On January 20, 2011, Robin Fields hosted a conference call to explain what was in our Dialysis Tracker and to talk about the ways in which people could use the information. The following are a recording and transcript of the conversation. If you have a question that was not answered in this call, please feel free to send it to us at [email protected].
Mike Webb: Hi, I'm Mike Webb, and this week we have a special edition of the ProPublica Podcast. Last week, our senior editor Robin Fields conducted a conference call with journalists and members of the dialysis community to talk about how they could use the information in our dialysis tracker. Our hope is that dialysis patients can use the data to evaluate their own care options and what questions they might want to ask their caregiver, and we hope reporters and researchers can use it to find out what facts are illuminated by the material and how they can relate that to their own localities.
Robin is a former reporter for the Los Angeles Times and she's been recognized for the high quality of her work with a variety of journalism awards. She's currently a senior editor here at ProPublica, and here she is to give you some background on the dialysis tracker.
Robin Fields: Hey, thanks, Mike, and thanks, everybody, for joining the call. As you know, since November we've been writing about dialysis care, and as part of our reporting, we requested data from the federal government on individual dialysis clinics' outcomes and performance. The government had collected this data for many years but had released only a small portion of it to the public. The rest was distributed and used by regulators and was given to clinics themselves, but was otherwise confidential.
In response to a Freedom of Information request that we filed in 2008, the Centers for Medicare and Medicaid Services released this data to ProPublica late last year, and we published it in our Dialysis Facility Tracker. The tracker gives 15 measures for each facility, including information on mortality rates, hospitalization, infection, and transplantation. Several of these measures are risk‑adjusted, taking the characteristics of patients at that specific facility into consideration. Others compare the facilities' outcomes to state and national averages.
The tracker allows users to look up a specific facility by name or to find those closest to a specific address to compare them on all the various measures. Where possible, the tracker also attaches full Dialysis Facility Reports (DFR) – reports generated annually by the University of Michigan's Kidney Epidemiology and Cost Center using government data going back as far as 2002 – allowing for deeper examination.
We feel we've only scratched the surface of the kinds of analysis and reporting you could do with this information.
Mike: You mentioned that there were some other resources people could use. Can you give us an idea of what those are?
Robin: Obviously, this data is just one way into the world of dialysis care and in examining dialysis care. Another way to get more information might be to look at inspection reports or just by visiting dialysis facilities yourself or talking to patients who are at those facilities or staff members. This data is a tool meant to be used in conjunction with other tools.
Rich: Hi, Robin. I've got a couple of questions. Number one, there is no doubt there's a need for people to see what the results are for individual dialysis centers. But, as you know, most dialysis patients in‑center are not engaged. They're getting more and more debilitated as time goes on. Hopefully, when people first start dialysis, they're well enough that they would be able to use the DFR. But my first question is what's the strategy for letting people know that the DFRs are available for them to look at, since people are so disengaged?
Secondly, one of the problems that a lot of people have regarding selecting a dialysis center is they have no choice, especially with people with primary insurance who are first starting. They typically have to go with someone who is in‑network and can't select anybody outside of the network, even if they so desire. So what are the issues around that that you see that can be dealt with?
Robin: Rich, those are really good questions. The first question has to do with how to get the word out that this information is out there to people who might not be aware of it or might be sort of in a mode where they're more passive than active in how they approach choosing care or figuring out aspects of their care. I think part of that relies on other media reports. Some of that might rely on patient organizations. I know some patient organizations have connected their members to this database and directed them to this database. I think one of the only things we can do on that is sort of remind people that this is out there and also push the government as well to continue to share more information as they get it and to update their own website, Dialysis Facility Compare, with additional information. I think it's a long‑term process of getting patients to take a more active role in their care, and that's a cultural change and will take time. But part of it is about getting word out there through media reports, which we hope people will take our data and generate media reports of their own.
