Journal of the American College of Radiology
Volume 9, Issue 1 , Pages 64-68, January 2012

Renegotiating Expertise: An Examination of PACS and the Challenges to Radiology Using a Medical Anthropologic Approach

  • Allison Anne Tillack, MA

      Affiliations

    • Department of Anthropology, History, and Social Medicine, University of California, San Francisco, San Francisco, California
    • Corresponding Author InformationCorresponding author and reprints: Allison Anne Tillack, MA, University of California, San Francisco, Department of Anthropology, History, and Social Medicine, 3333 California Street, Laurel Heights, Suite 845, San Francisco, CA 94143-0850
    • ,
  • Richard S. Breiman, MD

      Affiliations

    • Department of Radiology and Biomedical Imaging, University of California, San Francisco

Article Outline

Purpose

The aim of this study was to examine how the adoption of PACS has affected the professional relationships among radiologists and referring providers and to evaluate the effect of PACS on perceptions of radiologists' roles in patient care.

Methods

A medical anthropologic approach was used to assess the impact of PACS among radiologists and a community of clinical subspecialists at a large academic medical center (n = 40). Data collection techniques included 3 months of ethnographic participant observation during the routine medical practice of study participants as well as semistructured interviews and archival research. These data were then analyzed to identify behavioral and narrative patterns and themes among the study populations.

Results

The difficulty of establishing and maintaining relationships of trust between referring providers and radiologists due to the drop in post-PACS reading room visits emerged as a major source of concern for study participants. By interacting primarily over the phone or at weekly conferences, radiologists felt that they had fewer opportunities to build personal relationships with other clinicians. Meanwhile, the specialist referring providers stated they generally consulted only radiologists with whom they had established personal relationships and otherwise preferred to interpret their patients' images themselves.

Conclusions

Generating and sustaining relationships of trust and effective communication are vital for radiologists to communicate their expertise in medical imaging to referring providers. Because PACS have caused a reduction in referring provider visits to the reading room, radiologists must seek out new opportunities to form personal relationships with other physicians.

Key Words:  Medical anthropology , communication , expertise

 

Back to Article Outline

Introduction 

The widespread adoption of PACS has revolutionized the practice of radiology by increasing radiologist productivity, virtually eliminating lost films, and allowing images to be viewed throughout hospitals, in clinics, and beyond. PACS has also drastically reshaped how radiologists interact with referring providers. Despite studies that explore the efficiency, effectiveness, and reliability of PACS, there is a striking absence of research on the communication and interphysician relations that facilitate these practices and efficiencies. Therefore, the purpose of this study was to use the techniques and methods of medical anthropology (1) to examine how PACS adoption has affected the professional relationships among radiologists and referring providers and (2) to evaluate the effect of PACS on perceptions of the role of radiologists as members of the patient care team.

Why Medical Anthropology? 

Broadly, the field of sociocultural anthropology identifies and examines current patterns and processes of cultural change and human behavior, “with a special interest in how people live in particular places, how they organize, govern, and create meaning” [1]. A subset of this broader field, medical anthropology focuses on the study of human health, disease, and healing systems, seeking to understand them within larger social, cultural, political, and economic contexts. Medical anthropologists primarily use ethnography, a qualitative research technique that includes participant observation, interviews, and archival research to investigate how people's health shapes and is shaped by social and cultural forces and norms. Participant observation is often the primary means by which anthropologists gather data, which “involves placing oneself in the research context for extended periods to gain a first-hand sense of how local knowledge is put to work in grappling with practical problems of everyday life” [1]. Ethnographers “embed' themselves into study populations, facilitating observations of everyday behaviors and practices, the development of relationships of trust, and movement beyond standard rhetoric offered to perceived “outsiders.”

Unlike quantitative research methods that test the relationships among measurable variables, the goal of qualitative methods, such as ethnography, is to explore and understand the meanings individuals or groups give to social, political, or technical problems [2]. Qualitative methods, such as ethnography, are particularly effective for generating rich, detailed, context-specific data about how people speak, behave, and think about complex and emergent situations. Although qualitative techniques may lack the ability to assess statistical significance, the data that are gathered can suggest important themes, patterns, and hypotheses for future research that can be evaluated using quantitative methodologies. Thus, rather than perceiving “anecdotal evidence” or “personal bias” as problems to be overcome, anthropologists consider these to be important sources of data, as they provide insight into how people think, remember, and perceive events.

