Guy Harris, D.O.
President, Digital Medical Communications
Tokyo, Japan
harris@dmed.com

Imagine a wealthy country with advanced technological and communications infrastructure and a fully realized industrial base. Imagine now that that same country has a pathologist-to-physician ratio less than one-half the average of similarly developed countries–a country with a manifestly greater apparent need for telepathology. The country in question is Japan, but don’t be too surprised to learn that telepathology has not taken off in a big way. As this report will show, while some of the reasons are specific to Japan, others relate directly to the field itself, and thus to why its progress has been slow everywhere.

First, some background. The Japan Pathology Association currently certifies about 1,600 diagnostic pathologists. According to the Ministry of Health and Welfare, in 1986 there were 5.5 pathologists per 100,000 people in the USA but only 1.0/100,000 in Japan. The ratio of pathologists to all physicians is 2.6% in the USA, 2.0% in England, 1.3% in West Germany, and 1.4% in France, but only 0.7% in Japan.

Moreover, distribution is uneven . In 1992, The Ministry reports 40 pathologists per 100 hospitals in the Tokyo area but only 14 per 100 in Tohoku, the large area covering the north of Honshu, the main island. In Miyagi Prefecture, the center of Tohoku, all 25 pathologists are concentrated in either Sendai, the largest city, or its commuter towns of Shiogama and Natori. No other hospitals in the prefecture have a full-time pathologist, regardless of size. Elsewhere, in the northern part of Kyoto Prefecture, the number of pathologists servicing a population of 350,000 people is exactly 0.

As to an important reason for the shortage, note that remuneration in pathology under Japan’s national health scheme is much lower than for clinical specialties. The Ministry does not even consider pathology a clinical specialty. Moreover, in the absence of direct patient contact, pathologists miss out on a number of ‘perks’ their clinical cousins enjoy.

So if any country would seem ready for telepathology, it should be Japan. Yet growth is slow, and prospects not too clear.

Prof. Takashi Sawai is Japan’s most prominent proponent. He started his career in rheumatoid pathology at Tohoku University (TU) in 1974, and is now Chair of Pathology at Iwate Medical University Medical School, Tohoku’s largest teaching center. In 1992, TU’s Prof. M. Kyogoku (department chair at that time), selected two areas of prospective interest for the department to specialize in: molecular biology and telepathology.

For the latter, the department established links with one, and eventually two, regional hospitals and commenced a program of intraoperative diagnosis. In the two years prior to October 1996, the department (by then headed by Prof. Sawai) provided intraoperative diagnoses in 191 cancer cases, including 49 breast cases (26%), 38 lymphatic (20%), 22 thyroid (16%) and others. Endpoints were the malignancy of cases and the extent of excision required. Seventy-two (38%) were identified as benign and 119 (62%) as malignant. Prof. Sawai documented considerable savings in time and cost. More importantly, in a significant number of cases the need for follow-up surgery was avoided. These findings were sufficient to convince him of the usefulness of telepathology.

Given this apparent success, why has telepathology not developed further? Prof. Sawai identifies four reasons. The first is hardware. In TU’s case, the hub used a Nikon Telmics system while the spokes had Olympus Olmicos. The two systems did not connect, forcing TU to also install the Olympus equipment. Prof. Sawai feels that makers now support standards, and that in any case Internet 2 may obviate the problem through the universality of fast TCP/IP.

The second problem is cost. The Nikon system’s sending unit costs €15 million (approx. US$130,000) and the viewing station €7 million (US$60,000), but there is no reimbursement under the present scheme for the ‘tele’ part - pathologists receive approximately €26,000 ($230) per intraoperative diagnosis–insufficient to cover equipment costs and depreciation. Prof. Sawai’s informal reckoning suggests that break-even would require at least €45,000/consult, but there is no sign the Ministry is prepared to approve such an amount

The third problem is physician acceptance. Prof. Sawai reports that older pathologists wish to retain control over the microscope. This is less true of younger pathologists, who are accustomed to working with PCs.

The fourth and most important problem Prof. Sawai identifies is trust between the sending and receiving ends.

Overall, while acknowledging the problems, Prof. Sawai remains a proponent of telepathology and believes it likely that in time it will be incorporated into general pathological practice.

A more ambivalent view is provided by Prof. Kiyoshi Mukai, head of the First Department of Pathology at Tokyo Medical College. After 10 years post-graduate work in the U.S., eventually rising to Associate Professor of Laboratory Medicine and Pathology at the University of Minneapolis, Prof. Mukai’s experience with telepathology began in 1993 during his time at Tokyo’s National Cancer Center under Dr. Yukio Shimosato. His work in telepathology continues today. He is now constructing a consultation and education network among Tokyo Medical College’s four campuses, and as part of a research grant will soon study accuracy in 120 serial intraoperative diagnoses. The equipment is now out for bid, and will operate by TCP/IP over dedicated 1.5 Mbps lines within a TCP/IP-mediated WAN, using locally developed software operating over Unix. The units will use 21-inch 1800 x 1440 monitors.

