With the latest advances in technology allowing radiologists to tele-transmit images in virtual real time to any point on the globe, it may be only a matter of time before the interpretation of imaging is treated as a commodity to be sold transnationally for reading and interpretation on the open market. Many radiologists deny this could ever happen. But others are not so sure.

Whether globalization of image interpretation occurs or not, it is clear from interviews with physicians deeply involved in teleradiology that the transmissibility of images already has altered forever the practice of the radiologist’s art. And the rules of the teleradiologic game are far from resolved. For now, the ethics are defined by the players, as teleradiologists compete for the health care dollar but are wary of stepping on their colleagues’ toes. Wilson Wong, MD, an ultrasound and body imaging specialist, is the founder of Teleradiology Diagnostic Services (TDS), Arcadia, Calif, which supplies night-reading services to radiology groups throughout the state. “Our company is built on friendship,” Wong says. “I would never walk in and say, ‘We’ll do all your daytime reads.’ It is a matter of ethics. The rule is, you try to respect everybody’s territory.”

Curtis Harlow, MD, is a rural radiologist who offers on-call, after-hours coverage to small hospitals in his mountainous region of Colorado. Harlow and his partners conduct their own practice during the day, then take turns handling teleradiology on call at night. “I’m a free market kind of guy,” Harlow says. “I think all is fair in business. I would not welcome our imaging being sent out of town, but there is no ethical reason why referral sources should not do that. But I am going to win that battle. If I am physically here, I’m going to provide better service.”

Jeffrey A. Landman, MD, is a founder of INPHACT, an already large and growing company headquartered in Nashville, Tenn, that supplies not only reading but a whole gamut of teleradiology and picture archiving and communication system (PACS) services to hospitals and radiology groups. “There are no rules out there today,” Landman says. “I would never go to a hospital and say, ‘We can supply all this and you do not need your radiologist.’ But I will compete for hospitals that cannot afford a radiologist, or where the hospital does not want the radiologist it has. In that case, the hospital and the medical staff would have to be on the same page and supporting us. If that were to occur, I would think about it. That is where the ethics should be.”

Ethics is not the only touchy issue in teleradiology these days. State licensure is another. So is emergency department (ED) coverage. So is equipment standardization. But the truth appears to be that for all the demonizing of teleradiology, it has not yet turned into the monster that many feared it would. Instead, it has shown itself to be a huge help in the provision of radiology services, lessening the impact of the shortage of radiologists, reducing radiologists’ job stress at a time when the demand for imaging is on the rise, and prompting the creation of an electronic infrastructure that will carry over into all of telemedicine. Radiologists have been and still are the telemedical pioneers. To see just how useful teleradiology is, it is instructive to look at some of the many configurations to which this new technology is being adapted.

Night readers

Wong says 65% of radiologists in a recent California survey rated being on call as the most stressful part of their lives. Wong’s TDS has found a niche for itself by relieving that stress. Founded in 1997, the company now delivers night-reading services to 40 hospitals in California. Having begun TDS with only six hospitals on board, Wong’s company has grown quickly. Maybe too quickly. “I don’t like to grow,” he says now. The five radiologists on staff have all the work they can handle, and recruiting radiologists to work nights is difficult because of a shortage within the profession.

The bulk of TDS’s work is reading from 6 pm to 7 am for EDs. Wong says the company can fax a written preliminary report to the referring hospital within 15-18 minutes of receiving the image. All images are transmitted electronically. TDS contracts for services with the radiology groups at the various hospitals whose EDs it serves. TDS does preliminary reports only. Its clients do their own primary overreads in the morning. TDS keeps a log of any disagreements between its preliminary reports and the overread versions.

Wong says most of his clients purchased electronic equipment for teleradiology before he met them. Therefore, a big part of TDS’s effort has been to design software or purchase hardware to interface with the many systems in use by its clients. “The software that I write for getting the report back to the client is what I take pride in,” says Wong, a computer expert.

