ganey was talking about capturing and processing the key words and using rules to take these comments from the CAP survey and turn them into something that is actionable, that can fit on a dashboard for hospital executives, is an example of that. It doesn't just need to be clinical, it needs to be a way of taking what is produced in a way that is useful, that consumers can relate to, and turn it into something that the healthcare system can relate to.
DR. CARR: I think that is great insight, Matt. We have an objective measure, such as this person has a broken leg. We have a patient reported measure, the pain with this broken leg is level whatever. Really, the third measure is, how much does this matter to pain? The tolerance of pain or the motivation to drive your kids to school or get to work, then creates a hierarchy in which a five can be debilitating for one and insignificant for another.
DR. HORNBROOK: This also means that it is even more challenging when you're dealing with symptom-free disease. If a patient has hypertension or hyperlipidemia, sometimes the medicines make you feel worse than the disease itself. You have to know that. The physician needs to know that there are symptoms going on because of the medicine adverse reactions, or just general reactions, and needs to know what the adherence is and why the patient is or is not adhering. There is a whole set of chemical disease that we are creating, chemical in the sense of how we treat them, that put a real challenge into functional assessment, and patient well-being assessment.
DR. TANG: Your example of prostate is a really good one, picking up on what you and Justine said. One, is how often you go? Two is how much does it bother you? Three is, let me let you know what some of the side effects of the medication are. That's really much more informed consent in terms of shared decision making with the patient. That's your decision and here are some of the considerations.
DR. WARREN: One of the things that we need to be careful about, when you said, what are the side effects. Watching my husband listening to television commercials about drugs is an interesting experience. They will be doing a sleep drug like Ambien, and one of the side effects is drowsiness, and he asks why they would mention that. Isn't that the purpose of the drug? I think that that is something that people tried to tell us yesterday. You have to present it where it is meaningful. If you present the lawyer's laundry list of side effects, that is no longer meaningful. People turn you off. It has to be the key side effects that might be important to you, not the nausea, not the drowsiness, not the all of the things that practically every drug has as a side effect. I think there is that nuance of making it simple enough to understand.
DR. COHEN: I don't think human beings distinguish between side effects and main effects. It's an effect. The concept of side effects might be a clinical concept, but that's not the way -- this is the way it makes me feel, whether it is direct or indirect.
DR. WARREN: If you read the insert and listen to the commercials, drowsiness is a side effect of a sleep drug.
DR. COHEN: I know, but from the person's perspective, it's not a side effect. This is how it makes me feel.
DR. WARREN: That causes people to really wonder about the advice they get. I'm putting it out there. Here is a highly intelligent man who is blowing the whole thing away because it is a ridiculous statement.
DR. COHEN: One other comment. I think it is great that we are focusing on the doctor, direct provider, patient interaction, but I think we need to keep our eyes on the prize. This is a systems issue around what employers, what insurers, and what plans' responsibilities and roles are in measures that matter. That will change behaviors, and communities. When we talk about measures that matter, the focus needs to be on the core interaction, but what helps frame the context for the core interaction are all of these rules and regulations at a higher level.
DR. HORNBROOK: If you drill down into any kind of patient, you can find horror stories. We have been doing cancer-survivorship research, colorectal cancer, obese cancer patients with ostomies, they smell. If you don't look at the folds of their skin to see whether they have a fungus infection, you will think the smell is just poor hygiene from changing the bag. If the physician never asks the patient to undress, never asks the patient to show them what the stoma looks like, you can go for months with a yeast infection that is never detected and the person is driving everybody else away because they stink bad. Yet, it is a failure of primary care to do basic humane medicine.
DR. MIDDLETON: I think the other elephant in the room that we may have an opportunity to say something that will wrestle it down, is the matter of cost. The consumer needs to know what his or her exposure is. We saw the different approaches to it and the challenges with it, but we are so far from meeting that objective. If we could do anything, in addition to the stakeholder perspectives, let's get the cost issue on the table and be very transparent about that.
MR. QUINN: I think you are 100 percent right. This is so, so important with regard to the health insurance exchanges. A piece of testimony that really resonated with me yesterday was Lynn Quincy's example of taking a really complicated 100 page description of what is covered and what is not covered in a plan, and working it through three examples. Maybe it was supposed to-- the example where almost all of breast cancer coverage is covered, but pregnancy is not, that is meaningful for some people, but it's not.
I think the piece that is missing there, too, is to say that I'm a male, let's say that I'm a 24 year old male, I don't care about those use cases. How about some use cases that include getting hurt mountain biking, or I was diagnosed with some strange cancer use case, and being able to personalize that. To make it real, and to translate all of that stuff into something that is like that food summary on the side of a-- is there?
The other thing I wanted to bring up was the metaphor yesterday of the thermometer. I recall it as something like, if you don't understand the temperature concept, you can't build thermometers to measure it well. I still don't know how much good research is out there on -- we know some about how consumers search for information online, but how do people consume health information in their daily lives or make decisions on it or do they seek to? Maybe there is the evidence base out there. And not just people who are sophisticated, but people in different groups who have different health traditions, and people with disabilities, how do they navigate the system? Just watching them and really understanding the tools that they could use.
Agenda item: Begin development of findings and recommendations letter
DR. TANG: In order to try to capture all of this good information, this is sort of general impressions. Could we move onto findings, observations, and Matt, if you could create a list so that we know what list we're talking about. These are findings from the hearing yesterday. I'll make a starter like to echo how Mike started us off, which was related to Christine's story. The finding is that even the most motivated, knowledgeable, persistent consumer can't find information useful for making choices in health care. That's a statement. It captures it all. That's a problem. Other findings that we can list?
DR. HORNBROOK: One of the things we should emphasize is something we call plain language, which has an IRB meaning in the sense of non-lawyerly language, it isn't way above the reader's ability. Certainly plain language usually targets something below a high school education, sometimes ninth, sometimes even less than that, as a reading level.
DR. TANG: Try stating it as a finding.
DR. HORNBROOK: The finding is that if you get into very, very intense literature search, you go way above plain language, because everything is written in academic or legalese. There is nothing out there, even for the people who are most motivated, to get it in a way that they can understand it, unless they have a medical degree or a legal degree. That alone, if you are somebody who is a steel worker with a high school education and you are reading at the ninth grade level, what can you do to understand what is out there?
DR. MIDDLETON: A related one, I thought, that is sort of the words issue. I think the other one is a concepts issue. The finding is that I have no problem with the language, and yet I can look at my insurance coverage and I am befuddled. Somehow we have to simplify coverage so that the average consumer can understand it.
DR. TANG: Could I try a restatement of both of those? So the complexity of information makes it impossible to understand health related metrics, be they coverage or quality related.
MR. QUINN: Health insurance literacy versus health literacy versus literacy.
DR. TANG: It is complex stuff. It's not that this is easy to translate, but it is the complexity that is a problem. The net result is that they can't understand it.
DR. HORNBROOK: Part of the problem is that this country is really pushing back against standardized benefits. If we all had exactly the same coverage, you could talk to anybody around the table and share your coverage and learn more about coverage in an area that you hadn't been into yet, but everybody has different coverage. Around this table, I bet none of us have anything that is the same.
DR. TANG: By design, right? That is a hint of what a recommendation could be, in terms of the health insurance exchange has the opportunity to change that.
DR. COHEN: A couple other findings, providers don't give the information to patients that patients need to make good decisions.
DR. TANG: They don't give the information or they don't give it in a consumable, understandable way?
DR. COHEN: Both. That was my sense, they don't talk about quality of life issues that are probably fundamental to the patient. They talk about diagnoses and diseases. Even if they have that information, it is not clear how the patient or consumer understands it. The other one that struck me yesterday is – well, there is providing information versus activating decisions. We don't know how to activate decisions even if we know how to provide information. There was the issue around trust. I don't know whether that is a finding per say.
DR. HORNBROOK: One of the things that we should do, of course, is understand that patients come up against things new, for the first time they get a medicine, the first time they get a disease, but many of us have been living with diseases for years and years. You have a routine going. If you're lucky, it's the right routine; if you're unlucky, it's the bad routine, and you have never gotten on to the right track. That's where you need the information so that you can retune, re-understand why you're getting six medicines a day.
MS. GREENBERG: I think our first 30 or 40 minutes in which we were giving general observations, were findings. Many of them were findings and observations. One thing that I was just thinking about, and particularly thinking about the person from AARP, is that obviously the committee first advises the department, but we know that through its whole history, it goes beyond that. The recommendations, findings, and observations have often resonated throughout the system. We hope that they will. The department's responsibilities directly are for the elderly and more vulnerable populations.
When you link that in with the complexity issue and the difficulty of interpreting all of this, it is kind of profound. If those of us in the prime of our lives, technology, savvy, health literate, et cetera, are having trouble with all of this, think about the elderly population and the Medicare population, and then people with disabilities, people with children with special needs, et cetera. Their needs are profound. They are actually - people with multiple chronic diseases are disproportionately using services and affecting the cost curve. This level of complexity is really a challenge for them. In fact, it's beyond a challenge. It is not feasible.
I remember somebody, I don't remember who, saying that they thought the Medicare Compare website, they had something positive to say about that, but I think that the department in particular, is obligated to think about how the elderly people with disabilities, people with these types of multiple needs, can navigate the system.
If you solve it for them, presumably it will work for everybody else, too. It will be overkill for some other people, but if you can solve it for that part of the population and help them navigate, then presumably others will benefit. It is kind of that universal design concept. If it works for people with mobility and other limitations, the environment, it will be fine for other people as well.
DR. FITZMAURICE: I want to pick up on something Mark said that insurance coverage is different for all of us. It's probably a fair bet that Marjorie and I have the same insurance coverage, that is federal employee plan Blue Cross, but yet, if we have different FSA amounts, one might have $1,000 and one might have $3,000, we will feel differently about the out of pocket expenses, and might make different decisions about providers.
DR. COHEN: I was just going to say another finding from yesterday that I think is really important is that currently there aren't the appropriate incentives for insurers and plans to provide this information to consumers. That is a fundamental finding. Those incentives don't exist for Medicare or for the plans to provide this information.
DR. HORNBROOK: One of the questions, of course, is what is the currency by which the plans advertise? As a recent Medicare-enrollee, I get tons of mail about all the Medicare companies wanting me to enroll with them, even though I'm a member of Kaiser and don't want to enroll with them. It is mass mailing. It isn't a learning in the principles of health insurance, it is a matter of graphics and price. We can sell you this as marketing of a very generic, non-described benefit, and it is the lowest dollar per month that they're trying to get to you. People don't understand what they're giving up in order to get that lower price, until, of course, they have experience with it. Thanks to Medicare, they can choose a new program every year.
DR. MIDDLETON: I think the other fundamental finding is, and we have mentioned it several times in different ways, how patients and providers view quality completely differently. Related to that is this idea of giving the consumer more voice. We heard a variety of different survey, HCAHPS, and all the rest of the survey data, but I thought the consumer comment data was extremely interesting. If we are going to ask the consumer to engage more, with more accountability on selection and choice, et cetera, in health care, we really need to give them voice. What is the method of giving them voice? Perhaps it is more of this common based survey research, in addition to refined survey research. I wanted to make sure that we made that observation.
DR. WARREN: Responding to a little bit of Mark and then what you said, Blackford, is trying to get people to fill out these surveys. How do we present them with the surveys? Do they come in the mail? Are they by phone? I can tell you that if they are by phone, I don't even listen to the rest of the blurb, I hang up. As soon as I know that they are not someone that I know, they are wanting something that I don't have time for. I think we need to be very careful about that.
On the other hand, I think it is vital that we include them in the design of these things. As I was listening yesterday, a lot of what was being said was the things that we are beginning to try to do for our clinicians. It's the whole usability testing. I'm not sure we have ever done usability testing on some of these surveys and things like that. I do worry the same as Marjorie that we already have some classic surveys out there where you cannot aggregate data across them because they measure the same thing differently. You can't follow a person's functional status through those.
I was struck also by the man who was talking about how measures were developed in fact and then complaining that the EHRs didn't support them. Nowhere in his methodology was there ever a step to say that we have modeled this to fit an EHR information model.
DR. COHEN: What David didn't tell us was that he developed the measures before EHRs existed.
DR. WARREN: I know, but if we are saying that this is still the same methodology, then I think we need to start moving people to the information system that we have that is out there. We are still not sure of all the ways it could be used. When you put together a database, it gets pretty powerful on what front ends you can put on the front of it and how you can repurpose that data. At least, that is the vision that we have all had, enter once, use many. We could have some of these surveys in doctors' offices while they are waiting. The patients don't know they're waiting because they are participating in their care, filling out these surveys that are automatically put into their chart of the EHR. I think we really need to take a look at some broader pictures of some of the usability issues that are in there.
DR. MIDDLETON: We touched upon this several times, following Judy's thought. Several of the presenters talked about their survey instruments and readily acknowledged what is proprietary and what is not proprietary. We need to say something. The finding is that some useful tools are proprietary, yet we need to say something about what can be made generally available to the country at large in our industry, and used across the board.