The second question as to how much realistic choice patients have, first of all, I suppose that choice in terms of providers is an ongoing thing. It isn't only something you can do at the beginning. You can continue to exercise choice as time passes. I also think that this information doesn't just lead you to make different choices; it also asks you to ask better questions about the choices that you have and about the places that you are getting care. This isn't only a tool to pick one place over another; it's also a tool to let you understand the place that you're at better than you do.
Mike: OK, we have Carol Bradbury in queue. Carol, go ahead.
Carol Bradbury: Do you have an opinion overall on which is the best dialysis provider overall? I know you can go clinic by clinic, but if we're shopping for a provider for our family, do we just have to go by the location, or would it be better to pick between the different companies that have clinics?
Robin: Generally speaking, people do choose dialysis services close to where they live, and I think that you'll see if you look at the tracker and the information on the tracker that it's almost impossible to make those kinds of generalizations looking at overarching trends in terms of providers. I think the tracker itself is helpful in telling you how to look at clinics within a particular area and make more‑informed decisions. But I certainly would not be in a place where I should be offering an opinion. I would certainly go to your physician and ask what he or she might think of the information that the tracker gives you. And the other thing is you also have other patients to use as a resource in that regard.
Mike: OK, Pison Anderson, go ahead with your question.
Pison: I'm a reporter for Better Buy. I just had a few questions. I was wondering how many people do you have currently using the tracker, if you know? If you have any of the demographic data on the people who are using it, such as distinguished by race or level of education and income. Also, I've heard anecdotally that, for example, black men usually do not fare very well on dialysis for long periods of time, and if you have any of the data on the data tracker actually organized by that set of demographics.
Robin: Those are all really interesting questions. I wish I had more of the information that you're looking for. We know that thousands of people have gone on the tracker. We can tell that. Who those people are, we have considerably less information about, although that would be interesting to know because it gives you some insight into who are the most active patients in terms of seeking out information about care. In terms of demographics offered by the reports themselves, they do give some demographic breakdowns of the patient populations at each facility. It's also possible, in aggregate, to get more information on how different types of racial or ethnic groups may do on dialysis I think for the system overall, although not clinic by clinic. That data is just not that fine‑grained yet. It may be in the future, although those would be very small numbers and might be difficult to read into.
Mike: OK, Roberta Michaels is next. Roberta?
Roberta Michaels: Thank you, Robin, for having this series. Since the series came out, I have had many calls from patients throughout the United States that have reviewed the Dialysis Facility Reports and as a result contacted me on where they could obtain information, as an example, on infection prevention, because their facility might have had a high rate of infection. So they are becoming more educated as to what the staff needs to do to prevent infection. I really thank you for this. This has been very, very good.
Another thing that I notice in review of the surveys – I have the actual California surveys on my website for the last couple of years, and in reviewing and analyzing those surveys, I'm finding that a lot of them do not have the Dialysis Facility Reports. I'm wondering if CMS gave a reason for not providing those.
And then the other thing is I would like to see more media involvement in this and more transparency to support provider accountability.
Robin: Well, thanks for calling, Roberta. We really appreciate it. The question that you asked about making sure that the PDFs get attached with the Dialysis Facility Reports to that facility's measure information, one of the things that we've run into is that occasionally, or not as occasionally, there are facilities that have had more than one provider number. We're in the process of trying to make sure that if that's the case, that the PDFs follow that facility, and we're working on that. So I think part of what you're seeing in terms of the PDFs not being attached may be an imperfection in our connecting up of the data that we are working on.
Mike: OK, John Sadler, you are up.
John Sadler: Well, I wanted to say that you have to be careful about the data you're using. I'm a nephrologist, and I've been doing this for over 40 years. The quality of the data is highly variable, and that's why on the CMS Dialysis Facility Compare site there is very limited stuff in there, because those of us who were on the committee that helped them set it up pointed out that some of the data they wanted to use was not accurate enough or complete enough to be beneficial. I suspect that the same is true of what you are getting. For instance, the infections in dialysis have only just fairly recently begun being reported in detail so that any information from before that is likely to be fairly spotty.