Background 

Medical imaging technologies, such as radiography, CT, and MRI, are a cornerstone of modern medical practice. Traditionally, radiologists have been responsible for and considered experts in the interpretation of medical images. Before the capacity to digitize and distribute medical images (via PACS and other systems) was developed, a clinician requesting an imaging study was usually required to consult a radiologist and look at the appropriate films on a view box located in a radiology reading room. In his study of radiologists' behaviors when interpreting film-based images, Saunders [3] found that the images themselves were an important source of professional authority for radiologists, who acted as the “voice” of the images and translated the scientific knowledge generated via CT, MRI, or radiography into the language of clinical medicine. Face-to-face interaction between radiologists and clinicians was frequent, and radiologists had an integral role in communicating with and among their professional colleagues. Indeed, radiologists have historically been able to lay claim to and control the majority of medical image production and interpretation, a major source of both economic and knowledge-based power [3, 4, 5, 6, 7].

With the development and adoption of PACS, clinicians of all types have the ability to examine and interpret medical images without interacting with radiologists. Also, with radiologists using the speech recognition software that usually is adopted together with PACS, clinicians receive radiologists' reports as soon as they are dictated. As early as 1999, studies showed as much as an 82% reduction in the rate of in-person consultation for general radiography [8], sparking a wave of concern among radiologists that they might be “considered as disembodied functionaries, more akin to servicing technicians than professional colleagues” [9]. Still a relatively new technology in 1999, filmless radiology and PACS have now been widely adopted and are seen as improving productivity, lowering cost, and improving patient care. However, concerns remain that radiology is vulnerable to the erosion of its domain of expertise. As the importance of imaging informatics technologies continues to grow, “turf wars” over the control of image production and analysis [10] raise questions about the current role of the radiologist and the future of the specialty. The crucial change involving the transition from the view box to PACS workstations involves both a redefinition of radiologists' expert diagnostic interpretation and also a shift in the practices surrounding that visual expertise.

Recent research in medical anthropology, science and technology studies, and medical sociology indicates that medical technologies are an important site for redefining medical knowledge, with conflicts often arising over practices and agendas [11, 12, 13]. Furthermore, social scientists have found that when engaging with new technologies, people enter into unfamiliar and uncertain terrain, producing concern over the possible destabilization of identities, professional futures, and control over resources [11]. The aim of this study was to build on this area of research and examine the specific ways that PACS has changed how radiologists define their professional identities and establish claims of visual expertise with referring providers, as well as to explore the potential impact of these changes on patient care.

Back to Article Outline

Methods 

For this study, the primary author conducted 3 months of ethnography at a large tertiary care academic medical center in northern California. Ethnographic observations were focused on radiologists and a community of clinical specialists (neurologists on the neurovascular service or “stroke team”), who rely heavily on imaging for their work. These two study groups included both men and women with a broad range of levels of medical experience and training (n = 40). Observations of the daily practices and interactions of these two groups took place in reading rooms, at multidisciplinary conferences such as tumor boards, and on daily patient rounds. In addition, semistructured interviews with 10 radiologists and 5 neurologists specializing in stroke were conducted. The interviews focused on perceived changes in interactions among radiologists and referring providers and perceptions of radiologists' roles in patient care before and after PACS implementation. In addition, archival review of scientific and popular literature was conducted to identify discussions relevant to this project.

Analysis of ethnographic data is an iterative process that involves detailed reading and analysis of interview transcripts and field notes for the purpose of identifying recurrent themes and patterns in people's behaviors and speech. Therefore, verbatim interview transcripts and field notes (the written record of daily observations and interactions of the ethnographer) were analyzed and coded to distill emerging areas of interest and patterns. These themes and patterns were then correlated with relevant literature. Special attention was paid to discrepancies between what people said and what they did.