DR. HORNBROOK: That is like patents. You get exclusive rights to your invention for a fixed number of years in order to provide enough incentive for their technological advance, but copyrights are forever. You own the intellectual property forever when you copyright it. There is no provision, unless you forget to reapply for your copyright, for going into public domain, unless you donate it to public domain. We are finding people like Judy Hibbard who has taken the Patient Activation measure as her way to fund her retirement. She has made it a business entity now. She has not put it into public domain. As far as I can tell, she doesn't plan to.
That does get at the issue of trying to work out a parallel treatment of providing incentives for developing questionnaires, but also putting them into public domain. AHRQ paid for a lot of Judy's development work. A lot of the work that went into making the PAM measure was federal money. Sometimes you go back to the final reports for those projects, and you can use that version of the questionnaire in public domain, but then she goes out and uses additional private money to make more changes to it and copyrights that new version.
In any case, it is a general finding that we are going to have to deal with in research and clinical care, that interface. Social science research is generating those questionnaires that we need now to deal with patient well-being, patient activation, patient decision-making, and patient experience. Of course, CAPs has been funded by the government, so it stays in public domain, thank goodness.
In any case, I think the other issue that you really need to think clearly about is that all of us react to surveys depending on who the sponsor is. If the person calling you won't tell you who the sponsor is, you hang up. I've gotten into the habit of asking can you tell me who is paying for this, and if they can't give me the right answer, I don't want anything more to do with it because the sponsor knows that if I knew I wouldn't answer anyway. What we find inside Kaiser is that if you use Kaiser Permanente to a member, it's logical why you should know their name and their phone number because they know that you are their health insurance and medical care provider. If you tell them the reason why it is important to ask them these questions, they are very cooperative. Universities and government agencies use that social contract to get good response rates on scientific surveys. If you are a contract researcher with an unknown client or a hidden client, then your response rates start dropping considerably.
If you are doing legitimate survey research that can clearly be convincing to the respondent that it is in their best interest to be a participant, you will get high response rates. You won't get 100 percent, and you can also learn that as long as you have them on the phone, you better give them the questionnaire right then and there because if you allow them to get off the phone and mail them something, you will lose a proportion of those people. If you never get them on the phone, you have to keep going at them from the mail and email, whatever you have a way at them. There will still be people who simply do not pay attention to you. You have that increasing marginal cost of going to the last non-respondent. We face, in getting these data, the same problem you face in healthcare. When does the marginal cost way exceed the marginal benefit of the information you're getting, and you stop? The same way with one more heroic measure, may cost you $300,000, but it gets you one hour of additional life to live on average.
MR. QUINN: I was just going to mention that there was a recent National Academies study called “Health Care Comes Home: The Human Factors”. It was actually Teresa Zayas-Caban at AHRQ who was the lead on it. It was specifically around how we drive usage and safe, efficient, effective usage of all of these consumer tools that are out there, beyond just delivering information but capturing information from consumers.
I read that by the end of this year, there are going to be 15,000 consumer health apps in the iTunes store. This is a proliferation, some of them useful and some not. Today there are no guidelines or standards or anything around what works and what doesn't work.
The other piece, and that is something that is already in the queue of people like me at NIST, to do, not that we have funding. This idea of needing to do usability testing on data captured for consumers and clinicians, this is a parallel issue with the HRs. How do we get information into the system? There was a great AMIA workshop on data entry. One of the challenges of EHRs is how you get the data in the system. A lot of people like to do narrative things. It's unstructured data, versus structured data. This is a parallel issue here. There are costs associated with surveys. What we have seen is that people will share and share a lot, if it is unstructured data, these comments. The question is how we balance getting the data in the system and thinking about how to present it both to providers and consumers. Getting the data in-- Christine was looking for doctors who had a lot of reviews, found some and couldn't find others. I bet, somewhere hidden, people have shared things in threads of blogs and comment series. How do we move that into ways and share that information, the same way that it is on Yelp about restaurants, and Trip Advisor. I think that there is a greater opportunity to capture and make useful unstructured data from consumers than there is inside the healthcare system because that's the way that people, we heard yesterday, make decisions.
DR. WARREN: I was just responding to Mark and several of us are on the list serve lately that has been talking about Elsevier and copyrights and ownerships and whether or not we should allow proprietary tools to be developed with federal dollars. There is legislation out there that is being highly contested, a very active group of healthcare librarians who are working at trying to perform a lobby. Elsevier who are one of the leading publishers out there trying to lobby for the law and the librarians are against the law. I think we need to take a look at that. We are finding some very valuable diagnostic tools are no longer being used because the people who own them have decided to charge for them. They key one is the mini-mental status exam. We no longer teach it to students because the schools have to pay a royalty in order to teach it to students. We were surprised. We quit teaching it to students.
When you look at the amount of research and predictability that is out there on that particular exam, and it is no longer being used, and people are out there now trying to figure out if someone has closetly used it on paper in an organization. If they find out that they have, they could be facing a fairly significant law suit. I think we need to address some of these issues along with this. It is one thing for Hibbard to publish her work, and there is a form that we could use. That's fine as long as there is a form. It should be copyrighted, but it should not be licensed. There's a difference there. You want it copyrighted to protect the reliability and validity of the tool. If she wants to create other tools to fund her retirement, fine, let her do that. The work that has gone out there, that should be available for everyone to use for health care.
I think we need to re-look at that whole model of where ownership is, where we put our intellectual property rights. Every time we look at this problem there is another thread to pull out that impacts the final piece.
DR. MIDDLETON: Excellent point. That could be a major finding from this letter and this meeting, that actually spans not only quality measurement, but clinical decision support and other types of things, that we really seek sharability, but are not yet there.
I wanted to return to one of Matt's comments. The finding or observation that I had after the day yesterday and from my own experience to a degree, is that the consumer is actually assess quality already in lots of different ways, the Yelps, Angie's,, and other things that are out there. Those are naive tools and not tested instruments, and yet consumers seem to be finding them to be useful. Somehow, we need to acknowledge and address that fact that consumers are assessing quality already. How do we either get on that train and help it to be done better, or how do we say that it's okay, but here is how we really think you should do it? I don't know.
DR. COHEN: I think that is a great point. People are making decisions in the absence of the information we want them to be using or that we think they should be using. Once we provide them with the information that we think they should have, it doesn't necessarily mean that they will make the decisions that we want them to make. That's not the goal. I think the goal is to provide more information so that they can consider it. I think that is an important issue. One other theme that emerged yesterday has to do with the sum of these individual decisions and population health in the community. I don't think we should lose that thread. Certainly, the Rochester experience and the comments from that from Kaiser about this is ultimately a population health issue, we're talking about getting individuals to make better decisions, but this needs to be seen in the broader context of improving population and community health. I think that's a key that we need to focus on as well.
MS. GREENBERG: This is perhaps a little creative of me to segue from that statement to the ICF, but we knew it was going to happen. Classifications are obviously relevant to population health data. We have been talking about standards, et cetera. I don't know whether-- I wouldn't say I'm agnostic about it, but at the same time I have to be neutral here. With everything that we heard, I still think there is only one classification or international classification for sure, that does attempt to classify in a standardized way the different aspects of functioning, in quite a broad way, from body functioning all the way to environments.
We were talking also about some of the issues related to licensing and all of that. You mentioned yesterday, Paul, that you might want to-- we haven't really gotten to recommendations yet, but if you're going to make a recommendation, you have to have some kind of observation that there were findings. I throw that out as to whether there is anything you want to say or think needs to be said about the value or use, barriers, all that, to ICF.
DR. HORNBROOK: Just to go back one step, one second. Is there such a thing as a public domain copyright?
MS. GREENBERG: People talk about things being in the public domain.
DR. MIDDLETON: The thing to look at is the creative comments architecture for maintenance of IP. There are different levels in the creative comments framework which I find very useful, actually, for blogging and stuff, as do anyone who puts up good stuff. That may be helpful to us. Whether it has been used or stressed in this way is not known. I don't know.
DR. HORNBROOK: I am just thinking of what we should put in the letter. Is there something in here that we should say something about what is already possible and call it out, or do we need to call for some kind of new legislation that enables this concept? I don't know. It's something for us to think about.
DR. MIDDLETON: Or be more resolute, if it's publicly funded, don't we all get a right to use it?
DR. WARREN: That is the law that is being debated right now in Congress. I can look that up and get the citation.
DR. FITZMAURICE: If it is funded by AHRQ and AHRQ has the clause in a grander contract, which normally we do, then the government has rights to whatever is produced. If you change it beyond what was produced, you can copyright that improvement. I don't think there is such a thing as a public domain copyright, but a federal agency could copyright something and make it available to the public. Any private individual could copyright something, but yet make it available for free to the public.
DR. TANG: Maybe that is one of our-- we could act on that. We could go to whoever funded Judy's work, and go with the last-funded output and grab it.
MR. QUINN: One of my projects at AHRQ actually was a contract or task order to identify, and this was specifically in the health IT domain, it recognizes that the primary dissemination mechanism for health IT research has been articles or white papers, when oftentimes the product of that research is intellectual property. It's code sets or tools that are built with code or specifications for measures, or all of these things, and the project, which I haven't seen the final report yet. The idea was what are best practices and what is possible legally and how should AHRQ internally look at funding, describing it in the requirements of task orders and contracts and grants, with the recognition that we want to disseminate this. Taxpayers paid for it. It's great if an article comes out of it, but the code of the actual module about which the article was written would be great, too. How do we share that? I don't know of anybody doing that today in government beyond some of the data liberation stuff. It's another frontier that I think would be of huge value to the taxpayer.
MS. GREENBERG: I guess my feeling on this that the committee could definitely articulate the problems that they see, the barriers to things being used, and what the goal is. I think it might be-- I don't know if it's out of scope, but it might be premature for the committee to support one approach versus another, given that we haven't really had hearings on that. It's complicated because, and particularly if it's not under contract, if something is developed under a grant, which is the way of most of the NIH work, which really supports a lot of this. Often people are actually participating in this research because they can then use, promote it, what have you.
I think the bottom line is what are the problems that result in this. One solution doesn't fit all. There was a SNOMED kind of solution that the government essentially bought it out. It probably shouldn't do that for everything. Obviously, you want to lower barriers, but you also have to realize this is complex also. There were people who felt that CPT shouldn't be named as a HIPAA standard, that that wasn't practical, but it was given its use-- what it tried to do through HIPAA was to lower the barriers to the things we used. I would caution you not to get too far out without doing more exploration on that.
DR. TANG: We have to portion our time so we're not spending all of it on proprietary interests. Let me move to another topic, which somewhat undermines all of our work in quality measures, which is what the consumer union said about coverage concerns and out of pocket costs trump quality. They do not get to step four when they get stopped at step one. We just have to deal with this somehow. It is still probably in this standardized realm, and we may come back to that as part of our recommendations, but we have to cap to the fact that it won't get to us.
DR. COHEN: I think, well, I don't know if it is a cup half full or cup half empty. The research showed that consumers were looking for value. It's not just a bottom-line cost. We need to frame cost in terms of value. Quality is a key component of value. That's the path we should try to encourage.
DR. TANG: They didn't even get to that line because they were so befuddled by not even understanding coverage, that was problem, which is solvable. That's the cup half full. Let's recognize that we don't get to where we want them to get to.
DR. CARR: I've just been following on themes that we've been talking about and I want to come back to two things. One is engagement with the individual on acknowledging a problem. The second is engaging on what to do about it.
I go back to what Mark said earlier, when you think about acknowledging the problem, there is what the doctor observe or know from data, what are the symptoms of the patient, and how much does it matter to the patient. I was struck by the hypertension thing because there the doctor is very concerned. The patient has no symptoms and may not be concerned. The ideal is that if you have a broken leg, I can see it and feel it, you can see it, and we all know it is a problem. There are all permutations on that.
For example, there may be something that the doctor-- chronic fatigue syndrome, maybe there is nothing that the doctor can observe, but the patient feels it and it matters. Or there might be something where there are no symptoms, the doctor doesn't see anything, the patient doesn't see anything, but the patient is worried about something that might happen. That triad, it's not just the signs, symptoms, and doctor's observations, it is how much it matters. That's acknowledge the problem. I think there is sort of a similar triad on what to do about it. It is the cost, the benefit, and the risk. This is the cost, this is what is going to get better if you do it, this is the risk of not getting better or the risk of harm if you don't do it.
Somehow, we think in a two-dimensional world, and somehow this third dimension of how much it matters to the patient's way of thinking. I put that forward as a framing device, because I think without recognizing these third dimensions, it all seems so simple. You have hypertension, you should take this medicine or we have a solution, this is what it costs, this is how much better you'll get. Some of the recent chemotherapy costs $100,000 and you'll get three months of life. It's the value proposition for that individual as to whether that is worth it. Or, it might have a side effect that could kill you in two weeks, is that worth it?
DR. HORNBROOK: I just want us to remember that technology is moving very fast and medicine is moving along with it, not just new medicines, but smart phones. The day of having to prick yourself for your blood sugar is going to be gone pretty fast, because you can put your finger on the smartphone and it will read your blood sugar through spectrographic analysis of light waves going through your finger, the blood in your fingers.
