The US: #2 in MRIs, #15 in Health Ranking

Pt 1 | Pt 2 | Pt 3 | Pt 4

- - -


i-f6aa68cec0112080597a5d626453a221-MA slice.JPG

Part 4 with Kelly Joyce, author of Magnetic Appeal: MRI and the Myth of Transparency, follows below. All entries in the author-meets-bloggers series can be found here.

WF: Is your story of MRI one that is common to medical technologies? Common to the commodification (in an economic sense) or routinization (in the sense of technical practice) of machines in daily medical practice? Common to a story about the consequences of health care practice in the 21st century? Anything?

KJ: The integration of MRI into medical practice is a story that shares issues of commodification and routinization with other medical technologies. But, at the same time, MRI also has unique features. The fact that the data is represented as anatomical pictures elicits cultural ideas about sight, knowledge, and recognition that other forms of technological output (e.g., graphs, numerical measurements) do not. It is interesting to wonder if MRI would have the status it has if its output was portrayed as numbers or in some other representational form.

My work focuses on the United States. Examining other countries and contexts will lead to different stories about MRI--both in regards to perceptions of the technology and its use in health care.

WF: You're extending your work to other countries?

KJ: Yes, I'm currently working on a project that examines perceptions and uses of MRI in Japan. Japan provides a good case study because it has national health insurance yet invests in MRI machines. Usually, these two entities do not go together in this way.

For a better sense of the international distribution of MRI machines, check this out:

i-e2ccab0c064d4e138f18be2279d174bc-MRIs per mill.bmp

Source: CIHI, Medical Imaging in Canada, 2004 Report.

WF: I see what you did. The figure is a visual aid that makes it obvious how MRI use differs in a way a simple list of numbers wouldn't have achieved. You used the visual to demonstrate the value of the visual. Do you know if or how that distribution maps onto the rankings of health care in different countries?

KJ: There are different organizations that gather and compare national health statistics. The World Health Organization is one of them, and, as you can imagine, is considered a reputable source. The WHO health statistics shows that Canada, Italy, and many other countries who invest less in MRI systems have better health indicators (e.g., life expectancy at birth, infant mortality, expectation of disability at birth) than the United States.

In World Health Report 2000, which reports on years close to the ones talked about in the CIHI report, overall health system attainment rankings were:

1 Japan
2 Switzerland
3 Norway
4Sweden
5 Luxembourg
6 France
7 Canada
8 Netherlands
9 United Kingdom
10 Austria
11 Italy

The U.S.? Number 15.

More specifically, World Health Report 2000 reported that life expectancy at birth for males in Canada was 76.2 years whereas life expectancy at birth for females was 81.9. In the United States, life expectancy at birth for males was 73.8 years, and life expectancy at birth for females was 79.7.

Subsequent WHO reports suggest that Canada, Italy and other countries continue to do better than the United States. In the World Health Statistics 2007, for example, Canada, Italy, the United Kingdom, Japan, Germany, (and more countries!) all had better health indicators (e.g., life expectancy rates, infant mortality rates, maternal mortality rates) than the United States. (You can check out WHO health statistics here.)

WF: Here's something else, about the history of technology and invention. In my Intro to STS class we sometimes read an article about the birth of MRI as a lesson in the non-technical factors that shape technical developments (it's by Tim Heppenheimer, from American Heritage's Invention and Technology, 2000). This guy Damadian seems really interesting in that story - he's been fighting for inventor credit for decades now and with some unusual tactics. But you actually interviewed him. What's your take?

KJ: Damadian and his research team were part of the development of MRI. They attended early international meetings, published articles, and worked on transforming the technology into a diagnostic tool. The struggle over inventor status that occurred and still occurs in relation to MRI is a great case study about what counts as priority, credit, and the like. As a physician, Damadian started off as an outsider (in comparison to the chemists, physicists and other researchers who worked on early MRI development). He also chose some strategies (e.g., taking an ad out in the Washington Post and New York Times to contest the 2003 Nobel Prize in Medicine) that are not part of scientific norms. These issues (e.g. strategy choice, professional training) may have cost him credibility.