Back to Article Outline

Results 

Analysis of the collected data indicated that radiologists in the study believed that they interacted with referring providers less frequently than they used to and as a result were less likely to have relevant clinical information about patients. Furthermore, radiologists found it more difficult to communicate clinical recommendations, relevant research, or alternate diagnostic options with the patient care team. In addition, all radiologists expressed concern about the impact of these trends for the future of radiology and on patient care.

Although radiologists believed that the ability for referring providers to access images via PACS was important to improved patient care, they also worried that without radiologic guidance, referring providers might miss important findings. For example, one attending radiologist noted that “specialists can look at the images without talking to us [radiologists] now, and I think they can be lured into a sense of comfort and miss critical issues that might be outside of their own specialist knowledge.” Similarly, a radiologist in her third year of residency said,

I think they [specialists] can get misled by thinking that they have more knowledge than they do. PACS does allow clinicians to see images more often, and sometimes they develop a sense of what is important and what isn't, but they can also get misled. And if they hadn't been looking at those images, that wouldn't have happened, because they wouldn't have thought that they saw something that actually wasn't real.

Observations of and interviews with neurologists in the study supported this perception; neurovascular service attending physicians and a large majority of neurology fellows and senior residents expressed belief that PACS allowed them to see images frequently enough to develop significant expertise in interpretation. At the study site, images were not embedded in radiology reports. Often, the specialists would look at the image but not the dictation, preferring to rely on their own “reads” and clinical knowledge of patients. The dictation was consulted only when images had unusual or ambiguous findings. For example, one neurologist commented, “We call them [radiologists] if there's something difficult … but otherwise we don't really talk to them, since images are much more a part of our daily management of patients, just because of their accessibility with PACS.” Similarly, another neurology attending physician stated,

I tend only to talk to them [neuroradiologists] if there's a really complicated vascular picture I'm trying to work out [on a patient's images], or if it's something I thought was a stroke that turns out not to be a stroke, and I want to figure out what the hell it is.

Yet another member of the stroke team said that,

In the acute setting when someone has a stroke and is in the emergency department, we have our residents look at the images on PACS and then a senior person, an attending like myself or one of the fellows, views the image as well, and then makes a decision about emergent treatment. We make a lot of decisions from home … we have our Web-based PACS that we can look at from home, so I wake up at 1 am and stagger down to the computer and look at the thing. In the middle of the night, there's no neuroradiologist around. There is a radiology resident, who has less experience reading these than we [neurologists] do. So I look at the images, and then tell folks what we're going to do.

In this instance, the neurologist is reading his patients' images and making treatment decisions without consulting a radiologist. In fact, examination of images was a standard part of daily rounds observed by the primary author. After visiting and examining each patient, the team gathered around the PACS station located on the ward, went over any imaging studies done the previous day, and made decisions about the care plan for the present day. However, one neurologist admitted that this practice of viewing images without consulting radiologists could have negative effects on patient care, saying that,

Getting a PACS station up into the Neurology ICU, [intensive care unit], which changes how we round, did disconnect us from the verbal discussion with the radiologists and has led to some medical errors. Specifically, we look at a scan and say, oh, it looks negative, and the wet read isn't there or we don't look at it, or the wet read was done but the night-time radiologist missed something, but the formal read happens but then we don't go back and actually look at the formal interpretation, and so don't realize that the discussion we had on rounds was wrong. That's rarely lead to anything significant, but it's a frightening potential side effect.

However, it is important to note that attending neurologists and fellows only felt comfortable interpreting types of images that they saw frequently and relied on radiologists to interpret less familiar imaging modalities and body locations. For example, one attending neurologist on the stroke team said,

I see a lot of non-con [noncontrast] head CTs, head MRs, and CT and MR angiograms. So I'm pretty comfortable reading those types of studies. But if for some reason my patient needs a CT abdomen/pelvis, I'd be lucky to identify the spleen. So if it's out of my comfort zone, I always look at the report [from the radiologist].