Consulting with ED physicians over the phone is constant, despite the faxed written preliminaries, Wong says. “I see our network as an alliance with clients and the ED doctors. We’re on a first name basis,” he says. Wong adds that most ED doctors can read radiographs but not the more sophisticated modalities like CT, MRI, and ultrasound. A disturbing trend to him has been that the ED doctors are ordering more and more imaging, routinely turning his night readers into diagnosticians. “We are losing the art of medicine,” he says. “The ED doctors are relying on us instead of using their own judgment. It’s disturbing. I think it’s a problem nationwide.” It is a problem exacerbated by patients who have learned that going to EDs is one way to skirt the HMO gatekeepers. So much so that Wong estimates the ED setting now accounts for 10% to 15% of all health care. “In the old days nobody would get a bone scan at 3 am,” Wong says. “Now they are asking us to do that and it is costing everyone a lot of time and money.” In essence, ED radiology is becoming a subspecialty, he adds. Like most of the physicians interviewed for this story, Wong declines to give exact figures on the numbers of reads his company does each day or week. He estimates that 40% of TDS’s work is CT of the head. Pelvic and abdominal ultrasound accounts for another 20%, and pelvic and abdominal CT another 20%. The remaining 20% is composed of an assortment of plain film studies.

Rural teleradiology

At Red Mountain Radiology in Canon City, Colo, Curtis Harlow, MD, and a partner service St Thomas More Hospital. They also service a clinic in the nearby mountain town of Westcliffe. A third radiologist, not a partner, is contracted to serve Gunnison Valley Hospital, in the town of Gunnison, over the mountains about 120 miles north of Canon City. At night the three practitioners use teleradiology to do on-call reads. The hospitals are connected by old-fashioned analog phone lines to the doctors’ homes, and the night on-call duty is rotated among the radiologists, all of whom also work daytime.

“Our ED is very, very busy,” says Harlow, noting that Red Mountain handles about 3.5 after-hours cases per night on average. Each case results in a written report sent back to the referral source by fax. Additionally, each case results in about two phone calls for consultation, Harlow says. The night reads are preliminary and the primaries are done the next day.

Recently, through a grant from a Catholic charity, St Thomas More was able to install a $100,000 scanner to serve the 16,000 residents of Canon City as well as the people of Gunnison and Westcliffe, so that the teleradiology, though serving a population that is rural enough to be “fairly out there,” as Harlow puts it, is quite up to date except for the phone connections. “I think it has definitely increased the business we do at night,” says Harlow, noting that many of the ED patients are borderline clinic cases. “People can’t get a doctor’s appointment or they can’t find a doctor.”

Teleradiology Toolkit

There are two important considerations when designing a teleradiology system. First, one must assess what type of electronic roadway is needed to carry the images from the imaging machine to the interpreter and then from the film interpreter back to the referral source. The dimensions of this electronic roadway will depend on the scope of the work that must be done. For preliminary reads, for example, a night on-call radiologist in a rural setting, the system can be a simple electronic path costing perhaps less than $50,000 to implement. On the other hand, if remote primary diagnoses (RPDs) are to be required, the roadway will not be so simple, and the costs will not be low, more like $250,000 or higher.

Michael J. Cannavo, president of an Orlando, Fla-based consulting company, is primarily a PACS (picture archiving and communications system) consultant, but his work requires expertise in teleradiology.

Cannavo emphasizes that before ordering any teleradiology devices, practitioners must consider the second important factor in the decision-making process, which is how much the system will cost or save the practitioners in terms of their own time. Practitioners must also analyze any system to see how it might be integrated with other office functions-like voice transmissions as part of an expensive T1 transmission line-so that the highest value is obtained from any hardware or software put in place. “It’s so much more than the equipment,” Cannavo says.