Despite this activity, though, he says he has never been interested in real-time telepathology. At the start of our interview he stated quite emphatically, "Telepathology is not the solution for shortages of pathologists in Japan." Telepathology, he says, is valuable for consultation and education, but is too inefficient for clinical use in regional hospitals. He quotes data provided by Dr. Yasunori Tsuchihashi at Kyoto Prefectural University of Medicine, who found that the average time to diagnose a frozen section was more than 30 minutes for telepathology (using 64 Kbps ISDN) compared to 10-15 minutes for standard evaluation. Later he softens his approach, saying telepathology may be the solution in country areas–but personally he is not interested. It may be important to note a significant difference in the clinical situations encountered by the two pathologists. Unlike Prof. Sawai, who has been spent his whole career in an area of pathologist undersupply, Prof. Mukai does not face the problem directly and views it only in the abstract.

Further, like many other pathologists, Dr. Mukai questions the diagnostic accuracy provided by telepathology. "It is easy to say something is there, but difficult to say something is not there." Largely it is a question of trust in who is preparing the slide and selecting the views. In the U.S., he would accept a slide from a pathologist assistant but not from a histotechnologist, while in Japan he would not accept a slide from a technologist or surgeon he did not know (many surgeons in Japan do their own frozen sections). Such questions of trust in the transmitting end are of concern to telepathologists everywhere, but experience has shown them to be largely dispelled as familiarity is gained over time.

Prof. Mukai is candid about the profession’s less rational objections to diagnostic accuracy, averring that domestic studies that would prove or dispel this concern have simply not been done. In the absence of data, he describes the opposition as "superstition," and quotes a Japanese catchphrase, kuwazugirai, renderable as "I haven’t tasted it but already I hate it." He states fear of job loss as one motive. As if this weren’t enough, he says that any pathologist involved in telepathology is by that reason alone an object of suspicion among colleagues, possibly unhinged or even downright off the deep end. Why would anyone want to mess with the tried and true?

A third view is provided by Ms. Yukako Yagi, a Japanese researcher in telepathology now based in the U.S. She started her career without formal training, developing microscopic applications at Nikon’s Optical Instrumental Division in Tokyo. Her work in telepathology started in 1987 and intensified two years later when Nikon quickened the pace of development at client request. While still with Nikon, she worked with Japan’s National Cancer Center from 1990 to 1995, where she was involved in supporting the NCC’s links with the Armed Institute of Pathology’s (AFIP) program. She then moved to Georgetown University for two years and the International Consortium for Internet Telepathology (ICIT), and is now at the Division of Pathology Informatics at the University of Pittsburgh Medical Center. Among other roles, she was at the National Kidney Foundation’s Cyber Nephrology Center for a year, and is Japanese Liaison Officer for the American Telemedicine Association. She is thus uniquely qualified to discuss the position of telepathology in the two countries.

In general terms, Ms. Yagi considers her biggest challenge is to put telepathology into routine clinical work. As to why telepathology has not taken off in the same way that teleradiology has, she states that whereas radiology images are easily converted to (or are natively in) digital format, without the need for much additional work or equipment, the same process in pathology is expensive. Radiology images are sent in their entirety, or in predictably compressed form. These factors affect the quality and quantity of information provided and are largely determined at the time the image is taken, i.e., before ‘tele’ transmission. In contrast, final diagnostic views in pathology need to be selected (or ‘edited’) and are affected not only by factors at the time the sample is taken (analogous to the taking of the radiology image) but subsequently at the time of transmission, namely in the fields selected within the slide and the depth of focus within each 3-dimensional field. Moreover, human influences render the standardization of image quality difficult in pathology. With current technology, it is not possible to have the same information at a distance as is obtained by a glass slide. Telepathology cannot provide the same information.

In other words, whereas the ‘tele’ part of teleradiology can be virtually invisible to the process, particularly for S&F modalities, it is decisive in telepathology. This difference powerfully explains the relatively retarded state of telepathology versus teleradiology, and suggests that, despite superficial similarities, it is perhaps invalid even to compare the two.

As to contrasting the Japanese and US situations, Ms. Yagi says that the needs behind telepathology are different. In Japan, the primary justification is to make up for the lack of pathologists. Thus, the purpose of Japanese telepathology is the support of hospitals without pathologists. In the US, in contrast, there are sufficient pathologists, but they still need to consult with specialists. Thus, the primary functions of US telepathology are to support isolated pathologists and consultations between general pathologists and specialists.

Given the field’s relatively greater development in the U.S., does this difference suggest that telepathology will only succeed where a sufficient level of pathology infrastructure (i.e., human resources) is already in place? If so, what are the implications for telepathology in developing countries?

Ms. Yagi says that development in the U.S. depends on telepathology’s integration into Laboratory Information Systems (LIS). The question is whether this integration is so valuable that it justifies spending a lot of money to set it up. UPMC has for some time been referring images to its LIS for documentation, and is using the images for other purposes as well. Thus, starting telepathology has not been difficult compared with other hospitals. In Japan, in contrast, Prof. Mukai reports that Dr. Toshiyuki Kondo, president of SRL, the country’s largest lab services provider, has abandoned his company’s commercialization of telepathology, saying a business case cannot be made.

Perhaps Ms. Yagi has the final word, saying "People should seek out and focus on the advantages of telepathology, not the disadvantages. Also, various telepathology systems existing in the world - we need to understand what we can do with each of them."