I just wanted to make the comment that when people talk about different provider companies, you have to realize they are not monolithic. Their central management may be a characteristic, but in individual facilities the people that work there do cast a large shadow and influence the nature of the care and the atmosphere of the place a great deal, and so there may be considerable differences. I lead a group that has just eight facilities, so we're small potatoes. But we are a non‑profit organization. It just takes care of patients, and we don't have to provide profits for our shareholders.
But I wanted to add that having read the articles that Robin wrote and having spoken to her before when she was researching this, I was really dismayed because you've drawn a lot of negative generalizations from some obviously clearly negative situations. I think that it's not as universally bad. I started doing this when everybody whose kidneys failed died, and when we found that we could keep them alive with dialysis we were very pleased. I think the patients were, too. For many people on dialysis, it still is a major nuisance, but it allows them to be functional.
Part of our problem is that when Medicare started covering people for dialysis and transplantation, it let down the floodgates and there's been no real selection of people for benefit from the treatment. We have a lot of people with a lot of co‑morbid conditions who would be debilitated if their kidneys worked. So it's a different population, and it's hard. I really would ask you to take a broader look at what's going on out here and not just damn us all because you found some bad places.
Robin: Well, Dr. Sadler, it's always good to hear from you, and I appreciate what you're saying. Obviously, I feel like what we wrote reflected the thrust of my reporting, but I appreciate your call, and I appreciate your participation in the conversation. I think what you have to say is important.
Mike: OK, we have a question that's been e-mailed to us from Lindy Washburn. Lindy writes, "How can we get access to the inspection reports? Do we have to OPRA or FOIA them for specific centers? What is the frequency that inspections are required, and if the date is not indicated on the data that you have, what does that mean?"
Robin: The inspection data is available publicly. It's all public documents. You can request them in a couple of ways. You can go to your state health department that supervises and oversees regulation of dialysis facilities and request them that way through public records requests. There are a couple of states that put theirs online, and those include Pennsylvania and Florida, and I think possibly Colorado, and there may be some others. You can also request them from CMS regional offices. If the material is not reflected in the PDFs or on the tracker, what that tends to mean is that they may be older or they may not have, for one reason or another, followed the facility if its provider number changed or something like that. The information in surveys can be really interesting, and I think that it's well worth seeking out.
Mike: Next up is Morgan Loew.
Morgan Loew: Hi, I'm here. Thanks for taking my question. I'm a reporter for the CBS TV station in Phoenix. I'm looking at ways to push this story forward in this area in light of a couple of things: one, major state budget cuts to our health services department, and two, the state potentially pulling out of the Medicaid enrollment program.
I'm wondering if you have any ideas for what to look out for about how either of those issues may affect the level of service here with these facilities and whether you have concerns after doing your reporting that the oversight with huge state budget cuts may affect the way states are regulating and looking at these facilities.
Robin: Well, certainly that was the implication of a story that we did earlier in December about oversight and about frequency of inspections. Going back to the previous question, the target is to do surveys about every three years on average. Certain facilities are considered higher priorities and get more inspections, whereas other facilities, maybe, the average interval may be closer to four years. I do think that, particularly in states where there are going to be cuts in jobs and cuts or furloughs or those types of things, the expectation would be that it would be difficult to keep up with all the inspection burdens that departments of health are asked to keep up with. I would watch that carefully, because that's been a big issue in California, and it's likely to become one for many states which are facing very difficult choices about how to use the limited resources that they have.
In terms of Medicaid and the effect that Medicaid cuts could have, this is really something that's sorting itself out in real time. I'm not sure that I am 100 percent qualified to answer that question. There are some patients that might also qualify for Medicare. Medicare is the primary payer. It looks like some patients get both Medicare and Medicaid. They rely on sort of a hybrid. Then there is a group that is Medicaid only, and those patients potentially would be affected by Medicaid cuts.
