Digging its way in: lab digital
  imaging

May 2005
Feature Story

You don’t know it yet, but if you’re skeptical about digital imaging or even admit to being a technophobe, there’s a nearly 50 percent chance you’ll have a change of heart in the next three or four years. At least, that’s one plausible inference from CAP TODAY’s recent survey of anatomic pathologists.

Even some who now regard digital imaging as a gimmick or a useless frill in pathology could shift their thinking. That’s because the technology is well beyond being the "next new thing" and on its way to being firmly embedded in the laboratory, the survey suggests.

In answer to the survey’s lead question—Does your group currently capture one or more images of the histologic sections associated with surgical pathology cases...?—34 percent of the pathologists who responded said yes, their group captures such images to store or integrate into reports.

Of those saying no, 12 percent said they would consider deploying an image capture system within a year, and 35 percent estimated within one to three years. "A picture montage greatly enhances our reports," was a typical response.

On the other side, 50 percent insisted there was no chance. In answer to the question of when they would adopt such a system, said one: "Never, if I can help it."

Nevertheless, that still suggests digital imaging is conforming to a classic technology adoption curve—the formula that has accurately mapped the adoption of tools like VCRs, cell phones, and DVDs, says Bruce Friedman, MD, professor of pathology at the University of Michigan Medical School and director of pathology data systems for the University of Michigan Health System.

"The general rule," he says, "is about one-third are early adopters, and they’re willing to tolerate the inconvenience and technical challenges of the technology. The questionnaire results reflected this." (This would correlate, he notes, with the technology enthusiasts, visionaries, and some of the pragmatists as defined by Geoffrey Moore in Crossing the Chasm.)

"What it means," Dr. Friedman continues, "is we’ve got ten through the easy part, and I’d predict about a third will hang back for a brief period and wait until the technology has proven itself, and a third are uncomfortable with change in general and will be very reluctant to adopt this new technology. The way these adoption curves move, it’s probably taken six or seven years to get the first third of people to adopt, and my guess it will be two or three years for the next third to come around."

Dr. Friedman told CAP TODAY he found the questionnaire results, which ranged all over the map, to be a fascinating collage. "As I was looking at the survey write-up, I thought of the ad line, ’Certs is a candy mint. No, Certs is a breath mint.’

"People in pathology look at digital capture, storage, and retrieval in different ways and to accomplish different goals. Some people look at it and say, ’Is this a means to enhance the quality of my small biopsy or tumor reports?’ Some look at it and say it’s great for recording gross images of specimens.

"Others say it’s great for teaching conferences, and others say it allows me to acquire and publish more information about a case, understanding that an image is just a different type of information. Different people come up with different reasons why they want it."

And why they don’t. The question, "Why haven’t you adopted it yet?" brought a flood of negative descriptors, among them "time-consuming," "nonreimbursable," "superfluous," "inefficient," "far too cumbersome for our surgical pathology volumes," and "not clinically useful." "Worthless," was one pathologist’s succinct appraisal.

Dr. Friedman was struck by the volume and variety of these narrative comments and said the questions in the survey "clearly struck into an emotional vein." One commenter dismissed digital imaging as "window dressing," while another termed it a "relatively useless ’bell and whistle’ for the report."

"I would be hard-pressed to think of another technology adopted by one-third of pathologists that another third would regard in such a dismissive way."

"I don’t really blame many of the pathologists who question the value," Dr. Friedman says, "because the technology for digital image capture and publication has been very user-unfriendly in recent years, and it disrupted the workflow in surgical pathology reading rooms."

He finds it interesting that in radiology there’s almost no debate about the value of capturing images and making them available to the ordering physician. "You have to ask the question, If it’s the norm in an allied field like radiology, why not in pathology?"

"Part of the reason is surgical images do not start as digital. A paraffin section looked at through a microscope is essentially an analog image, the opposite of digital, so several steps are necessary to convert it to digital."

In contrast to radiology, pathology also has a sampling issue, he admits, because a paraffin block and a paraffin section contain millions of images. "But I believe it is really important in diagnostically challenging cases to present in the report those key images on which the diagnosis was based," he says. "When reviewing such a report years later, it’s a relatively simple matter to reconstruct the decision process of the pathologist signing the report."