Another stroke team attending physician stated that,

I think that I could read a CT scan compared to a general radiologist better for things that are within the neuro axis, but things within the face, maxillary bone, neck, I wouldn't do anywhere near as well as a general radiologist probably. Because I'm just very focused on a particular part of the puzzle. Angiography, looking at blood vessels, I'm probably better than most general radiologists. In terms of neuroradiologists, I'm probably as good as [a neuroradiologist], in vascular imaging, because I've done it for 15 years and I've seen a lot. But, I would defer to or potentially challenge a neuroradiologist about a particular finding, and if we get to that level, I'll often get another image to confirm and solve the dilemma, if there is one. And I would say, that in those cases that I disagree with them [neuroradiologists], they usually win [laughs]. And so it keeps me humble.

Interestingly, trust emerged as a major point of concern among both radiologists and neurologists who participated in the study. Neurologists said that they were unlikely to seek out the opinion of a radiologist unless they already had a solid professional relationship with that radiologist and felt that they could “trust” that radiologist's interpretation. As one attending neurologist stated,

I feel like with PACS, it's put into very sharp focus, which I don't think used to be there, the difference between good radiologists and great radiologists. And so, you used to say, really, anyone can help me with this [question about imaging], and now you get to the point where you say, I know just as much as radiologist X, Y, or Z, but this one [image] needs to be shown to Dr So-and-So because he has such a better perspective because he's seen more of radiology than anyone … more neuroradiology than anyone. So it's created a situation where I'll go down to the reading room not just to find out what the radiologists have to say, but to find a specific person who I want to ask a question to.

By interacting primarily over the phone or at weekly conferences, radiologists felt that they had fewer opportunities to build relationships of expert authority and professional trust with other clinicians. In addition, radiologists would talk about clinicians they “trusted” to order the appropriate study or give an accurate history on the request form. As one senior radiologist commented,

We [radiologists] knew all the clinicians intimately before. And then with PACS, this intimacy disappeared. Before [PACS], I knew the face, name, wife's name, and kids' names of all the clinicians, but now I don't know who you are if you joined the medical staff after we got PACS. Now we're operating in a void because there's no history of the patient on the written image requests. Before, when a clinician showed up, I could ask them and find out what's really going on with the patient.

With PACS, an important opportunity to find out “what's really going on” with patients has been diminished for radiologists. Yet the implications of this PACS-related shift in patterns of interaction extend beyond communication about individual patients. The primary pre-PACS mechanism, by which radiologists established professional trust and relayed their mastery of knowledge about imaging, involved daily, face-to-face meetings in the reading room [3]. This core set of interactions provided the basis for other types of interactions to work, for example, calling the reading room for a “wet read.”

Anthropologic studies have shown the importance of social interaction in establishing and maintaining jurisdictions of expertise and professional trust [14, 15, 16, 17, 18]. Expertise is knowledge based, but also interactional and performative. Without opportunities to create and reinforce relationships of professional trust and displays of expertise in image interpretation with which to “convince” referring providers of their expertise, radiologists are at risk for losing their status as imaging experts.

Back to Article Outline

Conclusions 

In his book The Scientific Life: A Moral History of a Late Modern Vocation, historian of science Steven Shapin [19] argued that “we cannot understand how various scientific and technological knowledges are made, and made authoritative, without appreciating the roles of familiarity [and] trust.” Although it may seem old fashioned in this era of numeric accountability and performance metrics, trust and effective communication among physicians remain more important than ever, especially given the trend toward the increasing specialization of medical care. Trust is not simply a matter of exchange of information but involves certain expectations, risks, and actions that must be actively reproduced and renegotiated through social interaction [20]. Care of patients requires trust in the competence of other specialists on both an individual and a group level.

PACS has revolutionized medical imaging, but it has also disturbed the mechanisms by which radiologists formerly communicated their competence and expertise to referring providers. This development has already begun to erode the relationships of trust among communities of radiologists and referring providers and threatens to further marginalize radiologists as a member of the patient care team. To combat or reverse this trend, this research suggests that radiologists must find new opportunities to cultivate professional trust and form personal relationships with clinicians and patients. Radiologists must capitalize on existing opportunities to interact with their colleagues as well as create new opportunities for interaction. Some techniques for becoming more visible or more involved in patient care include embedding reading rooms in clinical areas, becoming more visible through active participation in multidisciplinary conferences, meeting with referring providers to solicit feedback on reports and asking how radiology can help them provide better patient care, and taking on the role of patient advocate, especially regarding the appropriateness of requests for imaging. The primary author is currently engaged in research evaluating which of these techniques are perceived by different communities of referring physicians as most useful, effective, or desirable.