Then, of course, there's all the stuff that's going on with the health reform plan and whatever effect that might have, although that obviously won't affect the Medicare population. I'd be guessing, and it may be a state‑to‑state thing.
Mike: OK, Jim. Jim Hall, you are next.
Jim Hall: Yes, I am. Thank you very much. Hi. I work at a daily newspaper in Fredericksburg, Virginia. Your database lists six dialysis centers for our service area. One of them seems to be clearly superior to the others; its numbers are much lower than the others and the state averages. And then one of them seems clearly inferior to the others; its numbers are higher than the others within our area and higher than the state averages. My plan at this point is to try to talk to and visit the two centers and write, I guess you might call it, a comparison story. Do you believe that to be a valid strategy?
Robin: I certainly think it's a valid jumping‑off point for a story, and I would just continue to dig and look for explanations for why those differences might exist. I think that the data can take you a certain distance toward understanding that, and I would urge you to look beyond just the baseball‑card stuff and into the demographics of the patients and what you know and what you can bring to an understanding of the area and the demography of the area, the socio‑economics of the area, and so forth. But I think that that is a very interesting set of circumstances, and I would look at how much real choice patients have in terms of which facilities they can go to.
The other thing that you may want to look at in terms of comparing facilities is what types of services each one of those facilities offers because some may offer different modalities of care. Some may only offer in‑center hemodialysis. Some may offer nocturnal or night‑time shifts, things like that which may change their level of appeal to the patient population. So if you use the data as a jumping‑off point, though, I think you've got the potential for a really interesting story.
Mike: OK, Kate Howard, I see you're on the line, and I also have an e-mail from you. Go ahead.
Kate Howard: The first question that I had is that as I'm looking through our inspection reports on some our local facilities here in Jacksonville, I saw a handful of them had, after their inspection, been notified that they could be terminated, that their certificate could be terminated, if they didn't correct within a certain number of days. All of them have. I haven't seen any that were actually shut down. But I wondered what sense you got of how seriously states handle issues such as sanitation problems and patient records and things like that. Did you often see that these facilities were dealt with severely when they had problems like that?
Robin: Kate, this is a good question, and we dealt with it a little bit, I think, in our earliest story when we talked about the rarity of actual terminations. Usually, the warning that they give and the time limits that they give to give a plan of correction tends to be the final step. One of the things that we talked about also in that story is that in terms of choices of enforcement tools, CMS is more limited in the types of enforcement tools it has for dialysis facilities than it is for certain other kinds of facilities such as nursing homes, where they could impose fines and do various other types of enforcement.
With dialysis, basically, they can ask for a plan of correction. They can put a time limit that is more time certain on a plan of correction. In certain cases, if they walk into a facility and inspectors find what they consider to be very, very significant issues, they can say that they're not going to leave until those things are remediated to their satisfaction, sort of deal with it almost on an emergency basis.
The only other step that they can take if a clinic doesn't remediate those problems in the time frame that they've given in the way that they've asked is termination. But that's really a hard decision for regulators to make because they're very concerned about access to care and disruptions in care. So in the last decade or so, they've really terminated about 17 facilities, I believe, off the top of my head. So that's a very, very rare step.
Mike: Kate, do you have a follow‑up?
Kate: The only other question is the one that I sent via e-mail, which is just as you've been looking over this and trying to get local media to hop on, I wondered if there was anything in particular that you saw as the best opportunity to localize what you've started.
Robin: Well, I think that people have talked about looking at differences that exist within their communities. Another area that I think is particularly interesting is to look more closely at the first year of dialysis, which I think has become much more of a focus within the dialysis community in looking both at outcomes in the first year, which have been a significant issue, especially in terms of high levels of first‑year mortality, and also at the care that patients were getting before they began dialysis.
I think the dialysis system sort of inherits some of the burdens of a fragmented health care system in which patients aren't being sent early enough to specialists and are coming into dialysis on an emergency basis. What you may find is that those first‑year outcomes very much reflect that, and that's an interesting story, both about dialysis and about the larger health care system.