Lack of reimbursement should not be an issue, in his view. "A number of pathologists commented on the survey, ’Well I don’t get reimbursed for image capture and publication.’ But radiologists don’t get reimbursed either for that specific component of the consultation process. It’s part of the process they accept, and many clinicians place a high value on being able to review the radiology images themselves. And I believe the same is true in surgical pathology."

And he says dismissing digital imaging in surgical pathology as "marketing" isn’t wise: "It’s a weakness of our specialty that we don’t always convey to clinicians the value of our services. Then we complain we’re not adequately appreciated by them."

"My own view," Dr. Friedman adds, "is that a report with one or more images, microscopic or gross, is a higher-quality report than one that does not have images for complex and challenging cases. And I feel in a relatively short period, the quality of reports will be judged in part on the basis of whether relevant images are included."

Still, he knows that some pathologists’ doubts about digital imaging are not ill-founded. "Capital is scarce for most pathology departments and health care in general, and there are a tremendous number of financial reasons for reluctance to adopt new technology, particularly one such as imaging without a very favorable return on investment.

"If a pathologist is not extremely enthusiastic submitting a capital request, he may assume, a priori, that hospital administrators will turn it down. And it is also very easy for administrators to turn down early technology, particularly if they don’t see additional revenue." For the most part, imaging systems are an additional expense and not a revenue generator, so they have to be justified on the basis of clinician satisfaction with a higher-quality report, he says.

Technical issues too can set up roadblocks, the survey revealed. One pathologist reported that problems with image integration into Cerner AP reporting had prevented the adoption of digital imaging, while for another hospital inadequate computer storage capability of the LIS and hospital information system was an obstacle.

On the other hand, just plain embarrassment could be an incentive to make the leap into digital imaging. One respondent who now uses digital image capture only for teaching said his hospital was considering expanding its use.

"Since we don’t incorporate gross and microscopic photos in our reports," this pathologist said, "the clinicians always point this out when we get reports from some small city in Mexico that does have gross and microscopic images in its reports (most of good quality and in color)."

Thirty-five percent of pathologists using digital image capturing said in the survey that integrating images into reports has been successful and cost-effective. But even more (42 percent) cited success in using images for conferences or teaching. Fourteen percent listed documentation of gross images.

The respondents’ descriptions of how they are using the technology showed a shift from recent years, Dr. Friedman says. Small biopsies and tumor cases were cited often, as was gross imaging, while skin cases and cytopathology were cited less frequently.

The popularity of skin biopsies as a favorite use for digital imaging has dropped as a percentage of total use, he says. "Images of skin biopsies launched the technology, and now they’re being supplanted in popularity by small biopsies and use of images for gross tumors."

Cytopathology also ranks at the bottom in terms of popularity. "This reflects, I believe, a lag on the part of cytopathologists in appreciating the value and importance of images in the report, but I look for that to start climbing."

It’s no surprise, he says, that the greatest use is for tumor cases. "Gross images of tumors are very, very important in selected cases. And one of the nice things about capturing a digital gross image is that the digital image can then be used to indicate where tissue has been sampled as a substitute for the crude line drawing on a piece of paper." Such paper drawings, traveling with the case, he notes, can be a source of contamination.

As one of the respondents pointed out, there is tremendous potential for correlations between positive cytopathology cases and subsequent definitive surgical procedures, Dr. Friedman says. For example, with a history of a positive Pap smear, "when you do, say, a uterine resection, you could then correlate the image of the previous positive Pap smear with an image of the uterine tumor."

But the value of images goes beyond direct patient care. From a clinical research perspective, he says, "the presence of clinically relevant pathology images in databases would be extremely valuable when cases are retrieved and analyzed."

And finally there’s the value of pathology images for patient education—the opportunity for the clinician to sit down with the patient and have a discussion about the results of surgery.

"The patient can’t recognize the tumor," Dr. Friedman says, "but the image might help the clinician in explaining, for example, that there’s a tumor in the resection margin or in a lymph node; therefore, we have to treat you with chemotherapy or radiation therapy."