Entry-level teleradiology

With that caveat in mind, Cannavo advises that for entry-level teleradiology involving preliminary reads only and utilizing existing phone lines, the expense would involve purchasing a “frame-grabber” (a simple image capturing station) costing about $20,000, and a single-screen review workstation, which could be an off-the-shelf PC and associated software, located in the radiologist’s home. The PC might cost $8,000 or so and the software another $2,500. So the most basic system would be “between $30,000 and $40,000” to implement, plus the costs for the use of phone lines and so forth, Cannavo says.

Still talking simple, if the basic system were to be augmented by the addition of a low-end data compression device to speed up transmissions and a low-end film digitizer to allow use of faster phone or cable system lines, then the costs would escalate by another $20,000 or so.

“If you want to do a primary diagnosis, those systems can easily run $250,000 and up,” Cannavo says. “Ideally, you’d want to store your data, which calls for a PACS.” Since primary diagnosis requires comparing old patient studies with the latest study, the viewer/reader would need a fairly sophisticated high-resolution viewing screen-and there would have to be dual screens to compare old studies to new. Cannavo estimates this cost at $40,000 to $60,000. Digitizers and data compression devices, along with high-speed digital transmission lines, would be required. Special T1 transmission lines can cost $1,000 to $1,500 or more in monthly fees, Cannavo says.

Additionally, the digitizers, whether laser or HD-CCD (high definition charge coupled device) versions, would require vendor service contracts. To become DICOM (Digital Imaging and Communications in Medicine) compliant, which nearly all teleradiological operations of any size are doing now, would require a DICOM file server with compression, adding $30,000 or so to the setup expenses.

Cannavo envisions a low-end RPD system as costing somewhere in the neighborhood of $110,000. Slightly upgraded versions might cost in the $150,000 range, and more sophisticated versions could double that price.

Radiology groups and hospitals interested in installing a teleradiology system are well advised to patiently analyze the costs of installation and operation in terms of the time to be saved in the radiologist’s billable hours. A teleradiology system is not worth doing if it is a consistent financial drain. Consider the standards of the American College of Radiology or the FDA, according to Cannavo and other experts interviewed for this story.
-George Wiley

Harlow says teleradiology has been for his group much more boon than demon. He does not think the technology will ever replace on-site radiologists.

A large-scale service

INPHACT founder Jeffrey Landman, who is now chairman and chief medical officer, says that the company was the first application services provider for medical imaging, and the first to market Internet-based radiology services to hospitals and radiology clinics. INPHACT could be called the plug-in computer of teleradiology. For an installation fee or under a long-term rental contract, INPHACT will spend up to $1 million installing all the equipment a facility needs to become an up-to-date teleradiology services provider. INPHACT’s services include a PACS, a radiology information system (RIS), as well as digitizers and compression equipment to move images electronically. INPHACT also contracts with radiologists or uses its own staff of more than a dozen physicians to provide reading services, if desired. Transcription is also available. The only thing INPHACT does not do is install the phone lines into and out of a facility.

Landman, who now has more than 40 hospitals of varying size signed on, says taking a 200-bed hospital teleradiologic can be done for around $1 million. INPHACT does its homework before completing any contract and generally requires a 5-year contract. The hospital or clinic client pays a set fee for each study completed. INPHACT does no billing for services, instead assigning billing rights to its clients and then requiring that each client pay each month the agreed-upon fee for any service package. “I’m giving facilities access to data and putting the capital expense on my budget,” Landman says.

According to Landman, INPHACT already has the capability through its existing customer base to impact more than one-third of the estimated 350 million radiological procedures performed each year in the United States. “We’re reading now around 30,000 cases per month. Three years ago, we were probably reading around 2,000.” INPHACT just signed up with the Veterans Health Administration (VHA). “Its 1,000 to 1,500 hospitals will be able to use whoever they want for teleradiology,” Landman says. “The VHA won’t force compliance, but the hospitals will have to bring us to the table. Basically, for us, it is a hunting license.”