Intel has also worked out putting accelerometers on your wrists and your ankles and then connecting those to the web. They have worked out the most important thing which is the program that translates all those movements that the accelerometer picks up into some interpretations that have medical meaning. They developed this for professional sports training, too, but they are looking at it for increasingly frail elderly. How do we know somebody fell when nobody is witnessing it? How do we know when somebody is having gait problems? How do we know when somebody isn't lifting their arms above their head anymore?
Eventually, we are going to have this and it's not too far away. We're going to have to decide, again, with healthcare costs who are the people who will benefit the most from this technology because we can't give it to everybody.
DR. MIDDLETON: I love it because the research question has not been addressed. We are talking about a lot of innovation and a lot of things are happening such as 15,000 smart phone apps. Just show me the study where a diabetes module on the web reduces one hospitalization, then every insurance company should pay for it. There has been some interesting stuff.
MR. QUINN: Do we want to make an observation about the timeline for the health insurance exchanges? Seeing that in black and white, that states have to have their-- say that they're ready to have a plan on January 1, 2013, is some pretty scary stuff.
DR. TANG: And put the corollary, which is otherwise the federal government will stand it up, is that correct? That is an opportunity that will come up, I think. Let's take this opportunity to move over to conclusions. Think of that as one level up that can-- we must have 20 findings up there-- one level up that can make a statement. As an example, let me summarize the one we just finished which is the proprietary. It is a combination of the lack of standards and proprietary interests that inhibit, or some feel even prohibit, the development of useful, comparative information with economies to scale. The reason that I added economies to scale is that if we do a bunch of these one ups, neither the consumer can deal with it, nor the clinicians. The only way is to generalize it, and you can't do it if everyone is protecting their siloed interests, especially if it has been paid for by public funds. That notion that we need it standardized, harmonized, and deal with the proprietary interests.
MR. QUINN: Would it be too far to say lack of federal funding has resulted in proprietary tools?
MS. GREENBERG: Actually federal funding has resulted in proprietary tools.
MR. QUINN: But only taking it so far. We heard yesterday the question around whether or not there are electronic versions of these. No, we have the one that was published in the paper, but we don't have the one that fits into electronic tools.
DR. TANG: Can you finish the sentence? So I said lack of standards and proprietary interests inhibit or prohibit development of useful comparative information because it is the useful comparative information that people were asking for.
DR. COHEN: I actually think we should separate these into two. Lack of standards and the proprietary interests are very different concepts. Would it be clearer if --
DR. TANG: That is fair.
MS. GREENBERG: Getting back to what Matt was saying, although there certainly has been federal funding, and I'm thinking of the PROMISE Initiative, there hasn't been a lot of, comparatively, there has not been a lot of funding for research into the types of things that matter to people, whether it be functional status, quality of life, all of that. If you would compare-- you would know better than I do Mark, but I'm thinking back on the report of this committee in 2001 on collecting functional status information. There was a whole research agenda that the committee laid out. None of it, for the most part, was every funded. It wasn't a multi-million dollar research agenda either.
In other areas, too, there was some really interesting stuff that was presented yesterday about what value people put on different things. Comparatively, I'd say, if you compare the funding that goes into disease, disease-specific funding or whatever, it's a drop in the bucket. I think in that sense, I agree with Matt.
There has been a lot of work done on case mix, but not a lot on how functioning, if you have functioning information or functional status information, how that impacts on case mix. We had a hearing on that when Lisa Iezzoni was chair of the Population Health Subcommittee, and actually the second day of it got snowed out. Some of the research that has been done that I know has been done in other countries, but we're not putting a lot of research dollars into this right now is my impression.
As I said, there is this RFP that the Defense Department has put out, and it is really the first one of its kind that invokes the ICF in its relationship to assessment of functioning, et cetera. ICF has been out, it was approved by the World Health Assembly ten years ago, or more. There isn't a lot.
DR. TANG: So is the conclusion then that there is one, that there is an inadequate number of validated instruments in functional outcomes. The cause, you are saying, is the lack of funded research.
MS. GREENBERG: There are a lot of instruments, I think, though that is where you get to the lack of standardization.
DR. TANG: It is more the lack of standardization.
DR. CARR: I am wondering as I look at these conclusions, the first thing we might want to say is that there is a lack of standardized instruments. The lack of standards inhibits them and contributing to the lack of standards is the proprietary interests. I think our biggest concern is the lack of standards and we're speculating on all of the things that contribute to that.
DR. TANG: That is what that says or no?
DR. CARR: Well, the first one, you jump right in to proprietary-- before you say there is a lack of standards, you are saying why you think there is a lack of standards. First, there is a lack of standards. Second, what is the impact of that lack of standards? Third, what is contributing to that lack of standards?
DR. HORNBROOK: We didn't ask for anybody from the European scene to come and talk to use. In Europe, there is a very strong tradition, of course, of trusting government that allows them to focus on cost effectiveness and cost benefit much better than we do. They have preference weighted health status instruments that cover generic, like the health utilities index, but a lot of EQ5D, and a lot of the other disease specific indices that come out of Europe also have preference weights. There is an anathema in this country to developing preference-weighted health status industries which makes us very different than the European system. It goes along with the fact that we have comparative effectiveness research, not cost-effectiveness research. That is the resistance of the conservative side of the country to look at the fact that we can't afford everything for everybody.
MS. GREENBERG: I think the comparative effectiveness actually forbid the legislation. That whole area of preferences is an area unto itself. I know the committee has dabbled in it a little bit. It can be kind of risky. There are issues, obviously, and you know those better than I do, Mark, about preferences and can devalue a less than desirable state on some peoples' part. I know the disability community has some issues with that. You're right, the focus is quite different in Europe.
DR. TANG: So is one of our conclusions that we haven't been able to reach and engage patients in this whole area of using tools? Is that a higher level?
DR. HORNBROOK: The key concept overarching that, how do we help patients make valuable judgements? How do they know when they are getting a good value when they're dealing with their doctor, dealing with their insurance company? How do we help them put all of this complex system and medical data into an overall sense of worth and value?
DR. TANG: I would generalize a little bit more than just talking about value. That's one of the ways. Our problem is that we need to reach and engage patients-- that was used yesterday-- patients and consumers with the information that they need to make health related choices. Then we can drill down on how we engage them and value propositions, what are those, et cetera, experience, et cetera. That is one of the things that we probably have to solve in our recommendations.
DR. CARR: Following on the whole comparative effectiveness research, because I know I have been queried by any number of folks about why this is not effective, and now that we have done all of this comparative effective—I think we're missing a middle layer. We are directionally correct. We generated information, but we don't know how to dumb it down and also how to—that's even pejorative to say, how to simplify it into themes that are usable by individuals. It is really that middle layer. We have really smart researchers up here and we have really motivated patients down here, but we don't know how to present the data in a proper juxtaposition.
DR. COHEN: Stickers on cars that have miles per gallon so people can compare cars.
DR. WARREN: That was a slide yesterday.
MR. QUINN: The issue of preferences for people and I think preferences in the context of comparative effectiveness research or patient-centered outcome research are slightly different in that people flock to sites such as PatientsLikeMe because they want to interact with people like them. People self-organize into online communities of people like them, whether it be socio-economic or shared interest. They're looking for information that is relevant to them and not the average person. On one side, it is a scary thing for us to go down the line of “this treatment option is better for these kind of people”, but people make those decisions. People say that their friends, people like me, families like me, make these decisions about health coverage. They do it informally today. The more we can support those natural interactions, I think it's understanding how those people make decisions and doing research on how they do that and how they organize themselves, supporting those decisions with parameters that allow them to personalize.
DR. CARR: But it is interesting, those are very narrative sort of things, but when you think about it, Amazon does it: people who read this book also liked this. If you look for a hotel, if you are person for whom money is no object, you are only looking at the five-star hotels. If you're a person on a budget, you can stratify, and that is the middle where of sort of marketing people who understand how to take these complex concepts and put it into strata that a person can go to and they can peek over what it would be like at a high cost hotel or maybe there is a bargain, but actually, I'm limiting myself to these choices.
DR. TANG: Can I just add – Matt, you changed the words of number three. I don't think communicate includes the concepts of reach and engage, because engagement was one of the biggest points in one of those things. Then we can address what Justine and you just said about engage them in ways that matter to them, and that means that it is more specific rather than general.
DR. MIDDLETON: It is about time for me to say my usual provocative and outrageous comment. I'm going to suggest that this language is throwing a lateral pass. We're not throwing a home-run in anyway. We have said a lot of these kinds of things before, and I just went back and checked the last letter and prior letters. What is missing? I think Justine was hitting on it a little bit and others have as well. What is happening out there is very different than what we're talking about. What if there were a place where quality data were aggregated on behalf of the consumer across all plans and providers and they could allow them to go look Amazon-style or PatientsLikeMe-style. What are the requirements for creating such a resource? Who manages and does it? Do we have a measures clearing house? Yes. Is it functional? Not really. I worry that we're not making the home run pass here. I want to think, given the opportunity with the HIE, with exchanges I mean, and with the advent of an installed base of HIT, let's go big, even if we're painting a picture that is 10 or 20 years down the road. I think we're still doing laterals, and I want to go farther.
DR. COHEN: Not only for consumers, but the two other targets for this are the physicians, or direct providers. They need and we need to integrate this information that is meaningful to their patients and to their routine practice. The third piece is that we need to provide the incentives for the plans and the insurers to promote and include this information. Those are the three legs of the stool if this information is going to be properly assimilated.
DR. WARREN: I think you need to disseminate it to more than just the position. It needs to be all health care professionals, because they all have a piece of that quality. I'm not sure we always feedback the quality to the rest of the professionals, to pharmacy, to nursing, to physical therapy, to speech therapy, and whoever else is involved.
DR. CARR: Again, following on the theme, if you think about all the effective websites, and before you even set expectations of change, take what we have today. If you go on—I'm a 20 something year old with a high deductible, but my doctor says I need to treat “x”. Now you can segue into comparative effectiveness research. You can do it with mind-body meditation or you can take this medication or you can have this operation. Here's the cost. Now you can say that you've made your choices and you'll be fine, they can do mind-body because they're into that.
We have the moving parts, and before we get to things we don't have, let's focus on what we do have. Even if we made, back to your use cases, a couple of different use cases of who I am and what would be my choices. Like my choices will be framed by the cost, or my choices will be framed by proximity geographically, or by academic credentials. Whatever my preferences are, you ought to be able to slice the data, and that data exists. We could get Todd Park on this.
MR. QUINN: Reading back over our previous letters, one of the things that we observed was that measures, measures everywhere, but not a drop to drink. As I thought about this, we are in a similar situation. I think something that resonated for me was that not only do we need similar centralized coordination and harmonization for consumer peers, we need harmonization between and among the groups of measures. For health care systems-centric measures there is a corollary on the consumer side and they need to be able to translate and talk to each other. Without some sort of coordinating body or some sort of central actor to drive this, it is going to continue to proliferate and do so in inefficient ways that, frankly, is not going to support—just imagine 50 different health insurance exchanges.
DR. TANG: I want to not let Blackford's challenge to think outside the box, go past me. Here's a suggestion for a recommendation, and it piggybacks on the insurance exchange deadline, that federally supported health insurance exchanges should be required to set up standardized benefit tiers, which is what Mark was saying, with clear, easy to understand comparison charts, and that deals with the coverage problem that we couldn't get beyond, and to require what I'm calling next generation measures that matter to consumers in standard form with easy to read and understand comparison charts. In a sense, it is taking advantage of the moment, which is that we do not have a way or the incentive or the motivation or the stick to standardize in the environment, but we do. In 2014, there will be a number of states that will not stand up to exchange themselves, and the fact that the federal government is paying for some of this gives it the contractualability to standardize things that are in its and the country's and the consumers' best interests, meaning the benefits, but also the tools that can help people—one, you set up the standardized measures, but then you can write the tool that you just spoke of. Without the standardized measures, we can't write the tool.
DR. CARR: I would challenge our traditional concept of measures, which I think we have thought a lot about, the NQF, numerator, denominator, exclusions, et cetera. As we heard, miles per gallon is a measure. Number of procedures done per year is a measure. Distance from a big city is a measure. I think we have to get away from the traditional idea of measures because those often are confusing and confounding. A simple number or percent is also meaningful.
DR. TANG: We didn't capture it as a conclusion very clearly, but multiple people, almost everybody on the consumer, what matters to me making decisions, talked about how it needs to be specific. Unlike the city and highway driving, which is a nice goal, what we have heard is that it's just not going to work for an individual. That's what I heard.
DR. COHEN: But there is a way to deal with it, not just charts, but interactive software. I am a blank year old, blank male female transgendered person, I have these issues, these problems, describe yourself and then here are your options for what plans would work best for you. I'm concerned about, I want it to be on public transportation. It wouldn't be difficult to design--
DR. CARR: It is like Zappos shoes. I want medium heels, I want black, and I want this designer. Or, show me everything that is black. I think that's right. Even, my dates are flexible when traveling—I need an operation, or I need something that might be fixed but I can take two weeks off or I can't miss a day of work. I need childcare, I need--
DR. TANG: That is what the market can give us provided we give them the standard inputs. For example, let's say you are city-driving. You can plug in your GPS, which knows exactly your driving habits, down to the acceleration, you could plug that in and get your predicted MPG.