The importance of the visual also comes out in recognition struggles. Although Damadian was the first to imagine (and patent) using NMR to evaluate human tissue, Lauterbur was the first to consider turning that data into anatomical pictures. This innovation (along with the use of gradients to map spatial information) proved crucial in terms of scientific recognition.

WF: That leads me to ask: if I want to use this book or parts of it in my undergraduate courses to help show how technologies are shaped by contexts beyond just technical details (a common theme in the engineering school here), what should I do?

KJ: Check out Chapter 2, "Painting by Numbers; The Development of Magnetic Resonance Imaging and the Visual Turn in Medicine." This chapter focuses on the entanglement of researchers' creativity, contexts, and design decisions in relation to the development of MRI technology. Students may enjoy wrestling with this case study. Alternatively, Chapter 5 "The Political Economy of MRI" is interesting since it illustrates how health insurance reimbursement policies, advertising, fear of litigation, and research and development funding contribute to how MRI is used in clinical practice.

WF: Finally: Foucaultian or Foucauldian? We can settle this here.

KJ: Who says Foucaultian?

- - -

Pt 1 | Pt 2 | Pt 3 | Pt 4
Categories

More like this

I'm just not sure of the point of those statistics. Are the # of MRIs supposed to correlate to the US health rankings? How does our # of MRIS compare to # of x-ray machines? If I'm not mistaken, the US has more medical technology than most other countries and I doubt this is very MRI specific. What is the # of MRIs per hospital? With a total of 5850 MRIs in the US (300million*19.5) that seems like a median of 1-2 machines per hospital. Does the UK have fewer MRI's/hospital or just fewer hospitals since the population is denser? While this interview was interesting, it really hasn't brought forward a unique point of view or an interest in buying the book. Perhaps when full reviews are out, I might change my mind, but hopefully there's much more new than what's listed here.

As for other books. If readers are interested in the history of medical imaging across modalities, "Naked to the Bone: Medical Imaging in the Twentieth Century" by Bettyann Kevles is a phenomenal book. It covers some sociology, but is more from a historian's perspective.

If you want to read more about Damadian and his role in the creation of MRI, there is a good series of semi-lay articles in the Journal of Magnetic Resonance Imaging Volume 20, Issue 2, 2004. Some articles support him and some downplay his role.
http://www3.interscience.wiley.com/journal/109561264/issue

I should note that Dr. Joyce is bending history a bit too much toward her thesis about the centrality of images by writing Lauterbur was the first to consider turning that data into anatomical pictures. This innovation (along with the use of gradients to map spatial information) proved crucial in terms of scientific recognition. The innovation is what cut data collected times by several orders of magnitude made MRI practical to collect data from people. Sure images are nice, but this was the key discovery that turned MRI into a useful clinical tool. There is a good argument that Damadian was also worthy of inclusion, but that doesn't downplay the scientific and clinical significance of the other's findings.

BTW, in case you haven't figured it out by now, I do research with MRI.

I do research with a computer.

I do research with citation databases.

I do research with structuralist tools.

I do research with books.

Just because I use these various items to do my research, I would never presume to know everything there is to know about them their history, the ways in which they are produced, or how they function. How preposterous! If someone came along and said: Ive been hanging out at the Thomason Scientific [where ISIs web of science is pieced together] for seven years and Ive watched the ways in which the programs behind the databases have evolved, the ways in which weve learned to rely on these databases rather than card catalogs (or perhaps the cumbersome Chem abstracts), and the hidden voids and gaps in knowledge that result in trying to present a database that is truly encyclopedic [and of course its not, thats why rivals like google can construct their own version] I cant imagine having the audacity to tell that person that because I *use* web of science for my research that I somehow know more about it.