However, this is not just a problem of less communication but the kinds of communication and the ways communication takes place. Key questions for future research include exploring how trust and expertise can be established and maintained using alternate modes of communication and novel technologies, including e-mail, real-time video interactions, and telecommunication. Because PACS is here to stay, future research must concentrate on how radiologists can not only become more “visible” to their clinician colleagues but how to use alternate modes of communication in ways that enhance rather than distance radiologists as members of the patient care team.

Back to Article Outline

References 

  1. American Anthropological Association . What is anthropology?. http://www.aaanet.org/about/WhatisAnthropology.cfm Accessed May 9, 2011
  2. Creswell JW . Research design: qualitative, quantitative, and mixed methods approaches. Thousand Oaks, CA: Sage Publications, Inc; 2009;
  3. Saunders BF . CT suite: the work of diagnosis in the age of noninvasive cutting . Durham, NC: Duke University Press; 2008;
  4. Blume S . Insight and industry: on the dynamics of technological change in medicine . Cambridge, Mass: MIT Press; 1992;
  5. Barley S . Technology as an occasion for structuring: evidence from observation of CT scanners and the social order of radiology departments . Admin Sci Q . 1986;31:78–108
  6. Prasad A . Making images/making bodies: visibilizing and disciplining through magnetic resonance imaging (MRI) . Sci Technol Hum Val . 2005;30:291–316
  7. Burri R . Doing distinctions: boundary work and symbolic capital in radiology . Soc Stud Sci . 2008;38:35–62
  8. Reiner B , Siegel E , Protopapas Z , et al.  Impact of filmless radiology on frequency of clinician consultations with radiologists . AJR Am J Roentgenol . 1999;173:1169–1172
  9. Baker S . PACS and radiology practice: enjoy the benefits but acknowledge the threats . AJR Am J Roentgenol . 1999;173:1173–1174
  10. Krestin G , Miller J , Golding S , et al.  Reinventing radiology in a digital and molecular age . Radiology . 2007;244:633–638
  11. Brodwin P . Biotechnology and culture: bodies, anxieties, ethics . Bloomington: Indiana University Press; 2000;
  12. Lock MM , Young A , Cambrosio A . Living and working with the new medical technologies: intersections of inquiry . Cambridge UK: Cambridge University Press; 2000;
  13. Ong A , Collier SJ . Global assemblages: technologies, politics, and ethics as anthropological problems . Malden, Mass: Blackwell; 2005;
  14. Strathern M . The tyranny of transparency . Br Educ Res J . 2000;46:309–321
  15. Rabinow P . Making PCR: a story of biotechnology . Chicago, Ill: University of Chicago Press; 1996;
  16. Evans R , Collins H . Expertise: from attribute to attribution and back again . In:  Hackett EJ ,  Amsterdamska O ,  Lynch M , et al. editor. The handbook of science and technology studies . Cambridge, Mass: MIT Press; 2008;p. 608–630
  17. Carr ES . Enactments of expertise . Annu Rev Anthropol . 2010;39:17–32
  18. Hogle LF . Introduction: jurisdictions of authority and expertise in science and medicine . Med Anthropol . 2002;21:231–246
  19. Shapin S . The scientific life: a moral history of a late modern vocation . Chicago, Ill: University of Chicago Press; 2008;
  20. Beard RL . Trust and memory: organizational strategies, institutional conditions and trust negotiations in specialty clinics for Alzheimer's disease . Cult Med Psychiatry . 2008;32:11–30

PII: S1546-1440(11)00387-5

doi:10.1016/j.jacr.2011.07.006

Journal of the American College of Radiology
Volume 9, Issue 1 , Pages 64-68, January 2012

Article Tools