Asked how they deployed their system for capturing images, 49 percent of the respondents said they had an internally developed system. Only eight percent said the system came from their current laboratory information system vendor, while about a third (35 percent) obtained a system from a different vendor.

"One would think the LIS vendors, since they had already installed the system in a department, should have been on top of this emerging technology and made it available as an easily used feature in the LIS," Dr. Friedman says.

Why isn’t that the case? "Perhaps the LIS vendors didn’t understand the technology, or it didn’t integrate well with their technology, so they didn’t offer it." Or, he speculates, "They didn’t see there was sufficient demand from customers for it, or perhaps the technology is so challenging that specialized vendors had to evolve to develop it as best-of-breed stand-alone systems."

Most laboratory information systems have been managed by clinical pathologists because such a large percentage of information generated in the laboratories comes from chemistry and hematology. "It may be that LIS vendors did not see sufficient demand for imaging modules because they were interacting primarily with clinical pathologists."

In teaching and conferencing, digital imaging has enthusiastic converts. "We do use digital image capture extensively to teach and conference," wrote one pathologist, who added, "I don’t know how we got along without it in past years. I recently had a biopsy case that came out on a Tuesday and I integrated that same case in a lecture I gave on Wednesday at 8:00 AM."

Walter H. Henricks, MD, director of laboratory information services and surgical pathologist at the Cleveland Clinic Foundation, told CAP TODAY his hospital, too, is poised to have pathologists use digital imaging to create their own collections, and most specifically for residency, medical school education, and tumor board purposes.

Using primarily Diagnostic Instruments’ Spot Insight camera, "we’ve phased in digital photography for the microscope over the last one to two years," he reports.

Dr. Henricks says it’s important to distinguish digital imaging’s educational purposes (teaching, conferences, or tumor boards), quality assurance, and personal educational collections from putting images in reports.

"To me, that’s a whole separate realm of behavior, goals, usage, and results compared with putting images in reports. A lot of people wouldn’t argue with the value of using digital imaging in a big center where you can load images on a network, show them in conferences, and use them for teaching. There are all kinds of obvious benefits."

"But when you start talking about putting images into reports, it’s very controversial."

Yes, it is. One pathologist surveyed wrote, "Pictures are not representative; they are misleading for most nonpathologists, and can be misinterpreted by others." Another respondent who sized up digital imaging as "marketing fluff" commented: "Who is going to look at it—a surgeon who couldn’t tell myocardium from onion root cells? Try signing out 90 to 100 complex cases a day. Leave the picture taking to Michael Moore."

Dr. Henricks, who is a founding member and former president of the Association for Pathology Informatics, sympathizes with some pathologists’ concerns.

"The pro side is that this adds value and helps some clinicians communicate with patients. It serves as good patient education. Referring physicians like to see photos in reports, it jazzes up reports, and the commercial reference laboratories have this capability."

"The flip side is, What purposes does it really serve? I think groups have to ask. When you really look at what it does for patient care, I’m not convinced of its value."

As one survey respondent wrote, "We’ll probably end up doing it because it’s progressive, but the medical justification for the extra time, expense, and overhead remains to be proven."

Moreover, Dr. Henricks says, some pathologists are concerned that photos in reports can be misleading. "We’re dealing with field selection, and you can’t always encompass an entire case in one or two microscopic fields."

"I’ve seen referred cases of very nice reports with very beautiful photos, but the diagnosis was incorrect. The most important thing is the accuracy of diagnosis and staging information if there’s a tumor."

Dr. Henricks is a skeptic on how likely clinicians are to use images in talking with patients: "I find it hard to believe that most clinicians will sit down to explain a picture they themselves may not understand. Also, it kind of diminishes us to say we can sum up the work we do in pathology in a single photograph."

Installing digital imaging to survive competitively, however, is understandable. "It might be required for survival for some practices," Dr. Henricks says, if the group’s client base is at risk of migrating to other laboratories that provide this capability or if it wishes to compete for new business. His own laboratory plans to implement the technology to have it available for competitive reasons.