While INPHACT has grown quickly it remains under pressure to lock in its market segment. To do that, it has just received more than $17 million in capitalization from a Canadian investment company, Landman says. The reason INPHACT must secure its position is because the company has been so successful that it is attracting competitors with bigger names and deeper pockets than its own, Landman says warily.

Subspecialty reading

At the Cleveland Clinic Foundation in Cleveland, Michael P. Recht, MD, section head for outside imaging, presides over 60 to 70 radiologists who do subspecialty reading only. They read for the Cleveland Clinic’s chain of a dozen hospitals and for hospitals and clinics across the country. “We’re not reading any plain films at this point,” Recht says. “We only do images that are DICOM [Digital Imaging and Communications in Medicine].”

Recht says his clinic began marketing its services to others only after it had put together a complex teleradiology system for its own needs and then realized that its services could be marketed and delivered electronically. The foundation interprets subspecialty imaging for clients all over the country where hospitals and clinics lack the subspecialists in house. “We look at MRI scans while the patient is still on the table,” Recht says. “Everything is in real time. We don’t do anything that is not in real time. Everything we do is a primary read. We don’t do overreads. People don’t send something to us out of the blue and say, ‘Will you read this?’ It’s not a trivial task. They must connect to our RIS. We read thousands of cases per year. The geography changes daily.”

Consults are done over the phone, but the foundation is laying the groundwork to switch to consults over the Internet. “We’re not trying to replace local radiologists,” Recht says. “If there are subspecialty radiologists already on site, then we don’t have a role to play in that community.” He adds, “We think we’re providing a good service. We think it’s a win-win for everybody. The ethics book is being developed, but it hasn’t been written yet. We certainly want to keep radiology among radiologists.”

Clinical trials

Mark A. Goldberg, MD, is a radiologist, but he has not practiced for nearly 10 years. Instead, Goldberg, a former president of the American Telemedicine Association, has found a niche for himself as president for advanced technology in the Informatics Group at PAREXEL International Corp, headquartered in Waltham, Mass. PAREXEL, which has more than 1,000 employees, is a contract research organization (CRO) that specializes in clinical trials for pharmaceutical and medical device companies.

Goldberg runs the radiology segment of the CRO. It is a very important segment since researchers have learned they can use imaging to much more quickly determine the impact of a particular compound under trial than they could if they waited for the patients to demonstrate outwardly a positive or negative response to the drug. Studying changes in tumor size is only one obvious example of how imaging systems can speed up the clinical trials process and get drugs approved and on the market faster than waiting for patients to recover or die from their diseases.

Goldberg says the images coming into his centralized reading rooms arrive in all forms, from plain film to images transmitted over the Web. Once in hand, all images are digitized and electronically studied and stored. Extreme efforts are made to make sure that trials taking place are comparing oranges to oranges and apples to apples. “We’re a leader in the application of imaging,” Goldberg says. “We provide consulting as well. We guide companies and tell them which imaging modality is the best choice. It’s not just imaging, it’s having done imaging in a reproducible and analytical fashion. That increases the statistical power of the trial. The more noise you have, the harder it is to predict the statistical value of the compound.”

For all the attention to the details, Goldberg still calls his radiology segment of the clinical trials “a rabbit hole,” as in Alice in Wonderland. He notes that historically there has always been a 10% to 15% disagreement ratio between panels of radiologists asked to interpret the same images. When these disagreements occur between the researchers actually conducting the trials first hand in the field and the readers back in Waltham, they must be resolved, Goldberg says. Most often, if patient treatment decisions must be made, it is the primary researcher in the field who prevails, Goldberg adds. But he notes that the centralized reading also plays its part in decreasing variability and that central readers can sometimes see things that the lone observer in the field cannot. “That’s why it’s a rabbit hole,” he says.

PAREXEL is not required to supply fast reads back to referring physicians or to meet state licensing regulations for patient care, but it has been able to do already what the some in radiology may be waiting to do. Its clinical trials take place in many different environments from the United States to Europe to East Africa to Asia, Goldberg says. “We’re worldwide.” As such, PAREXEL is pioneering the Web-based imaging conduits that in the not too distant future will serve all of radiology. “Moving toward the Internet becomes the choice of efficacy,” Goldberg says.