MS. GREENBERG: Are we just talking about coverage decisions?
DR. TANG: No. Remember somebody gave essentially the traditional package for cost, for appendicitis and something else? That could be, if you plug in all of these inputs—now, that was having to do only with cost, but let's talk about performance in the care spectrum—you could get for your health behaviors, here is what it would cost.
DR. CARR: We should talk to Todd Park on Friday because their datapalooza is coming up June 8th or 7th. If he has somebody who is a developer who would like to take some of these concepts, we could work with them and get an example.
DR. MIDDLETON: I think four is on the verge of profound, Paul. I really do like it, and it does leverage existing activities and the funds flow and the regulatory opportunity, but I guess it would be worthwhile to dwell here for a minute. To achieve that vision, there are a number of things that really have to go into it. Matt, I think even number five is part of it. We need standardized benefit design. We need standardized interoperability of those benefit design statements. We need the exchange of all of that information into a resource or repository where the consumer can access it, just details under the covers there.
DR. WARREN: I am confused by what is meant by “standard benefit tiers”. Does this mean that we want to dictate to the insurance company what benefits they provide?
DR. TANG: No. My understanding is that insurance exchange will set a floor of coverage. You can add all of the enhancements you want. Let's pretend you have three floors, and they might have different deductibles, something pretty basic, and then people can deviate. Everybody would have their comparison chart against the floor, bronze, silver, gold. The consumer, one, can chose if they are high deductible or not, and then go off, and what happens is--
DR. WARREN: I am coming back to the complexity that we had in the standard subcommittee of defining what is a plan. I keep going back to that. Is the plan the contract that the individual signs, which would have deductible and all this? Is the plan the broad plan that the employer negotiates?
DR. CARR: It has to be patient-specific, and that's what is missing. You sign up for something, and when you go for whatever the thing is, they don't cover it, or there is a co-insurance, co-pay, co-premium--
DR. WARREN: If it is patient-specific, then I'm still not understanding how one would go about specifying tiers. It just seems to me that it is an enormous issue of complexity.
DR. MIDDLETON: Let me rephrase what I think I heard Paul saying. I think the idea is that first, we know tiering is happening all the time. Every plan designer and any benefits manager has a negotiation about what are the products for the plan, what are the benefits the employers need. There is tiering going on all the time. It's not just being done in any standard way with any standard inputs, and it is not interpretable to the end user at all.
If we said that here is a parts library with which you design plans and benefits, somehow at a simple level, and it doesn't have anything to do with competition, because you're saying here are the parts and you can still compete, but at least it would be interpretable.
DR. WARREN: I would suggest if you talk to AHIP, which is the Association of all groups, they may have an informal menu of here are the widgets that go into the plans. Here are the things that we have traditionally put into our negotiations. That might be a place to start. I'm just trying to think how we get there because when I read four, I don't even know where to start. I don't know the meanings of some of the words. I think people may push back on benefit tiers, but I think if we collaborate with the industry, they probably already have it, just not in a formal way.
DR. TANG: Let me try one analogy and see if it works for you. So with Dell, you basically get tiers. You sign up for this machine with this horsepower, or these three, and then you add on. You are basically knowing that you're getting a games playing machine or a basic—that's the “tier”.
DR. MIDDLETON: Let me rephrase it slightly, because Paul, what you described there is Dell's benefit plan. Gateway may have a similar but slightly different benefit plan.
DR. TANG: Yes, but they are all defining the game tier and the basic tier.
DR. MIDDLETON: No, that is where I differ. I think actually they are describing different benefit plans, but they may be doing it in a common way. I actually don't give a damn if we design the same benefit plan for everybody, it just has to be designed in a way that is a standard description so that I can interpret it. I can interpret the Gateway plan versus the Dell plan.
DR. COHEN: When you put in your individualized preferences, it will spit out saying that you can get the Gateway 640M and add on this or you can get the Dell 520--
DR. TANG: The trouble is you don't know the difference between the Dell 520 and the Gateway 640.
DR. COHEN: But you don't need to know if you put in your preferences. Essentially, the program can--
DR. WARREN: If you think about it, you do. Everybody has a basic laptop model that is their bottom line. Pretty much, if you pull up all of them, they have the same stuff, memory and all of that. They're not going to vary too much because then they are going to be undersold, somebody else is going out of market. Then they add on, such as wanting a certain kind of image card or more flash memory, more disk memory, more whatever, more power, a bigger screen.
DR. TANG: The problem is that most people can't do that kind of ordering. It goes back to the complexity. If you gave them ways--
DR. WARREN: But I have never seen the insurance companies give me that kind of plan.
DR. TANG: There are three benefit tiers, and they are absolutely prescribed, and you can go look between the different companies on those three. Then you pick the company you want and you modify tier two, because that's what you compared and wanted from each one, and then you tweak tier two--
DR. MIDDLETON: Somehow that is anti-competitive feeling. Why as a plan, how do I differentiate myself by the options?
DR. COHEN: I want a plan that provides me gym memberships and $10 co-pays for my medicine. I design my benefit plan and see what companies will match that need and how much it will cost me.
DR. MIDDLETON: It is a standard set of descriptors as opposed to the standard tiers.
MR. QUINN: It is almost like—the approach that happens today is that you are presented with a bunch of choices. I go to the OPM website and it has a list of 26, and I look and there are some decision points. At no point does it say, tell me about yourself, tell me what you're looking for.
I read an article that health insurers, I think Highmark is an example of one, is setting up brick and mortar stores to help people resolve their insurance issues, as well as to answer questions. What if you could go into a store like that and someone, real or virtual, could say to tell me about yourself and what you're looking for. Tell me about your family. What kind of coverage do you need? Describe it. And then say, based on that, here are our suggested top options for you, and here are some choices that you could either go up or down. Here are the associated choices. Here is the network. The plus and the minus is a difference of network, deductible, and these other parameters. Starting with what you want is different than what is available, and matching—there might not be a perfect match, but there is a better one than I could probably figure out on my own given a list.
DR. WARREN: I was just thinking at the University of Kansas, you are given an option between two or three insurance policies, different companies. If we had this match that we're talking about that is with company A and these are the things that are covered this way and I can come in at a couple of different tiers based on what they pay for, deductibles or whatever those parameters are, if I could have that comparison, that would help, but I'm only going to have to look at the three, the KU offers, because those are the only choices I have.
I think that would be an enormous benefit as well as something in a way that Matt is talking about, to encourage companies that are contracting not with the large employers, but with smaller businesses to have these things. I think that would be a huge benefit. I would let me take a look at these because the brochures that we get are not organized the same and you can't tell where the similarities and the differences are.
MS. GREENBERG: I am thinking back to some of the other hearings that we've had, and in particular on the one on the medical home, and the issues that were raised, starting with the first person about wanting someone who could help them navigate the system and make recommendations, treat them as a whole person. I think there was a lot of stuff about the whole person, too. Is there anything in this exchange or with the minimum benefits or whatever, that speaks to that?
Because at the end of the day, you can only give people so much information, and they can only deal with so much information. They aren't experts in all these areas. Once they have selected their coverage it is what happens to them, their experience with that coverage and with those providers. I still think that if people are having to wing it from the point of view of coordination of care and lack of a medical home or the functionalities of a medical home, they are still pretty far from where they want to be.
MR. QUINN: I was just going to volunteer that Sarah Thomas from NCQA, is somebody who really—we could go back to her and get some of this, as well as Glen Nichols, who is a man who is pretty deep in this area.
MS. GREENBERG: I am just thinking that some of the recommendations that we made or observations—I'm not sure that we actually made recommendations, Susan, in that letter. Some of the things that came up in that hearing are very relevant to this. I think that is almost the biggest problem. If you don't have many health problems, you don't have many problems. But once you do, the lack of coordination of care and all of that is the biggest problem people face.
DR. MIDDLETON: I think back to Justine's description or depiction of that quality circle and different stages along the way, one of the things that I suggest we might want to do in our preamble or whatever, is to say that there are milestones that we have gotten to so far, and there are milestones that we see ahead, and how does this fit into that overarching arch, if you will, on progress on quality measurement. I think we are talking about—that would help us ground these observations, recommendations, and assessments.
It's clear to me that we're trying to move providers along in all the interesting ways about value, but we're also trying to move consumers along and we should be very articulate about that, cognizant of what the limitations are with this move, this incremental move, versus what might be the ultimate destination.
DR. WARREN: I was just thinking of when Marjorie was talking about the medical home, many of your large insurance companies provide case management to the high end users of their providers to help them coordinate are they are getting the right care or the right people involved, and to let them know what their benefits cover and what they don't. Many times, people don't tap into that service because they don't know it is there. All they see is that insurance companies are denying them coverage.
Sometimes that may be true, but sometimes it may be that you have an engaged case manager that is trying to help you navigate. If we could pull that service and say that there is one resource that we have, we also have case managers over on the provider side. Wouldn't it be interesting if the two ever got together and started collaborating about how we can help this person through this activity? That might be a really good thing to happen. We have some standards out there that will support that, prior authorization and stuff like that, that already has a place in our suite of standards for supporting healthcare.
DR. HORNBROOK: In Kaiser, we have two kinds of case managers. One is the kind that are standing at the door and they help everybody get through the doors. They are case-managing a transition, like out of the hospital, into the nursing home, out of home health whatever. Then there are case managers who stay with the patient. The patient is a high-flyer and they need personal attention on how they get access to manage their access. There are not very many of those, thank goodness, but it's a very different kind of case management. If we don't discriminate, we get a lot of semantic confusion.
DR. COHEN: Picking up on what you were saying, Blackford challenged us to think outside the box. One is that the first decision is how can we enable consumers to select their health plan? Two is, once they are in a plan, how do we enable them to select their doctor, and it's the same set of parallel strategies. Then three is, once they have a doctor, how do we make sure that the interaction with the doctor-provider, how do we make sure that that interaction works the best? I would set up the parallel structures for number four, both for selecting their providers and then working with their providers to make sure that they're getting the information that they want.
DR. CARR: Minor point, but just with the ACOs pioneered that model of the care manager both in the facilities as well as navigating the transition and then the community-care measures is the direction that we're going. Again, with these recommendations, I think that we need to have two parts. One is, what could we do today with what we have, and it could be realized through these mash of Google people right away, and make us 100 percent better to begin to test, are we pushing the information that you need to make decisions?
The second is the re-education of the healthcare community in understanding all of the things about patients and beneficiaries. I think we can make the biggest impact by trying to take what is available today, the idea of the exchange, put it in there, make it user friendly in a format that's familiar in other ways, and out of that, we will learn what are the important things, what we are missing. Then that informs the education.
DR. HORNBROOK: The HIEs, health insurance exchange and health information exchange, I think the insurance exchange, the most important societal impact of that is that you have a universal enrollment of everybody in your “population” for the first time. Many times, we don't know who the uninsured are. Now they are going to be enrolled on the exchange, if they make at least some effort to get known. That, I think, is the most important social advance that goes on. After that, it is standardized benefits and good coverage.
Health information exchange, I think the most important thing aside from our measurement issues here is teaching people the importance of linking their doctors together. The important thing is to give doctor “x” access to your records from doctor “y”, and make sure that all your doctors can see all of your pharmacy records, which, again, is something that you have to do. The pharmacy doesn't do it for the sake of fun. These are things that individuals because of privacy concerns have to make their own arrangements. It's going to be a whole important part of the education process to get them to do that.
If you get inside Kaiser, all of that is done for you until you take your care outside and then you lose it. If you're inside an organized system, all of these things are done automatically. The whole point of informatics is to give people in fee-for-service the same advantages, in some sense, as we get inside structure-organized care.
DR. TANG: Let me leverage what Justine just said and let's work to closure on quote, number four. The reward is a break. We can start with what Bruce proposed. The different things we need to deal with are coverage and selection of provider. What are we going to leverage about the health insurance exchange? One, we are going to leverage the time urgency, and that's a really good thing to leverage – that is immediate. Two, is the federal government's involvement in that. We don't know how many, but probably half or more. Three, that means they get control over some of the setting of rules. Let's go to the rules. The first rule is having to do with coverage because this is insurance exchange. What do we want to say? Let's resolve this. I'm not sure we can get automatically all the way to where you were saying to put in all of your preferences. What's the intermediate step that could be done? They already were going to have a floor, at least one floor. Do we want to go with that and make it comparable? Do we want to suggest multiple just because to respond to the quote, voice of the consumer? Where do we want to stand with coverage and standards?
DR. HORNBROOK: This is a question of how do you want the insurance market to work. A lot of people will under insure themselves because they have the positive cash flow and do not bet correctly against their future health expenses, especially as they age each year and their risks go up. They may get away with it for five or ten years, but someday it's going to catch up with them. I don't know whether we want to say anything about the fact that the low minimum benefit should not be set down in the basement or the sub-basement.