When asked why they had implemented digital imaging, the largest number of respondents said because of "interest in the technology." But that’s a vague answer. The second largest answer was more specific: pressure from clients, Dr. Henricks points out. "Competition from from reference laboratories might be one of the single biggest drivers for groups to put images in reports." Only 17 percent cited a desire to capture more detailed information about cases.

"So they’re looking at what competitors are offering, the sales and marketing departments perceive competitive advantages, and that’s what’s driving it. None of the answers really say it’s going to improve patient care."

Dr. Henricks sees himself as "a bit of a contrarian" on some digital imaging issues. "I’m not saying we should go backwards, but we should decide whether the investment of time and money in digital imaging in reports is justified. And I think there will be more valuable applications of digital imaging in the future, such as whole-slide digitization," he says.

"Now that LIS vendors are becoming more progressive, you can buy systems to integrate images without any need for an external interface. It could be that some respondents built their system before this was possible."

Those using digital imaging only for tumor boards, conferences, and presentations are not going to use the LIS to capture images, he adds. "They will be putting together their own cameras and databases. And that’s something else to consider as we migrate toward electronic medical records."

Telepathology is a separate issue from the digital imaging cited in the survey, he notes. There’s a definite value to telepathology, which is diagnosis at a distance using some type of imaging technology in real-time or in consultation.

But he and Dr. Friedman suggest digital imaging is likely to have a much larger impact. "My own belief," Dr. Friedman says, "is that for the most part, except for some specialized markets, telepathology is not a big thing."

"For rural areas that are understaffed, yes, and for large hospitals or medical centers there is some movement of images around the institution, but most pathologists who need to can walk to the specimen and see it their way. So the market for telepathology will always be a niche—but the market for diagnostic image capture, storage, and retrieval will be important."

Pathology information systems are now frequently sending reports through interfaces to other systems, especially in the hospital. "But those interfaces typically don’t handle images at this time," Dr. Henricks notes.

"It’s not that medical record systems can’t handle them, but its means of transfer might not prove feasible for images. In the future, though, I think we will be seeing pathology images in the EMR as well."

He is not referring to a universal standard electronic medical record, but to individual institutions’ systems that were formerly called hospital information systems but are now more full-featured and include medical orders, laboratory orders, and the results from those orders.

"They’re the mechanism through which a lot of laboratory pathology reports are transmitted to physicians, and physicians often access them through some type of electronic system."

To facilitate that access, it’s important to find ways to standardize image exchange, he says. "Sharing of digital imaging is viewed by many as a significant part of our future, and a first piece is to develop specifications for a digital imaging exchange system so images can be transferred regardless of how they were obtained." This standardization is one of the key objectives of the Association for Pathology Informatics (www.pathologyinformatics.org) and its Laboratory Digital Imaging Project.

Survey respondents were likely to report that response from clients has been positive. About 70 percent said the system they had adopted was either very acceptable or acceptable. Said one: "Clinicians love it, and they are now so used to it they demand it."

However, there is no unanimity on the question of what clinicians prefer. "Clinicians are not asking for images on reports; they want timely reports," one pathologist wrote. Another said the impersonality of relying on digital imaging is a drawback: "One-on-one interactions have been preferred at our academic center."

In Dr. Henricks’ view, there is some validity to what the holdouts are saying. "Sometimes people do digital imaging just for the sake of doing it. It’s something that has some value and is worth considering, but departments have to look at what are the reasons, what are the benefits versus the costs, and are they justified."

For example, as one pathologist explained, there is frequently just not enough time to shoot photographs. "It does impact pathologists’ workflow to stop and select fields and take pictures," Dr. Henricks agrees.

Don’t be too quick to dismiss those who are hesitant to adopt digital imaging, he advises. The survey’s respondents shared thoughtful reservations about why it may not be right to rush headlong into the technology.

Nevertheless, he sees strong signs that resistance to digital imaging is softening. "Years ago, people just didn’t want to adapt. But we’re now well beyond the stage where people are afraid to use it."

Anne Paxton is a writer in Seattle.