Turning back to the issues confronting teleradiology, it appears that with the exception of subspecialty consults ordered by referring physicians in exceptional cases only, almost all teleradiologists are being extra careful to adhere to state licensing requirements. TDS operates only in California, where its physicians are licensed. INPHACT contracts with licensed radiologists or obtains licenses for its readers who serve various states. So does the Cleveland Clinic Foundation. “We’re open to working anywhere in the country,” says Cleveland Clinic’s Recht, “but we will make sure we have state licensing. Each physician that reads will hold licenses in the states that they read for, even though each license could cost $200 or more.”

Jim Borgstede, MD, is a Colorado-based practitioner who has served for the past several years as chairman of the American College of Radiology (ACR) Commission on Small and/or Rural Practices. Borgstede says the ACR “endorses full state licensure.” But some states, he adds, have no licensure requirements when it comes to reading and interpreting electronically transmitted images. “Some states like Hawaii say you can all come through tele-radiology. Some states have no laws at all.”


More and more, teleradiology is being subjected to equipment standardization, and already many in the field are demanding that hospitals and radiology groups have digitizers (see related story at right) in house to encode film for transmission. Sophisticated hardware systems are being linked directly to MR, CT, and ultrasound scanners to feed digitized images to readers in distant locations. JPEG (Joint Photographic Experts Group) compression is another must-have capacity for any serious user of teleradiology. The equipment list can get quite sophisticated and expensive, as has already been seen when hospital information system and RIS databanks are electronically connected. The equipment situation varies greatly depending on the physical locale of the teleradiology user. Many rural practitioners are still locked into using analog phone lines. DICOM compliance is becoming the order of the day for those in all but the most electronically disadvantaged locations.


What this all boils down to is that soon teleradiologists will be working with standardized equipment allowing them to send and receive digital images, and quickly the Internet will become the transmission vehicle of choice for teleradiologists sending patient-protected imagery from one point to another. When that technology is routinely in place, there will be no barriers to the transnational interpretation of images, and the reading of films could become a commodity that HMOs and other groups, ever intent on cutting costs, will ship off to reading rooms in Bangladesh or Bombay or whatever reading site develops as the low-cost preference for the radiological users of the future. “The ‘Indian Scenario’ could establish itself,” says the ACR’s Borgstede, “in the states that don’t have licensing requirements. The potential is there for the HMOs to put pressure on readers, and if they could get services cheaper, they might want to do that. That is a theoretical possibility, but I can’t state that I know it’s occurring.”

That probably has not occurred yet, but is it inevitable? There already are signs that globalized companies are eyeing the reading and interpreting market, according to Landman. INPHACT recently put out a press release announcing a purchase and supply agreement with a new vendor for computed radiography and PACS equipment. What makes the signing of this agreement interesting is that INPHACT had previously bought much of this type of equipment from a different vendor, one that still owns 3% of the company, Landman explains. He says that the former supplier has decided to enter the market as a competitor. “Until now, we haven’t had any competition, but now we do, and that validates what we’ve been doing,” Landman adds. “[The vendor] will be our only real competitor. They say they’re going to be in the business.”

If big global companies are getting involved with the marketing of teleradiology systems that include reading and interpretation, just what are the implications for radiology? “Going forward, I think a lot of radiology will be a commodity,” Landman says. “It will be driven to where the capacity is. The negotiating parameters will be the fees in the future.”

Others, however, would not concur. Harlow, in fact, believes patients and referring physicians will prevent radiology from being wrested from the hands of local radiologists. “They could replace us with a box, but the referring doctors don’t want that and the patients don’t want that,” he believes. “They want a real person there. They want somebody to consult. They don’t want just a computer. I think this is a universal.”

George Wiley is a contributing writer for Decisions in Axis Imaging News.