DR. TANG: We are not in the policy business for insurance per say. Let's figure out what it means to the measures that matter to the consumers.
DR. HORNBROOK: Then you should look at the Checkbook approach. They take the federal health plan and put it out into, as far as I can tell, relatively simple scenarios that people can look at and apply to themselves.
DR. CARR: I think we should model the Princeton choose-your-college: big school, small school, in the country, in the city, tuition.
DR. HORNBOOK: That's what Checkbook tries to do, in this case it's acute disease, chronic disease, healthy family, et cetera.
DR. WARREN: I was wondering if we couldn't use an example like you see on various labels that the federal government has mandated, what information has to be on those labels so that people can compare it. There are the food labels, the drug labels that have to be present and that may be, in the interim before we get to the lovely shoes thing or Amazon, if that might be the first step. The other step will come once people realize there is common stuff. They are going to figure out how to market themselves better, and it's going to be to go the Amazon way or the shoe way or whatever. Maybe that is what we try to put together, to encourage the Secretary to put together whatever she does, to come up with a standard reporting format for what a plan covers.
DR. HORNBROOK: Can we give them examples of the distribution of healthcare expenses? I don't think people understand that having a newborn is very expensive, that being a teenager is very cheap, but after being teenager your expenses inexorably go up.
DR. TANG: We could embellish Judy's suggestion, so there is a standardized label, and then we can apply it to some prototypical use cases such as we saw in somebody's presentation. That's the second — because we're going to use that same approach with our quality measures.
DR. HORNBROOK: You also have to understand that at any given age, there is always somebody out at the three million dollar level and there is always somebody at zero. Not everybody can be at zero, only a few people can be at zero.
DR. TANG: Those are concrete. Any change in that for the immediate?
DR. MIDDLETON: Well, there is one little thing about making all of these standards interoperable.
DR. TANG: Right. How do we state that? Not only the labels, but standard definitions. How easy is it to make—you can't just say standard definition for everything, is there some commonly used set? That's where you say we ask AHIP?
DR. COHEN: We need to run these all by consumers to make sure that they understand these choices.
DR. HORNBROOK: The EPA estimates are not done by the manufacturers. The government buys some cars and actually runs them to test them.
DR. TANG: How does OPM, does OPM have a standard? They enforce this, right, in the contracting process?
MR. COHEN: Somebody needs to. Unless it's the states with all the insurance commissions at the state level. I don't know.
DR. TANG: Could staff work on this offline in terms of how it is set up in ACA and get us that kind of language? The notion we are proposing is standard labels, standard definitions, and the label of prototypical use cases for maybe--
DR. WARREN: That gets at we shouldn't prescribe how they do this. Let them do that, but there are some values that we want them to consider. One of the values is that the insurance companies be involved in helping set the definitions. Once those are done, we do some consumer usability to make sure they are understandable.
DR. HORNBROOK: Don't we have examples we could point to such as the Washington State Health Care Authority, CalPERS, and OPM? Those are examples of very large, multi-state insurance programs.
DR. TANG: They have to be around, if that is something that the staff can help us with. Let's move on and do the analogue in the provider selection space. The analogue is that we need to have standardized definition and standardized labels of what?
DR. MIDDLETON: Just getting a provider directory would be a big step.
DR. COHEN: What do consumers want to know about the doctors that help them make a selection for doctors? We heard some of that information. I don't think I understand fully what those parameters are. Some of it has to do with others' experience, including the experiential information that needs to be available, and the physician/provider-specific outcomes, how well the do, those measures of performance. I think we can figure out a minimum package of information that all consumers should know about every provider so that they can decide what is important to them. Do they speak the language that I do? Are the available on the plan I chose? Perhaps the most basic question.
DR. CARR: Male, female, age, medical school, board certification. Those aren't measures, those are just information, attributes.
DR. MIDDLETON: This is one where the providers could push back drastically. If it's not something I don't already have, they will not be bothered.
DR. HORNBROOK: I cannot use the physician pictures on our outside website to assign race to our physicians for research.
DR. CARR: I was just saying when you look up any doctor now, no matter what you want to know about them, you get whatever their profile is that has all that information, but it's not in searchable format. I think we can already see that the fact that these consumer profiles have developed, these are things that are important. We can take those, but make them choosable. I'm going to be showing you Zappos black pumps in a minute. I think it gets very patient-specific, the language, community, et cetera.
MR. QUINN: One of the things that we heard from Christine yesterday and others was the narrative descriptions, the comments, were really useful. That's what she tried to use in aggregating across multiple sites and multiple media. She was able to collect a group of comments on the doctor that she was looking at, but they were all over the place. One of the issues was how to get a representative sample of people to provide useful feedback, but the other is how you make sure that it's available so that others can make decisions on that. We know from usability testing that people value that narrative in addition to, I think it was Lynn Quincy talking about the balance between star ratings and structured information like that and being able to read.
DR. WARREN: To follow up on that, one of the concerns I have is that we have been doing a lot of looking into teaching our students about social media and how the professional deals with it. As beginning students, they don't know what being a professional is, and they think of social media the way they use it personally. Right now in the social media, there is a lot of bullying that goes on, disinformation that goes on, as well as honest, sharing, and positive things that happen. When she was talking, what I heard was that anybody can go out on the internet and start doing queries and they will get a range. How do you know what part of the range to trust so that you can select? So that you don't pay attention to the one person who really hated this doctor, for either the way they were treated or for the fact that they lived next door and their dog, whatever. How do you handle that and how do we educate people to understand what is happening in social media? Do we provide a trusted haven, and I don't know how you would provide a trusted haven on the internet because it's wild out there right now. I think those are some of the issues that are going to happen as we look things up about doctors.
DR. CARR: If we reflect what the world is like today, if you go to a restaurant thing. you will get all the reviews and you can sort by best to worst, worst to best, and you can quickly see that there are a bunch of good ones, a lot of bad ones, and the reasons they say it's bad are important to me. That is the world that we live in. It's an interesting thing because, in a way, you would engage clinicians to have patients rate them so that you would have representative samples. If I look up Paul Tang for example, here on the web, none of his patients have rated him, but I can know right away where he is, how many years of experience, where he went to school, his publications, the insurance accepted. I think we need to learn from the media. We're struggling to figure it out and they're already doing it. I think the challenge for us is how to avoid the pitfalls so that the data is right and representative.
DR. TANG: Provider selection – what is it in our domain, in measures that matter that we want to say there in the context of health insurance exchanges? Clearly, it can be tied to whether or not they participate in that plan, because they would know that information.
DR. COHEN: Does this have to be in part of the health information exchange?
DR. TANG: Well, this is recommendation four which is immediately actionable and has a lot of impact.
DR. COHEN: I guess I would think we need three black box recommendations, one about selecting the insurer, one about selecting the doctor, and one about the interaction rather than trying to rope them all into one. That might be easier to operationalize and think about.
DR. HORNBROOK: Should we also stop using the term doctor and instead use medical home? When you select a medical home, you select a team, you don't select one doctor.
DR. MIDDLETON: I am not sure it means anything to the patient yet.
DR. WARREN: I had a brilliant idea that just evaporated. One of the things that was mentioned yesterday by the first woman that spoke, and she ticked off some of the severe problems that she had, she wanted someone who knew how to handle that. I'm wondering if in this labeling for doctors we actually put down what they're good at. Right now, I go to a general intern. She is very good at managing hypertension, diabetes, cholesterol—I would say 80 percent of her patients have that problem—and on health maintenance, she's very good on health maintenance. But if I have a hearing problem or something, or if I have a musculoskeletal problem, she's going to turf me off to the person who has that.
Maybe what we should ask the doctors to do on their labeling is to say what diseases they are really good at managing. So, if I have a rare condition, I might want to start looking for a doctor who really has experience in dealing with that. It's kind of like the old thing we used to do with surgeons, how many surgeries did you do of this?
DR. CARR: That is already on websites. If you go to any of the major centers--
DR. TANG: Is this in our scope?
DR. MIDDLETON: I am worried that we potentially get a lot of unreal data. I think, listening to you, I was coming back to the eBay idea, just the doctor ratings. The sellers can rate themselves on eBay, but really what is much more informative is the buyer rating.
DR. COHEN: Do we routinely, as part of the interaction, ask patients to rate their doctors? That should be part of the--
DR. MIDDLETON: Beyond HCAHPS and all that--
DR. COHEN: Yes, a bit beyond the occasional satisfaction surveys that the plans might send out. Maybe that should be part of the ongoing interaction, having patients continually rating and developing their providers.
DR. WARREN: I don't think that happens out in the doctors' offices.
DR. CARR: Yes, it does, and again, part of the Pioneer. Ask me anything about the Pioneer.
DR. TANG: In a competitive environment, yes. In California, we have to.
DR. CARR: It is done. It is on the web. I am speaking for Massachusetts, the Mass Health Quality, MHQP, partnership, every other year surveys, but there are size limitations and so on. There are episodic things, but again, the world is going to real-time feedback.
DR. COHEN: Part of the encounter is an evaluation form when you leave the office. It gets put into the EHR, and the sum of all the encounters is a rating for that provider.
DR. TANG: Let's go back to what would we recommend the secretary, as part of the health insurance exchange, to post. What's a reasonable thing to say that we will help you set up this exchange, these are the things we need to know. Clearly, all the things related to plan, and they have the right to do. What's reasonable to ask them to build into it with respect to providers that would support patients and consumers selecting the right--
MR. QUINN: I think that the critical thing is that the health insurance exchange is usable and useful in a broad sense, meaning that it helps facilitate the goal of why it has been put there, people choosing the right insurance, people being able to use it. It absolutely, positively must be tested with actual people, the intended audience for it, before it is fielded.
DR. COHEN: The insurance exchange was intended to help individuals select doctors.
DR. TANG: This is our out of the box, bold—
PARTICIPANT: Why?
DR. TANG: Because it is related to health, and this is part of the health performance --. The doc selection influences your plan selection. The question is whether we stay in bounds as far as what are the measures that we would propose.
The attributes are that they are standard, that they are without cost, which gets to the proprietary measures, and that they pertain to individual conditions. I'm just naming some of the things we heard, and maybe those are how we start listing this.
So, standardized without cost and pertaining to specific conditions, in addition to general experience of care like CG-CAHPS. Now, it seems like we need to be a little bit more precise, so the more actionable our recommendation is, the more likely—we say these things, so an example of that would be standardized without cost, Matt, and pertaining to specific conditions. We could at least give some examples.
DR. COHEN: I guess I am still stuck with—we have people using the exchange and they have gone through the process of selecting a plan that fits them. However we do with the standardized measures, et cetera, they make their choice and they select their plan.
Now, I would say that all the plans would be required to provide them the information so that within the plan they can select the appropriate providers rather than having it be—because people who are already in plans shouldn't have to go back to the exchange to select their doctors. It is the insurers and the plans that should provide this information tool that, once they are already in the plan, here is how you can evaluate the providers.
DR. TANG: The rationale is that they have an insurance exchange, and our hook is that in helping our patients select the plan, they are selecting the plan and checking what doctors I get to use for the plan. It's together. That's the rationale.
DR. HORNBROOK: You can imagine the map with all the locations of health plan access, whether it's planned facilities like Palo Alto or physicians that practice. Then you click on that, and you find out what the facility that is, and you click on that and you can find out—if a physician makes you click on any one physician you can see the physician's profile, right? You're going to be looking at, probably, physicians in the neighborhood of your house, physicians in your neighborhood of yours and your spouse's work.
DR. COHEN: So the plan and the physician are part of the same choice.
DR. HORNBROOK: You can make it that way, but it can be integrated--
DR. COHEN: That's not the way it is traditionally done, but that would be certainly a very different kind of model. I say let's go for it.
DR. TANG: What concrete examples, perhaps, can we say that fit those three?
DR. HORNBROOK: The patient-centered presentation of information on access.
DR. TANG: On access?
DR. HORNBROOK: Where are the physicians relative to where I am in terms of getting access? Are they five minutes away or are they fifty minutes away? The geographic dimension can't be underemphasized here.
DR. TANG: Matt - that would be a different thing, because standardized is a stand-alone thing, a major point we're trying to make. Geographic dimension is one.
DR. HORNBROOK: In your example here, you say to view accepted health plan. That's a link in your profile. That is sort of, in some sense, if I really was keen on you, I could then go and find out which health plan you subscribe to and could then pick among those plans and look at their benefit structure. You want people to navigate multiple ways through depending on how they're thinking and depending on what they find in each note. Clearly, you would think that somehow there was something about either living next to you or having seen you already, or the location of your medical office, that got them near to you, or they just happen to stumble on you as they were looking up physicians in your particular area.
MR. QUINN: I think that that is the point, though. We know a little bit, but we really don't know a lot about how people go about this process. It has to be well-aligned with the mental models that people have for doing this. Lynn gave us some of the research, but I don't know if there is enough to really know. If we think we know, then we have to test whether it is right, and whether the measures are just right, a little bit right, or the process flow is just right. There is nothing more frustrating than trying to backwards work a process.