Mike: OK. We're going to take some more e-mail questions, but Linda Butt is up next. Linda?
Linda Butt: I had a question about the source of the data. Is that self‑reported to CMS and then CMS released the data? Or is it actually something that CMS goes in and determines on their own? How high‑quality is the data, is what I'm asking.
Robin: The data comes into CMS through a series of different kinds of forms, including claims forms, death notification forms. It, I think, is subject to some verification. One of the points that Dr. Sadler made, which is that certain aspects of the data that the agency feels more confident about than others...I think they've felt, for example, that the hospitalization data, they wish it would be more nuanced and things like that. That it would be cause‑specific. And I think that they're working on making their latest data collection system to try to make it so. I think we'll continue to see data get better as time passes.
I think that there's a high degree of confidence in the mortality data, there's a high degree of confidence in the clinical benchmark data. And I think there's a growing level of confidence in some of the other categories. But it's a process, and they've been expanding the categories of data collection over time. The data is not only collected by CMS, it is analyzed, as I said, by the University of Michigan Kidney Epidemiology and Cost Center, and they've been doing this for quite a while. I think that there's a reasonable degree of confidence in the data, enough that it's been in use by regulators for a number of years.
Mike: OK. Our next question is from Jennifer Calhoun with the Fayetteville Observer. She writes, "I saw on the tracker that one of the dialysis in our city showed that it had performed worse than expected for first‑year mortality versus expected. The percentage noted for that row was 101 percent. I'm a little confused on what that means in real‑world terms. How would I define that for my readers? Obviously, it can't mean that 101 percent of patients died in the first year."
Robin: I would refer you for a little more explanation on that to the actual PDF, which may be more enlightening on that than just the sheer numbers. I think what you're looking at is how that clinic did in actuality versus what would have been expected of them. I think what you'll find is that figure reflects that they had more than 100 percent more deaths than expected, based on their patients' characteristics. But I would urge you to go back to the explanation that we give on the tracker of the statistic and to the PDF. If that still isn't clear, call us back for more conversation. But I think that's what you're looking at. It's not a raw number saying that more patients than actually existed died. Absolutely, that would not be the way to interpret it.
Mike: Let's go to Shea Todin. Shea writes, "Why are the infection rates not labeled similarly to the mortality rates as higher than expected or lower than expected? And does the infection data collected include information that spells out whether infections are directly related to mortality rates at each facility?"
Robin: I think that the reason that they don't categorize infection rates as higher than expected or lower than expected is because infection rates aren't – as you'll see in the PDF report – really compared to an expected level of infection. Where you have those phrases attached to the numbers are where the reports compare actual results to expected results based on patient characteristics. These are raw percentages and they are compared to state and national averages. So that's why that's a somewhat different format.
And in terms of correlating one statistic to another, I think the reports generally stay away from that. But that's an area of data analysis that would be interesting to pursue on your own.
Mike: OK, the next caller is Nancy Scott. Go ahead.
Nancy Scott: I live Delaware and I've been on dialysis for seven years. I just wanted to emphasize that I had the opportunity to serve on a committee with Kidney Care Partners and we discussed first‑year mortality rates. And since we have this audience, I just wanted to give them an e-mail address where you can go and see some of the best practices that we did and it's Kidneycarequality.org. You'll find a lot of statistics about first‑year mortality rates. As a dialysis patient, I was very surprised at it. But I hope that this information can help social workers, providers, and dialysis patients.
Robin: Hey, Nancy, thank you very much for calling in and giving people that information. That's very helpful.
Mike: OK. I do have another e-mail question from Carol Bradbury. She wanted to know how often will be update the database and will we do any more investigation reports on those updates?
Robin: I think we will be updating the database in a couple of different ways at a couple of different times. First of all, we are going to do a better job making sure that all the PDFs get to the right places. We'll certainly update for the 2011 Dialysis Facility Reports when they become available. The other thing is we know that there are a couple of additional areas of categories of information that we'd like to add. For example, hours of operation or shifts, and different modalities of services that facilities offer that we'd like to eventually add. Certainly this is something that we have a commitment to long‑term. And we definitely have an ongoing interest on reporting more on dialysis.