This interview has been one of the most intriguing that Ive read in this series. As someone who dabbled a bit in the history of NMR and someone who unfortunately has had many run-ins with MRIs of late, I found the stories that Prof. Joyce wove together to be evocative and provocative. Evocative of the ways in which images and image producing machines do *speak* to us in ways in which more traditional machines do not (and this is of course why Damadian didnt get the Nobel along with his colleagues the production of the image won, not the idea of using it on biological entities). Provocative because it says a lot about the ways in which modern health care has evolved. Its not just doctors relying on these imaging technologies, but the patients as well. Think of the parallels to the growth in use of fetal imaging technologies. As people who chose not to use them throughout our pregnancy, my spouse and I were treated with both surprise and uneasiness. Why *wouldnt* you get one? But why would we? What can we learn from that image that we cant learn from simply being in touch with the development through other senses? What are the consequences and limitations of such technologies? While the fetal imaging technologies have been written about at great length, this is the first time Ive seen someone tackle the topic in another arena of medicine and its high time. What is it that we *see* in the MRI? The myth is that it provides someone an unobstructed view into the body. But transparency only exists if we ignore the massive amount of electrical, mechanical, and magnetic mediation that takes place. If MRIs are transparent, then why can only certain trained individuals see with them? Prof. Joyce has opened a wonderful black box for us full of all sorts of goodies. Thanks.

jar,
Part of research is to understand the limitations of one's own knowledge and to respond to relevant criticism with future work that either rebuts or supports your original thesis. My comment focused on some criticisms of Dr. Joyce's work as presented in this interview, but that doesn't means there isn't value here too. If there was no value, I probably wouldn't have bothered to comment. Assuming she ever reads this, I hope this provides some thoughts for future research and improving her current work.

I won't speak much more of Damadian here, but there are many reasons he didn't receive the Nobel. Perhaps the fact that he didn't produce anatomical images is one of them, but it's definitely not the only reason.

As for no one tackling these issues before, I suspect you'd also like the work of Kevles in
"Naked to the Bone: Medical Imaging in the Twentieth Century." As you can tell from the title, she also focuses on the issues of transparency and imaging.

The myth is that it provides someone an unobstructed view into the body. But transparency only exists if we ignore the massive amount of electrical, mechanical, and magnetic mediation that takes place. If MRIs are transparent, then why can only certain trained individuals see with them?

I do hope that Dr. Joyce and responds directly to what the myth is, but I'm not sure you have it here. Many data collection technologies like blood tests high significant amounts of electrical, mechanical, and magnetic mediation behind the creation of a few simple numbers. Do we say "blood tests: they myth of hematocrit levels"? And the beauty of MRIs is that anyone CAN see the images. Experts are trained to see things others don't. It's like saying only experts can look at artwork. An expert might be able to see and understand more about a painting than I can, but that doesn't mean the paining is completely inaccessible to me.

First, I want to thank people for their comments and postings. It is fascinating to hear about people's experiences with the technology and their thoughts on MRI.

MRI is clearly a valuable diagnostic tool. It is an important part of clinical medicine and should be included in medical practice. We didn't talk about my research questions and methods in the interview, so let me backtrack and discuss these here.

Instead of asking is MRI good or is MRI bad, my book addresses the following questions: (1) What is the history of MRI? How was the technology developed? (2) What are the common messages told about MRI in nonfiction and fictional media? (3) How do the health care professionals who work with MRI (e.g., neurologists, radiologists, technologists) understand the technology? (4) How do their views of MRI relate to common media narratives? and (5) Do institutional, political, cultural, and economic factors contribute to MRI use in medical practice? If so, how?