I really want to choose the doctor, I want to make sure my doctor is the one that I can choose as part of whatever plan that I get. You choose a plan, and then you have to drill down, and then you back out of that because you find out—it's thinking about those mental models. Maybe there is more, but we heard that there is a little bit of research, but not a ton.
DR. COHEN: I guess what are the ten key factors, do we know what the ten key factors are that consumers use to select their doctor? Location?
DR. WARREN: Accessibility, do they have evening or weekend hours?
DR. COHEN: With respect to measures that matter, what are the three that consumers use and what are the seven that we want them to be using. I don't know.
DR. CARR: You would learn if you put a whole host of choices out there, and you could learn over time what are the ones that people go to. I'm sure that is what is happening out there.
DR. HORNBROOK: But being a geographically defined health system, we can tell from our patient origin studies that the density is highest closest to our facilities, and the further away you go it gets less dense. We know, even inside the Portland metropolitan area, which is pretty small, that distance matters, especially at rush hour.
DR. COHEN: Clearly, their location is one of these keys. When they get into the health exchange, they might say, go right to the doctor and look up the doctor and then see if the doctor participates in the plan. That might be the way a lot of people do it. My next door neighbor says that Dr. Tang is really good, I want to see if he's—we need to give them different avenues to navigate. Then, she should know the location, the office hours, what language he speaks, what he's good at, what other people think about him. That's what we heard.
DR. TANG: Do we really believe that the job is done once we get a plan with essentially the deductibles, the out of pocket costs, and the location and hours? That might be true.
DR. CARR: Let me just explain one thing. In Massachusetts, within one mile of my office, there are five academic medical centers. Location isn't an issue in Boston. I'm using the extremes. I think it's one of the things, but I think if we start in with pre-conceived ideas, we're going to undermine the value of the data. If we put an array of things that even health grades and others have put on there, and we see what people chose or what they look for, we'll learn. I think what we're saying now is what we think patients want, they don't want. We don't have a mechanism for systematically learning what they do want or what they don't.
DR. TANG: What you are suggesting is research be done instead of making a firm recommendation?
DR. COHEN: I don't know whether some of the presenters yesterday know the answer to this question or not. They might have, certainly, more insight than we do right now, and the answer might be that we need more research, patient-driven research, but the studies have indicated that these are the parameters that matter to patients selecting their doctor.
DR. FITZMAURICE: I was just going to say the same point that Bruce brought up, do we want to recommend further research into this question of what consumers prefer? Do we want to synthesize the existing literature that exists today? Those two things.
DR. TANG: We may say some of the attributes that we did hear about, such as standardized, no-cost, and pertain to specific, because they did say that, and we list some examples of things that we think need to be—we ask for more research and list examples of things that would be included.
DR. WARREN: One of the things that the panel did say yesterday was that most of them wanted to know whether or not their doctor would communicate with them by email. I think that's becoming a huge thing. Will they answer my questions by email? Several people said that yesterday. Not so much office hours, but if I have a problem and I'm at work, or it's later in the evening and I suddenly have a question about how to do whatever, can I email them and get a response back?
MR. QUINN: Just think about some of the mental models that we heard yesterday. Christine said that she wanted somebody close, somebody who will answer my questions honestly, somebody who has some experience with my specific health conditions, and she wanted someone that she trusted. The mental model that my wife goes through is a little bit different because she doesn't have specific health issues. She says that she wants a female, number one, number two she wants someone about her age, number three she wants somebody who is convenient, even though it doesn't have to be right down the street, it has to be in the general area, and four, she wants somebody who is, of course, covered in her plan. That's a different mental model.
Somebody might say that they want someone with very specific experience in this health condition or somebody tied to this larger medical group. I think there are different mental models of going in this and I don't think they have been well-defined. Or maybe they have and we just don't know about it.
DR. TANG: I would like to propose we take a break. Let's talk about the rest of the schedule. We are breaking at 1 PM, so I think we take a short break and we'll probably—how are we dealing with lunch anyway? We don't have any lunch arrangements? I guess that means we don't have lunch before 1:00, but I think we need to get further than we are right now, and maybe after the break, let's try to close off on this, a final push for this, number four, and then get onto others. So, just ten minutes, 11:50.
(Break)
DR. TANG: At least, when I was doing computer programming, the times when I figured out my bugs was as I was walking out. Your brain just lets it—Justine has come up with the solution to number four. She's going to present what they do, but I think there is a way to generalize what they do.
DR. CARR: The concept is let's take what we have and use it, and learn from it. I'm disclosing this. This happens to be my health care system, Steward Health Care. We had a very academically focused website that the board members said they couldn't follow, we don't understand it. We met with our marketing people who asked us to tell them about quality and then leave them alone. This is what they came back with. You go to this, this is the landing page, and it is a little bit about what we believe in quality in health care in Fall River —well, I've already gotten to that. I chose St. Anne's hospital—this gives you a drop down for the hospital, or you can go on the landing page to a zip code, and it will bring you to our closest hospital. When you get to this landing page, we have three dimensions, and, again, trying to keep it simple, patient experience, quality, and safety. Each of these are drawn from publicly reported data.
On patient experience, this is the HCAHPS data, and you can see that nurses communicated well. Here is the national average, how many points above the national average are we? You can also say, what do the nearby hospitals look like? So, we're above average, but we're better than all of the hospitals in that area. This happens to be a very poor, low socioeconomic status, Portuguese-speaking area. It is very challenging to understand and meet the needs.
The same thing, explain about medicines, we're above average, and in the community around us, below average. This is simply publicly reported data. We don't do this. The one challenge is that it is out of date, and we are just about to update it now with the latest information.
We also show where we fell below the average, so if you look at patient's room kept quiet at night, we were nine points below, but if you look in the community, everybody is struggling with the same noise issues. Those are the HCAHPS data, and we do the same in quality. Here we look at a roll-up of the four measures, and it's overall recommended care, heart attack care, pneumonia care, surgical care, and we fell below in heart failure care, which we no longer do, I'm happy to say, but we then show you, actually, Brockton Hospital did well, the other hospitals did well, but many hospitals also struggle with that.
Again, it is what it is. For safety, we use Leapfrog, and this is an annual survey that we participate in. St. Anne's hospital meets full standards in all but one area, and this one was when we had just rolled out POE. You see that we didn't get the full four bars because we hadn't fully rolled it out, but you look at the neighboring hospitals and you can see that others are also behind in doing that.
As Paul and I were talking-- I think that this is a nice example for us of a way to — there's not a lot of upkeep except that as these data get available, we put them out there. What is frustrating is that as we are improving, these hospital comparers, six months behind, nine months behind, but I think we have made progress. You just put in a zip code here and you come up with the closest hospital we have in our system.
We were talking about what is publicly available, and if you begin to think about physicians, it is tricky, but we thought about the fact that meaningful use is something—are you a meaningful use, certified provider? Or from your board of registration in medicine page, are there any sanctions? Where did you go to medical school? Male or female? I think the concept is, and it is the same as what Todd Park is saying, we have all of this stuff available. The challenge is to think about which threads really begin to help people to make decisions.
DR. TANG: To generalize this, what we have heard is that you can't find it all in one place, even though there are some important, though not ideal, measures and information out there. Maybe the next step, which is the immediate step, this whole health insurance exchange, is to have CMS post the already available data. Meaningful use is not a bad surrogate—we used to call it got EHR, as one of your first patient-safety questions. Take advantage of CMS initiatives already, meaningful use, PQRS, HCAHPS, CG-CAHPS, and have it in one place, so that when the consumer is making the decision about selecting a health plan, they have this information that is part of making that decision.
What we're doing is just starting that path. It still has attributes of standardized, right? Already all of the publicly reported stuff has to be standardized. It becomes a repository and what can improve are the tools to use that information, and then more and more people will want to have information made available. It just starts the private sector in the right path by a public sector program.
DR. CARR: There is physician compare on --
DR. TANG: What does it use for the metrics?
DR. CARR: I think all it says now do you take Medicare, and it also says if you participate in PQRS.
DR. TANG: Okay. It has already started that. So in other words, we can take hospital compare and physician compare, plus others, and just start there. I would say that is a bold, immediate response that has a lot of downstream potential benefits. I think Todd would love this, too, so there's even a champion.
DR. CARR: I think we can sort of set the stage, but I think we should talk to him on Friday, about taking this and getting the people who have the time to go out and really look at what else is available and how it might be used, and then maybe get a reactor panel. That's kind of how we did our thing. Here are all of the things, is this meaningful, what about if we put it this way, and it is in keeping with that spirit of publicly available data.
DR. TANG: What would you think if we reword that and that becomes our quote, recommendation four, which may be recommendation one, but that recommendation does sound good?
DR. COHEN: Also make sure a couple of things, the usability testing for the consumer, and also leaving space to add other domains that the consumer wants to this model.
DR. CARR: I think the discipline of ours was saying, what of it matters to the patient, the patient experience, is it safe, is it high quality, just those three things. Then, we will be building on that with the other measures, PSIs and stuff like that are now on hospital compare.
DR. COHEN: Is that bold enough for you, Blackford?
DR. MIDDLETON: It's a good start. I do think, though, that what we heard from Lowry(?) and Lansky, goes way beyond what St. Anne's has done in terms of what are the measures that matter to the consumer because none of those were up there.
DR. CARR: I think that is right, but we have to start with something. I think, as you do that, now you are better than you were. Now you build on it and you think of those streams and how to get that data to be --
DR. COHEN: Maybe the letter should include short term steps and long term steps. Immediate things where we have the technology and capability and then another section on future directions where you need to--
DR. MIDDLETON: The aspirational and really set the vision.
DR. TANG: You could even have a section on standardized, in a sense, vetted, because all of these federal measures are subject to audit, and then there is another voluntarily submitted section that you could put your IATP for peace(?), whatever it is, and we could still figure out some kind of oversight in terms of audit, but then that would create yet more incentives to start competing on what--
DR. CARR: It does align incentives because the MHQP, for us, the Massachusetts Patient Satisfaction survey, it puts up whether you participate and what your scores are, if you have enough of a panel. I think if you start thinking that way, there are things that you can begin to build. It's not perfect, but when you look at all of these websites and you type in your name, there is a health grade's evaluation of you, four different vendors that do it. They pole where you went to school and whatever, and then there is a part for what patient's think of you, if they want to put that in.
DR. TANG: Then maybe this is our opportunity in the voluntary section to almost have attributes of measures that matter and right now there will be a bunch of blanks, but it would stimulate the measure-developers to go in that direction, to fill out the things that matter to individuals.
MS. JACKSON: Following up on the last line from David Lansky, his whole message was rather than the measures, it's the systems. That kind of falls into looking at the whole pattern rather than the individuals. Then you can start answering the questions, because I can imagine people coming to this hearing and then looking at the follow-up from that, and really wanting to see measures that matter to patients. You have to get there some way, somehow.
MR. QUINN: I agree with the approach that we've just described here and taking a good inventory of what's available--data is different than measures. I think that there is opportunity to leverage what we have. I don't want to lose the idea that we have better measures for consumers or to lose the idea that there needs to be that connector between provider measures and consumer measures, and consumer measures and providers. That's a gap that needs coordination just like the provider measure world needs coordination.
MS. GREENBERG: I don't remember who said it, one statement that really stuck in my mind, and it might have been David Lansky, I'm not sure, but we don't just have to have better measures, we need to have better measurement. They're not the same thing. They should be related, hopefully.
DR. TANG: Matt, could you just separate that last “f” as a separate section? There is a difference between what CMS, the required submissions, because those are essentially auditable, versus some of the others, but we're trying to stimulate the creation of new ones.
DR. CARR: I would like to also go back—I agree with whoever was saying that we talk about what you can do and where we are aspiring to go to. Getting back to the idea of use cases, you were saying that, and I think we went off into the insurance thing from this, but if we say, I have prostate cancer, I don't have a lot of money or insurance coverage or a very high deductible, what are their choices? Using what we have, pulling from AHRQ the clinical effectiveness research, but shrink it down to if you are looking for something that doesn't cost a lot of money, that's comparable in its efficacy and it's not a surgery, it's a medicine, it's not a medicine, it's a meditation, whatever it is.
Thinking about doing a couple of demonstrations on two or three conditions, that pull all of this sophisticated data but shrink it down into that decision tree of what is it that you want to know. When I go on the clinical effectiveness it is overwhelming, and I don't go with a question. But if I'm saying I just got this diagnosis and I signed up for this insurance with this enormous deductible, and I don't drive and I need childcare, what can I do? Now, that suddenly becomes relevant to my situation.
DR. FITZMAURICE: CMS runs the national provider directory and it might be a good thing for Medicare beneficiaries to ask physicians, maybe an annual update, with what insurance companies have you signed contracts. Consumers then would have a place to go to find what insurance companies have contracts with that physician.
DR. CARR: Or you could ask the insurance companies.
DR. FITZMAURICE: Yes. Once we get a health span ID in operation.