Mike: OK. I just got a question from Lamour Daphne. His question is, "Are you aware of any studies showing the effect of Dialysis Compare introduced in 2001 on patient choices or facility quality measures?"
Robin: I'm actually only aware of one report that was done by a consultant about Dialysis Facility Compare, and the focus of it seemed to be about how they could get more people to use it. I don't know that they'd had much usage. And I think that they were looking to ask both professionals in the dialysis field, patients and their relatives about how they could make the site more useful, more helpful, and more appealing. Not much from that report actually was added to this site, but they did do that report. But that was, I think, at least five years ago.
Mike: Lamour also asks, "I see your website includes historical facility reports in PDF form. Is this information available in a downloaded form for research purposes?"
Robin: Not currently for us. We were given a database that only applied to the most recent couple of years. The older data was given to us only in PDF form, and we have not turned that into a database. So, sorry, there's going to be some boot leather that you need to use on that.
Mike: OK. Pison Anderson?
Pison Anderson: Yes, I had another question. I was wondering what was the impetus to start this project?
Robin: That's a good question. I sort of became interested in dialysis and began reading about it a while ago. I think I became more interested in it as the health care debate developed because it seemed like such a unique system and a unique experiment in American health care. And then I sort of just went deeper and deeper and deeper.
Mike: OK, we just got a tweet from @NWKidney asking, "Will you be proactive in gathering new data?"
Robin: In terms of gathering outcome data, I'm not sure we really have a way to be proactive, in that the data comes in from facilities to the government and then ultimately ends up in the hands of both the Kidney Epidemiology and Cost Center and in the hands of the SRD Networks. I don't know that we have direct access to that kind of data. There are other types of data that we can certainly try to collect, but that type of data really is not accessible to us directly.
Mike: OK, we have another question from Carol. Do you want to go ahead and pose your follow‑up question?
Carol: Yes. I actually have two. One is, how does the data in the database compare to another site that's called the United State Renal Data System? And my other question is, how involved did you get the physicians in looking at the data? Or since they send the patients there, how do you compare what the physician has involved in what happens in the clinic as opposed to what the provider has to do with the clinic?
Robin: To start with, the U.S. Renal Data System. The U.S. Renal Data System is a tremendous resource, and actually a lot of the data that they put in their annual reports sort of aggregates all the data that you see in the Dialysis Facility Reports that are reflected on our site. That's how those resources relate to one another. In terms of the effect and impact of physicians, certainly I talked to dozens and dozens and dozens of nephrologists and doctors in doing my reporting. We used several doctors as sounding boards in creating the tracker and figuring out how to display the data, and also in how to describe various things in our stories. The medical community has been a tremendous resource for us, and we've really appreciated their candor.
In terms of figuring out who's responsible for what in terms of dialysis care, not only do physicians have a lot of responsibility, as they prescribe every part of a dialysis treatment, they act as medical directors of facilities. But there's also the effect of the patient themselves, who may or may not be doing what they need to do to be comply with treatment programs. So there's a lot of impact that other people have aside from the people at the clinic and the clinic itself. So in terms of what the data says about all of them, it doesn't really parse out who's responsible for what, and that's something to keep in mind when looking at the data.
Mike: And the last question is, what is DFR, Robin?
Robin: DFR, as I have tattooed on my head is "Dialysis Facility Report," the three most wonderful words in the world.
Mike: All right. Listen everyone, thank you so much for the great questions and for calling in and for joining us. If you have a follow‑up question that you would like to ask, just e-mail us at [email protected] and we'll pass those along to Robin and she'll get back to you. And also note that we're going to be posting an edited version of this conversation online. And that should be available next week. We'll have a transcript to go along with it as well. You'll be able to find all of our dialysis work at ProPublica.org/dialysis. Thank you very much.