To answer these questions, I conducted seven years of in-depth research. This research included: in-depth interviews with technologists, radiologists, and referring physicians; in-depth interviews with researchers who contributed to the development of MRI (e.g., Paul Lauterbur, Raymond Damadian, John Mallard); fieldwork at three imaging units and five MRI related conferences; content analysis of print and visual popular texts; archival research; and targeted literatures reviews of medical journals. I used a grounded theory approach to systematically analyze this data. Grounded theory was initially presented by Anselm Strauss and Barney Glaser in their book The Discovery of Grounded Theory: Strategies for Qualitative Research (1967). There are many aspects to grounded theory, but for the purpose of this conversation, the key idea is that a researcher starts analysis by reading through interview transcripts, field notes, and other qualitative date multiple times to determine reoccurring themes and codes. After this process is completed, the researcher methodologically codes each document to see how and when such ideas occur. This analytical approach is called an inductive approach because it starts with the data; the categories and analytical claims are thus built from reoccurring themes/issues in the data. Of course, the researcher is not a ventriloquist who speaks for the data, but instead performs intellectual work through the creation of categories and analytical claims.

What does this have to do with transparency? Well, I used a grounded theory approach to systematically code the reoccurring themes in print and visual popular culture (e.g., newspapers, magazines). One of these themes was the idea of transparency. That is, many sources talked about how the body is interchangeable with MRI exams, creating the idea the MRI provides a transparent window into the body. Individuals may accept or reject this idea, but its repetition in popular culture creates a climate where it is acceptable to think that the image is equivalent to the body. I was curious about this idea, and wanted to see what technologists, radiologists, and referring physicians who worked with MRI thought of it. Did it fit with what they knew from working with the technology in clinical medicine?

Through my fieldwork and interviews, it became clear that MRI exams were valuable diagnostic information but they were not "equivalent" to the body. Thanks, jar, for your insightful comments on this issue in your post. There are many layers of translation that occur during the production of MRI exams. First, technologists make decisions about how much of the body to measure and thus include in the exam. Second, artifacts can be present in the image and may at times look similar to pathology. Third, radiologists (or other MDs) interpret the images into a written report. Even the best radiologists make mistakes and may interpret normal anatomy as pathology or miss pathology in an exam. Finally, the referring physician must analyze how the findings gained from MRI use fit with other diagnostic information (e.g., information obtained from other tests, clinical exam, patient history). There are times when the results from the MRI scan do not fit with the other clinical findings. One such example is multiple sclerosis (MS). MS lesions in the brain often show up in an MRI scan and MRIs are part of a diagnostic work up for this disease. Despite this, there are cases when a patient has all of the clinical findings of multiple sclerosis, yet the MRI exam appears normal. As one neurologist I interviewed pointed out, "The MRI scan is probably negative up to 25% of the time in [MS] cases, so I would usually trust my exam much more than the MRI scan." In this situation, the physician has to rely on other indicators to produce a diagnosis.

The other important issue here is that each of these moments (technologists' choices, MDs' interpretation of images, referring physicians' synthesis of the data) are affected by institutional contexts. For example, slice thickness (one of the choices technologists make about the content of scans) involves a trade-off between data and time. With MRI technology, decisions about slice thickness are also decisions about the level of anatomical detail included in an image and the length of exam time; short scan times and increased spatial resolution are mutually exclusive effects of the technique. Right now, many U.S. imaging units are under pressure (by administrators, by the effects of decreasing reimbursement fees) to speed up production and move patients through MRI exams faster. While many imaging centers resist this pressure, it exists and can affect the quality of image content.

There is so much more to say about these contexts, and I discuss them in more detail and complexity in the book (chapter 3). In the book, I include many quotes from technologists, radiologists, and referring docs to illustrate the points discussed above, but time/space constraints prevent discussion of them here.

By presenting this research, I hope consumers will shift their questions about MRI exams. When consumers assume scans are transparent windows into the body, they may not ask questions about the quality of the imaging center, the quality of the image interpretation, or how the findings fit with other diagnostic information. If consumers understand the layers of translation that occur, they may ask, "Was the physician trained in the interpretation of MRI scans?" or "Is this imaging center known for high quality work?" Does asking these questions devalue MRI exams? I don't think so, but everyone will have to sort out their own views of this issue.