DR. COHEN: Part of the discussion yesterday was about the heterogeneity among consumers. People learn differently. Some learn by seeing videos, some learn by graphs, and some learn by narratives. Plus the fact that they are all levels of sophistication. I really like the notion of, maybe not in the letter, but as part of our ultimate report, providing these narratives and these use case scenarios, I think they will be very instructive about an approach to providing information to consumers in a digestible way. Here's an alternative, and you can have the graphs, and we can make these stories video stories, interactive stories, and try to build a library of use cases that might be helpful for individuals.
DR. CARR: I also think that because our theme is to liberate data, to use the data we have, the National Library of Medicine has a whole inventory of information on diabetes, hypertension, common conditions. It can be interactive, video, narrative. It's all there. All you have to do is link to that, and that has already been provided. There is no work to doing that and it is an example of using what we have.
DR. TANG: Let's make progress on – this sounds like creating, developing, researching, new measures that matter more to consumers. How do you want to phrase this?
MS. GREENBERG: Do we want to say anything about categories?
MR. QUINN: Focusing on prioritizing a handful of conditions for consumer decision support measures--
DR. CARR: Yes. Pilots – pilot some models of using available information in a consumer-friendly fashion and learn from it.
MS. GREENBERG: But like we saw, functioning is important. I'm sorry I'm coming back to this. There isn't much information available. If you're going to have to use existing information in the acute care setting, almost nothing is collected about people's functional status, except by therapists, occupational and physical therapists. There is where you may need to do some development of--
DR. COHEN: In addition to what's there, what's missing that would help people make better decisions? That would be part of the narrative.
DR. CARR: I think the functional status is related to the outcomes. I made this choice and I was better or I wasn't. I think it's a continuum.
MS. GREENBERG: The woman started off with what she really wants to do is be able to golf, but she has back problems. That's what she wants.
DR. CARR: I think we need to stay in the step that we are in. The first one is I need a care provider, and how does one get one that fits his or her needs. The second one is being diagnosed with a condition, where does one go and what information is available to me? The third one is for people who did this, what happened to them? I guess that's a piece that we don't have. Could they golf after they had this back surgery?
Yes, that's a missing piece, but if we define the continuum and highlight the gaps, but also use what we do have, we will get further. If we don't start using what we have, we're making assumptions. We can cover the continuum of what was heard yesterday, but saying where we are on that continuum, and now it's the consumer finding their insurance, now finding their care, and what else do they need for that, now it's a consumer in relation to the doctor .
MR. QUINN: One of the things we talked about yesterday was whether the validation or endorsement process for, for example, consumer measures, how it should be the same and how it should differ from the current NQF-type process. I don't think that we came to any conclusions, but that should also be part of what we discuss or recommend.
DR. TANG: I don't think the process needs to be different because, in fact, over 50 percent have to be consumers. They are only a receiver. They don't write measures. Maybe, is it a true statement, that we would recommend that CMS should contract for meaningful measures to use in innovation programs and potentially meaningful use as a way to get them into EHRs? In other words, is it true that we actually have to go get these measures developed, tested?
DR. COHEN: The two missing pieces that seem to be the biggest gaps for me are, yesterday it was consistently reported that people are influenced by what others say about their providers. Measuring other's experience and incorporating that to provide information to consumers is a real missing piece. The other missing piece is the functional status. There is no information about how well your overall health, rather than curing your particular acute condition.
DR. TANG: Is that true that we don't have the measure or we don't have the means to capture that data?
DR. COHEN: I think the latter. The first one was that we don't have that information and the second one was that we hadn't figured out how to acquire it and provide that information back.
DR. TANG: One, we could take advantage of measures that already exist, like functional status, and that is an area that meaningful use is going, to be able to grab that stuff electronically. We can make that recommendation.
The other is that there is new development to be done in order to measure other things that are important to consumers. That would be important for CMS because that would be consistent with what they want to get out of accountable care, all of the new, innovative programs.
DR. COHEN: These instruments exist, but they are usually done as surveys or in other contexts. They have never been integrated in an ongoing part of the care-delivery and directed and for the purpose of providing feedback to consumers. They have always been for the purpose of providing feedback to the providers.
DR. TANG: In this recommendation, we can discuss the whole proprietary interests, the cost issue, for using the tool, so that we one, explore how to use already publicly funded tools, and two, how do we create more of them?
DR. MIDDLETON: Where does it go?
DR. TANG: What do you mean where does it go? It's going to end up in that exchange repository that we have created for existing measures. We need to create new ones.
DR. MIDDLETON: Who runs that?
DR. TANG: Either the state or the federal government.
DR. MIDDLETON: I think we should be concrete about that.
DR. TANG: I think that builds off of one. We said here, make what is already available useful and accessible to consumers when they are making these decisions, the first is health insurance.
Second, we recognize there is a gap. There are some measures that have no systematic way of being capture efficiently and made available, like VR-12, et cetera. There are other dimensions of care that are not captured, and we need to develop, we need to do more R&D on those.
DR. MIDDLETON: AHRQ funded stuff?
DR. TANG: Correct, in such a way that makes it publicly available at the end, and deal with some of the stuff that is already publicly funded that is not available.
MS. GREENBERG: Do you want to make an explicit statement about involving and engaging patients, consumers, in this work?
DR. TANG: As part of the R&D, the development must hear from the voice of the consumer upfront, and have usability testing in the back end. Actually, it is usability in terms of the way that the information is captured, and the way it is presented to the clinician. It has to be usable on both sides, otherwise it is never going to be used.
MS. GREENBERG: Is there anything you want to say about meaningful use?
DR. TANG: These measures could go into meaningful use. We are actually working on this already, but it would be fine for a recommendation.
MR. QUINN: The paradigm change here is that this is going to be consumer-sourced data.
MS. GREENBERG: A lot of it is. How about a recommendation, also, that links into the population health community data? This is important at the individual level but also --
DR. TANG: Yes, actually, I am trying to think whether that could go into the insurance exchange. It could in the sense that if you understand the population, you could design better insurance products for that community. It is done on communities, certainly, at least, by regions.
DR. COHEN: How about the Rios (26:31?)? Would that be a site for population health information, aggregating?
DR. TANG: It is one of the functions that HIE organizations are trying to do. HIE organizations is still a struggling concept actually. I think that's good. That is filling in gaps now.
MS. GREENBERG: Did you have something on the standards? Yes, standardized there, okay.
DR. TANG: Yes. So those are some of the attributes for these new measures to be developed.
MR. QUINN: That National Academies report on consumer health tools and improving their usefulness is good stuff. I can send it around. It talks specifically about standards for things like privacy and accessibility and things like that.
DR. MIDDLETON: Do we take out the cost information? Is that what “without cost” means?
DR. TANG: Without cost means without license fees.
DR. MIDDLETON: Oh, I see, to use. Do we need a recommendation or is it contained elsewhere about transparency on cost for the consumer?
DR. TANG: I think that was up in the coverage table, because that is presumably what they're making the decision on primarily there.
What major themes have we missed? We have dealt with the problem of-- we're trying to deal with the complexity problem. We're trying to deal with the coverage concern trumping quality. We're trying to deal with the fact that it takes so much effort to go find this information, by putting it together. We're dealing with the standardized. We're dealing with the proprietary interests.
MR. QUINN: Is there anything around community, state, federal use cases? It was really nice to hear federal, big picture folks like as AARP and then to hear a community relate it to their specific needs. They boiled everything down to really one -- to high blood pressure and controlling that. That was their one metric that they were focusing on. That resonated. Their dissemination strategies and everything, was understanding their local situation and who they needed to reach.
We want this not just to be for big things. We want the pieces that we are talking about here to be reusable and usable at those levels. I'm just trying to think if there are any recommendations or thoughts on that.
DR. COHEN: ACA has the requirement that hospitals do community needs assessments, non-profits. Maybe this would be a good place to make that connection, that part of the community needs assessments need to incorporate-- I'm not sure exactly what they need to incorporate, but they need to build in the consumer orientation and summary kinds of outputs from this process.
DR. HORNBROOK: Maybe it is a new source of data.
DR. COHEN: And hospitals are looking for tools to do their community needs assessments.
DR. TANG: Actually, we have covered a lot of what we heard about.
DR. COHEN: Have we gotten the notion that changing the culture of the interaction between the provider and the consumer, that this is just as important as providing what your blood pressure is? The vital signs concept?
DR. TANG: We are indirectly motivating that by having these publicly-centered, consolidated places where you look. It just so happens that the consolidated place where you look has measures about those aspects of care. That's the indirect way.
DR. MIDDLETON: What about the consumer rating itself? In other words, it is leveraging Yelp and Angie's List and all the other kinds of activities which show consumer interest and energy in rating. Is this part of the health information or the insurance exchange? Is that going to continue to be done at the community or commercial level?
DR. TANG: I think one of the things we heard, and there was an example about this Kaiser doc, how it's not-- we don't know exactly-- people use it, but we don't know if the implication is good or bad. Actually, a couple of people talked about the implication of seeing one bad review, and all of the sudden everything else is covered. You could have ten good ones-- so, do we understand the cognitive process and is it a valued one? I don't know, but a couple of people brought that up.
MS. GREENBERG: Sometimes the committee – this is kind of time sensitive, we are trying to provide these recommendations as the department is standing up new systems and all of that. It seems like the areas of responsibility are different parts of the department-- we've usually said “that's for the department to decide”, whether it should be AHRQ, or CMS, whatever. Sometimes the committee asks for a report back from the department in six months or something, as to how and whether these are being addressed. As opposed to what I would generally refer to as the bread and butter notes that your mother made you send, thank you for the nice gift or dinner - I'll use it well.
I'm trying to think about how this doesn't just go into the ether. Of course, you all can also offer to work with the department. That sometimes is taken up and sometimes isn't. You might want -- I'm already into dissemination, I guess, and we haven't even done the letter, but it might be something you would want to engage with the Data Council. I'm thinking that this is really important. The committee was commended by at least one speaker for shining a light on this area, and I don't think you want to just send your letter and say you've gone home. I don't know. You may have done your job, but, well, only part of it.
MR. QUINN: I was going to say the proof is going to be in the pudding. I had an offline discussion that the absolute nightmare scenario is that people try to use the new health insurance exchanges and they don't meet their needs. That's bad.
MS. GREENBERG: Well, current systems don't, so that would just be consistent.
MR. QUINN: Right, but this has got to work. The more that they are designed in ways that support the way that people select, want to-- that enable them to accomplish their goals, the more successful everything is going to appear. That's a good thing. For us to provide positive input in some recommendations here, is a good first step, but it would be great to get information back. I don't know who the right people are to provide that information back.
DR. TANG: Blackford, push us on what area have we left?
DR. MIDDLETON: I think we have covered most of the basics. Coverage trumps quality, provider trumps quality, basic information needs for the consumer. Consumer friendly language. Keep it simple, stupid. We didn't talk about community, actually though. You did a little bit, maybe, connecting consumer level measures back to the community level stuff. Phenotype, genotype, behavioral profile, functional profile. We are a little soft on increasing functional measurement still, which was the number one for consumers or number two.
MS. GREENBERG: I'm assuming that this is an outline, a fleshed-out outline, I commend you. So often the committee hears this great testimony like we heard yesterday and then everybody goes off to the four winds, so I commend you for holding this session and engaging, and of course Matt and Susan, too, to get you this far half a day after it ended. I'm also assuming that they can go to the presentations and to pull out some of the detail. If there are areas that you want them, such as the area in functions, then you should tell Susan.
DR. MIDDLETON: What I would suggest now is that maybe the last gasp is to go back to the top and review. Let's start at the ones that we think are in the top five.
DR. COHEN: Before we do that, with respect to the letter, can we address Marjorie's report back? Does the committee, has the committee asked for the Secretary-- or in six months can they give us an update of how you have incorporated these suggestions into the--
MS. GREENBERG: Yes, you have asked in the past, with mixed results, I guess.
DR. COHEN: And if you would like us to continue working with your designees to do this, we would be happy to.
DR. HORNBROOK: Not just the department, but maybe the health insurance exchanges to whom we are making several suggestions.
DR. COHEN: Who is controlling the development of the exchanges?
DR. TANG: Who is?
MS. GREENBERG: I think it must be in CMS now because OSIO(?) wasn't able to be stood up as a separate agency because we were under a continuing resolution last year. It's somewhere within CMS, and I think, still, the office of-- it's somewhere there. We can find out from Lorraine Doo.
DR. COHEN: So for each of the recommendations we should identify a target organization and our willingness to cooperate with them and a report back from the department about what progress each of these organizations are making in incorporating these suggestions.
MS. GREENBERG: Some of the research things, it is probably AHRQ.
DR. COHEN: I think there would be logical target organizations within the department to address these recommendations for and work with.
DR. HORNBROOK: In some states, I know Oregon has already passed comprehensive health reform legislation, starting in 2013, to set up the insurance exchange and create a pool that takes all the uninsured and distributes them among existing insurers. The law requires every insurer that has a license to operate in the state to participate in the pool. There are going to be insurers picking up people, who they have never seen before, and provide coverage down to at least the ability to find people that are physically better --
DR. COHEN: Massachusetts certainly has an active exchange. I don't know what Vermont is doing for their universal care. They must have developed or be developing a way to hook people into care.