At a presentation I recently gave on MRI and the issues discussed above, an audience member asked, "So, when I look at an MRI exam, is it me or not?" It is a great question--one worth pondering.

By Kelly Joyce (not verified) on 08 Apr 2008 #permalink

First off, let me say that this interview was one of the most insightful interviews that I have read in a long time and I look forward to reading the book.

I debated responding to this forum and recently saw Dr. Joyce's response. I am surprised she was so gracious to one individual in particular. I am a researcher that works in the field of health administration and am quite familiar with MRI. BSCI, your point about looking at MRIs in terms of per hospital is blatantly wrong. Anyone familiar with health science research relating to technology and hospitals knows that researchers do not look at technology by institution, but by number of hospital beds, if looking simply within the US. How can you equate a hospital with 40 beds with a hospital with 500 beds. Moreover, imaging is also done at imaging centers that would not be included in your MRI by hospital consideration. This raises a particular issue in the context of comparing MRI machines between countries, since more machines may be housed outside of hospitals in particular countries rather than others. This is noted in the expression that is commonly used these days to describe healthcare in the US: "we do not have health care in this country, we have sick care." Therefore, there is more hospital ownership of MRI than in certain other countries, e.g., the US, than others. For purposes of standardization, the statistics regarding the number of MRI machines would be properly presented as machines/million, just as Joyce has done.

Unto efficacy and safety. As Joyce is a sociologist, she may also mean a more global definition of efficacy and safety. If efficacy is meant to mean is the overall treatment efficacious in producing better health outcomes using an MRI-here we have a more interesting point. Based on her interview, it seems like her book will show how each actor, whether it be the state, the hospital, the physician, the patient, lawyers the advertiser, the biotech company, technician, the patient, and different players in popular culture make sense of and relate to MRI individually and how they negotiate and cooperate to make sense of MRI in a larger context. When all the players are combined into the larger system and the players understood as part of a system is when the analysis becomes interesting-at least from my perspective. The best way to find out exactly what she is saying though would be to actually read the book before judging it.

I also cannot imagine reducing a complex discussion of MRI to a simple binary argument of MRI is good or MRI is bad. Instead, she seems to view MRI as a technology situated in contexts and describes those contexts. An MRI does not tell you what is going on or show you the problem, it is simply a tool that helps humans determine the problem and it is a tool with lots of human intervention before getting to the final product, unlike a simple photograph that has less intervention. Joyce desribes how human actors use that tool in a certain way and humans then tell you what is going on.

Is it a positive tool or a negative tool? The answer is neither. It depends how it is used by the humans involved. For example, a knife can be used to cut food or to open boxes, but it can also be used to stab someone. The knife itself is just an object. It is humans that give it meaning. As Shakespeare said nothing is good or bad, but thinking makes it so.

Again, I do not fully know what she means and I certainly would not judge a book that is not out yet by its cover, especially by its back cover. I also would not claim to be an expert or a sociologist on MRI and know more than someone who has researched the topic for more than 7 years! Joyce certainly also does not say that MRI needs to stop being used! Especially in cases of neurosurgery! Don't worry BSCI no one is trying to take MRI from you!

This is why I debated responding on this blog--because I did not want to take away from Dr. Joyce's interview and scholarship by addressing one individual's misinterpretations. Thanks for the interview.

RSC

I'll respond to Joyce's excellent post too, but I want to respond to RSC first. I wrote some critical comments here, but I tried to frame them in a way that was polite and it was possible to get a response. I also very clearly said I was working from my understanding of her work based on this interview since I do not (and cannot) have the book, since it isn't even for sale yet! I was hoping to get some clarification to my questions. Although it's not always easy to tell with writing, I tried to be respectful of her research while asking for more details.

I very well might buy the book at some point. I should also note, that I was mentioning another book here because I thought it might also be of interest to others reading this post and not to detract from Joyce's work.

As for my stats question, I was trying to say that # of MRIs per country vs healthcare ratings is poor comparison. I gave one potential reason why. By your continuing to list the complexities of this # of MRIs measure, I think you actually agree that # of MRIs doesn't mean much. As you said, this number is additionally confounded by the free-standing MRI centers. I'm curious if Joyce discusses this in her book, but one of the more interesting aspects of some of these non-hospital systems is the large number of them which are supported by whole-body MRI scans or other scans that aren't requested by doctors and are paid for completely out of pocket. The fact that many people in the US will pay for these scans, which have no clear health benefit, does explain some of the origin of the disconnect between # of MRIs and health quality. For that matter, Japan and the US also have fairly significant MRI research programs. I'd venture that at least 250 MRIs in this country are primarily devoted to research and have minimal or no relevance to patient care. Are these included in the statistics? I'm mainly trying to say that there is no reason why I'd expect the # of MRIs to have a clear link to national healthcare quality and I'm not sure why this was such a big focus here.

I also cannot imagine reducing a complex discussion of MRI to a simple binary argument of MRI is good or MRI is bad.
I think this is what jumped out at me in the publishers book blurb on Amazon and some comments in the interview. It seemed to be presenting a binary view that MRI is bad. I doubted this was actually the point of view of the author and I wanted clarity on this issue. Joyce mostly addressed this in her comments.

Dr. Joyce, Thank you for your detailed response. I'm sorry if any part of my original comments were taken as insulting. Nuance is fairly hard to do in blog comments (and blog interviews for that matter). I am obviously focusing on parts where I have questions or critiques, but there are many more things you write with which I agree.

The five questions you list sound very interesting and some of these were glanced at in the interview. As I mentioned in my last comment, it seems the publisher's blurb as listed on Amazon is a bit more of the "MRI is bad" variety. I know authors have varying control of what is included in these blurbs and I'm curious what you think about it.

I also noticed that your first question seems to focus on history of MRI and interviewing the earlier developers, but it doesn't seem to include the continued research and transitions of new methods into clinical settings. Did you look into the continued interaction between research and clinic? If not, was there a methodological reason or was it just that you didn't want to work for a Harry Collins duration of time before writing a sociology of science book on the topic?

Your multiple sclerosis example is interesting. From the MS doctors with whom I've worked, the MRI is a good benchmark, but I don't think it was considered part of the diagnosis. Much of the research I saw was on whether certain MRI scans would be able to better predict which clinically isolated syndrome (CIS) cases would become MS.

On the new technology front, methods such as anisotropy maps have the potential to include better information about MS. As an aside, the word "map" also seems to come up quite often in MRI work. Did you see this too in your documents?

Another similar example is Alzheimer's disease where there really isn't any golden standard of diagnosis. I've seen MDs get the symptom list and sometimes see an MRI and say it was definitely an Alzheimer's brain and sometimes be unsure. (Though MRI is helpful in distinguishing Alzheimer's from fronto-temporal dementia)

I'm definitely familiar with the compromises of resolution vs. scan time. Even as methods are developed to make it possible to get the same resolution in less time, they also make it possible to get lower resolution faster. I should note that sometimes longer isn't better. In a person or a body-part with more motion, a fast but low resolution image is sometimes better than a slow, high resolution image with motion distortions.

I really like your second to last paragraph and I definitely agree that making people think more about what an MRI is and ask questions to their doctors is beneficial.

"So, when I look at an MRI exam, is it me or not?" It is a great question--one worth pondering.
I have been pondering this one a bit and I'm curious if that person would say the same thing about a digital photograph that was light converted to RBG color data with many image transforms. In some ways MRI is more intimate because it is inside you, but in others it is at least as much "self" as a digital photo.

I am trying to find out where and how to get a low cost MRI done. I have no health coverage and no known problems but want a baseline to compare to or to discover a problem early. Years age in Salt Lake City i saw a billboard advertising MRIs 2 for $500. I want that. Please help me.