DR. TANG: They have a single payer system.
MR. QUINN: Their IT infrastructure was being built, and I know that they changed technology providers with that. I don't know where they are recently.
DR. COHEN: There is stuff out there.
MS. JACKSON: I think we can be more aggressive in engaging our liaison, in this case, Lorraine, for CMS to wrestle through some of this. As Marjorie and you were saying, this is a fleshed-out outline, that Susan with all of her notes, might have some themes. I was going to follow up on what Bruce was saying about communication, the changing in the culture of communication. That goes back to the first thing you were talking about in terms of communicating with the family. If your information doesn't go in certain way in the way you need it to, various aspects-- it's the whole communication box that I'm not sure has been fleshed-out. So we can keep in mind that things will be tucked in as you get to them.
MS. GREENBERG: I think some of the previous work of the committee, whether it be on the community health report, the medical home-- your letters, obviously, that we can tie together.
DR. TANG: If you go back to the community health panels, which was true of Rochester, one of their number one requests was to give them measures that they can use that are standardized, to be combined. The second follow up was to help them with some technical expertise or tools to deal with that data. Our emphasis on standards is addressing those needs.
Susan or Matt, other missing holes we have in what you're hearing? Then we need to do what Blackford was just suggesting.
MR. QUINN: How strongly do we want to-- one of our recommendations from the meaningful measures was this national quality measure coordinator. Do we want to harken back to that and say that it is even more important with this and to call out the connection between consumer and professional measures? If we have these as separate things then that connection won't happen. We're in a slightly different place with consumer measures. We have some and we have some gaps, just as we have in the provider world, some and some gaps.
DR. TANG: I think with consumers we haven't had enough motivation for folks to even create their silos. They just aren't created. We were hoping to motivate that by creating a place where it is accessible to consumers and then people recognize that it is pretty valuable here to be on there.
MS. KANAAN: I am interested in, at some point, fleshing and going further with Bruce's suggestion about the ACA recommendations about community health assessment and then somehow structuring that into your recommendations. I didn't know if you wanted to do a little more work on that now, or put that as something to flesh out as a sidebar conversation. Does it fall out of the scope of this report?
DR. COHEN: I think the Department would appreciate any guidance on implementing the community health assessment aspect of ACA. We can ask Jim tomorrow or whenever, but I haven't heard anything about any progress on that. I know in Massachusetts, the state health department is getting a lot of inquiries from the IRS and mandating the ACA mandate, what are we going to do here? I think it would be a valuable connection. I hadn't really thought about how to make it quite yet, but I think it would be great if we included that. It would be welcome.
MR. QUINN: So starting from the top in numbers and letters. Milestones so far and milestones ahead from previous hearings and recommendations, move providers along and move consumers along. What can we do today? We have re-education of the healthcare community. A place where quality data was aggregated for consumers' health information exchanges. That's pretty central. Integrate measures, meaningful information for patients, people, clinicians, healthcare professionals, purchasers and plans. So that basically sets out the scope of it.
Focus on what we have in use cases as framework for decision-making, that was going back to what David Lansky talked about. The meat of it is this number four that has a whole bunch of stuff under it. Federally supported health information exchange should be required to set up standard benefit tiers, coverage floor, and require next generation measures that matter to a wide array of consumers. Tools from standardized measures help navigate the system and then look at the medical home corollaries. Case managing a transition for all personal case managers for high use, that came from what we were talking about at Kaiser.
Select a health plan, or probably coverage, then selection of provider and interaction with provider. Leverage timing of health information exchange and federal involvement. Mandate what information is on labels for coverage. Labels must be present and standard in an operable standard definitions in conjunction with, probably coverage.
PARTICIPANT: Do we want to say coverage description?
MR. QUINN: Yes. Prototypical use cases.
DR. COHEN: Those could be defined by condition or demographics or both.
MR. QUINN: Also, by the family involvement. Bruce's point about how many people-- fortunately, I wasn't involved in this-- but when my parents had to select a part D plan, how many people walked their parents through this, trying to figure that out?
Thinking about who is involved and how people make those decisions with their families and care-givers, usability testing with consumers. Then the next step is provider selection, standardized provider directory, outcomes, plans, performance, languages, minimum package of information that every consumer should have access to.
Avoid the pitfalls we learned about. What are the providers' medical homes good at ratings? The standardized without cost, which means license fees, to use pertain to specific conditions, the geographic dimension. Overall, what are the key factors or mental models that consumers use to select their physicians? More research is needed. Synthesize existing literature, list examples, such as communicate by email, that was a parameter that we really hadn't thought about until we heard Christine's presentation.
What are standard inputs to enable the market? Benefit design, exchange into repository, and then we talked about getting input from AHIP and OPM on what are the standardized widgets or plan components?
The next is to have CMS share or post data that is available, required data, and make it usable and accessible as part of this. Some of the examples of that could be meaningful use, PQRS, the flavors of CAPS, physician compare, national physician directory, and update with plans with whom they have contracted, NLM condition library, hospital community needs assessment. Then others along the whole data liberation concept.
DR. MIDDLETON: Here, it is useful to clarify where we think there is new data or data which is not yet captured, and distinguish it from that which is already captured and we are just reusing, data.gov or whatever.
DR. HORNBROOK: Is there a way to get information on any physician who has a Board of Medical Examiners action against him?
PARTICIPANT: Through state board.
DR. HORNBROOK: Can you put a link to the state board inside these physician data services?
DR. MIDDLETON: Justine said earlier, maybe this could be not a repository but gigantic mash-up. That's a little on the implementation detail side.
DR. WARREN: It is who is going to want to put that link on their webpage.
DR. HORNBROOK: Well, you have to be protect yourself from physicians who will use your name to continue their career after they leave that state or that organization. That's what we had in Kaiser.
DR. MIDDLETON: What we haven't wrestled down is who is doing this. Who is self-interested enough in doing this on behalf of the consumer?
DR. TANG: Right now, only the government, only in the public.
DR. MIDDLETON: Who in the government?
DR. TANG: CMS.
DR. CARR: This is all part of what they are trying to do. They are looking for meaningful applications of the data that they have.
DR. MIDDLETON: But when we have a thousand data elements and they are used in 1,000 health plans or care systems, who is maintaining them, doing it, running it?
DR. CARR: The idea is that you draw from reliable sources that are updated and then you mash-up as--
DR. MIDDLETON: Who makes the mash-up?
MR. QUINN: I think what he is referring to is the operational component of this. We heard yesterday from Christine that many plans, the plans she looked at, didn't have updated information.
DR. CARR: One of the groups that we didn't hear from, and I'm sorry I didn't think of it, is the Commonwealth Fund, that has one of the best. That would be a very logical participant in putting this together. They have done a great job with taking-- that is where we, Steward, takes our data from WhyNotTheBest. They take it from Hospital Compare. I think at least as a pilot, getting the proof of concept out there, I think they would be very interested. We should definitely have them on our list, Anne Marie O'Day , or somebody like that.
MS. GREENBERG: List for?
DR. CARR: People we invite to hearings. The work that they have done, I think ahead of anyone else, has taken --
MS. GREENBERG: That was one question I had, whether you had anybody else you felt you needed to hear from or anything else you needed, information you needed to collect before-- or are we shooting for a June letter?
DR. MIDDLETON: Yes.
MS. GREENBERG: Some information could be gathered, also, by some other means than a hearing.
DR. MIDDLETON: We might want to draw out the connections to the National Measures Clearinghouse and what its role might be in maintaining some of these things. The other thought is, do we want to go back to the recommendation from the prior letter a year ago about there being a national Steward of quality measurement. What was that called? This could be something that that thing could do.
DR. CARR: Or, again, it falls under Todd Park. The fact that we have that, so quality becomes one of the-- what they have done, I was down in November or December, somewhere around there, and they had each agency of HHS present the data that they had made publicly available, and they had three people as a reactor panel. The thing that was so fascinating was the Institute on Aging had the same information as children.
In their mind, it made perfect sense that one talks about when you're old and one talks about when you're young. One of the things we said was what about just a person? Rather than being agency specific, it be person specific and fed by information from the Institute of Aging, but it would be agnostic to where the data comes from, but as you look at it thematically.
That was one of the things that they took back that they wanted to do. I think this fits in very well as well. There is such a huge amount of data and they don't know how to organize it, or they are asking for ways to organize it that meets the needs of the community. I think, again, I really think Todd Park and the Commonwealth Fund would be two key players who would have an interest and the resources to play a role in this.
DR. COHEN: So you think we should, before we generate this letter, talk with them about it? The other piece of information we had talked about getting to follow up on Matt's question was to assess the level of research about what consumers want, what information they want to know about in order to make good decisions choosing their providers, if that literature exists.
DR. MIDDLETON: Back to the federal participants, it strikes me that the ONC will obviously have an interest in this on a variety of levels, and whether or not there is a function in the ONC which becomes the Steward of consumer health quality data, is the thought.
MR. QUINN: Josh – his name escapes me-- Sideman from ONC was here yesterday. They have a new initiative around consumer health IT stuff.
DR. CARR: The only thing I would say is that, as we think about data sources, even survey data and things like that, it goes beyond EHR. You just want to make sure that the owner is thinking about all the data sources.
MR. QUINN: Another concept that we haven't touched on much is that health insurance literacy is different than health literacy, is different than personal literacy. I know that AHRQ has done a ton on health literacy, but I don't know how much on health insurance literacy and what we know from that in our experience in rolling out part D. That would be the other thing, to call out lessons learned from part D or from Massachusetts Information Exchange, or other places where they are further along.
DR. TANG: Are we –
DR. MIDDLETON: That was a pass through, we only starred one. It's either one recommendation or eleven at this point. We can do some collapsing. I'll help Paul lump. One and four are related.
MR. QUINN: Susan, maybe you and I in the next week or so can try to--
DR. TANG: One seems – I think 4 and 8 that are the two - the meat of it.
MS. KANAAN: I wonder if it is too late for me to raise a question about what might be missing here. I don't know, again, if it falls outside your scope because I know you are looking for where you can have the greatest impact, but as a resident of a rural area, I have to say that the provider richness and the information richness that you're assuming, is just simply not the case in many areas of the country. I wonder if there might at least be some kind of acknowledgement of the need for, or the possibility that government will help supplement information or give extra support to rural areas. I don't know what the recommendation would be, but so much of what you're talking about today has virtually no relevance where I live. That's not just my town, but my county, or much of northern California, as I think Paul can attest, and much of other areas of the country as well. Very few choices of providers, very little information, et cetera.
DR. TANG: How is that related?
MS. KANAAN: I am not sure. When you're talking about missing information, you make some assumptions about what information is already available. Maybe it is simply a matter of recommending, assessing the information that is available not just in a typical urban area with academic medical centers, but in other parts of the country.
DR. COHEN: The generic question, or the larger question is, is there a different set of measures that matter to people in areas where there is not a richness of providers in coverage options? If that answer is yes, then part of our job, maybe not in this document, is exploring the possibility of developing measures that matter for areas that are less provider heavy.
DR. HORNBROOK: I know that in the Oregon coast where we know a little bit more, it is sort of community knowledge that there is a certain level of acuity where you better get to Portland as fast as you can, and it's actually pretty low.
MR. QUINN: One of my formative moments, my first experience in healthcare, I called on medical oncology practices, and there were two in one little town in Oregon that practiced completely differently. I'm sure that they at one time were partners and decided-- but in this town of 40,000 on the way to Medford, where there is a big oncology practice-- I wandered into this world and asked how is this.
DR. HORNBROOK: I think the need is equal or more so.
MS. KANAAN: We're talking about a lot of people, or at least large areas of the country. Maybe it is simply in the future directions commentary or things like that, where we need to identify something like that.
DR. MIDDLETON: This is great, Matt. Thank you for taking the notes. I guess what we could do is circulate them with me and Paul, Susan, and perhaps try to refine a sort of first draft, fleshed-out outline, within the next week or two. At the end of that period, we'll find a time to get a call together to go over the fleshed-out outline that we've circulated on email before we commence even any writing, unless you want to do background stuff, just so we are all headed in the right direction.
The question for the committee at large might be at what point do you want to intercept, with a complete full draft, a first draft, or even at the fleshed-out outline stage? Say the former, and I'll be fine. First draft.
MR. QUINN: Do we also want to review it with the agencies involved, so CMS and AHRQ and other folks?
DR. MIDDLETON: In draft mode or when finished? Usually wouldn't it go from subcommittee draft, to Executive Committee to full board of NCVHS.
DR. WARREN: He was mentioning we have liaisons too.
DR. COHEN: If we are going to make recommendations that effect different agencies in the letter, then we should make sure that they see it sooner rather than later.
DR. MIDDLETON: That would be in our committee process.
MS. GREENBERG: I am assuming it would be agencies that already have a liaison, but if not we should facilitate that.
DR. WARREN: What you would probably want to do on a call like that is give the heads' up to liaisons that it may impact their agency.
DR. TANG: Okay. Good. Thank you, everyone.
(Whereupon, the meeting adjourned at 1:00 PM)