The true cost of a teleradiology system involves much more than just the initial capital outlay for a system and its ongoing maintenance costs. In fact, facilities implementing teleradiology systems often find that acquisition and maintenance costs are dwarfed by other expenses associated with operating a teleradiology link.
The hourly cost of a radiologist's time is crucial in determining the actual payback of a teleradiology system. There are two ways of examining teleradiology-related staff expenses -- the cost of a radiologist's time and the net value of that time. Based on published statistics, the average radiologist in the U.S. made just over $250,000 in 1998. Using a workweek of 50 billable hours, and factoring in eight weeks of annual vacation, this amounts to just under $115 per hour.
A typical "circuit rider" (a radiologist who travels between sites to read films) spends between 1 and 1 1/2 hours per day in travel time alone, or 330 hours a year in transit. Thus, a facility without a teleradiology system spends about $37,500 annually per radiologist in wages lost to non-productive travel, or $112,500 over a typical three-year return on investment (ROI) period. Under these circumstances a teleradiology system would pay for itself in three years, even at a facility with just one circuit-rider radiologist. When one factors in multiple radiologists who travel for a group, the financial returns of a teleradiology sstem multiply dramatically.
One can also look at the value of a radiologist's time in terms of earning potential lost while traveling. If we take the same 330 hours of annual travel time used in the example above, and multiply this amount by a radiologist's net earning potential of $500 per hour (which is very conservative by most estimates), it amounts to $165,000 per year in earnings lost to travel. That's nearly $500,000 over the three-year ROI period. Eliminating this expense would pay for a small-scale PACS network for most smaller hospitals, while the financial returns from installing a less expensive teleradiology system would be even greater.
Evaluating a system 
Looking just at the ROI model above might lead one to believe that any teleradiology (or even PACS) system would suffice for any application. This is untrue. A facility must evaluate several factors in order to determine whether a teleradiology system is called for or justified, including:
- Procedure mix;
- Payor mix; and
- Available wide-area network (WAN) infrastructure.
It is important to note that although teleradiology systems can transmit static images for review, dynamic modalities such as fluoroscopy, and interventional procedures such as biopsies, all require a radiologist on site. While radiologists might be able to perform these procedures remotely in the near future, the cost of doing so may exceed the value of using the technology for these procedures. In this case it is also important to remove technology from the equation entirely, and consider the human impact of replacing a "live person" with technology for certain applications.
There are two different types of teleradiology in use today. Preliminary diagnosis (or "on-call" radiology) is often used by radiologists at night to make a preliminary review of a study, usually a CT or ultrasound exam, followed up the next day with a more thorough review. Some plain-film review is also done, although the costs associated with reviewing plain film often exceed the billable charges for interpretation, unless the facility has a very high patient volume.
The number of preliminary-diagnosis teleradiology systems being sold today has also declined significantly in favor of remote primary diagnosis (RPD) systems. RPD allows a radiologist to make a primary interpretation from transmitted images, eliminating the need to review filmed images the following day. The vast majority of systems sold today allow RPD to occur.
Technology requirements
From a technology standpoint, the requirements of RPD and preliminary diagnosis systems are fairly similar. RPD teleradiology systems generally use DICOM interfaces to scanner modalities, although not always. One of the considerations here is that the modality must provide native DICOM output of imaging data. If a modality does not provide native DICOM output, then a third-party interface device must be used. This increases the cost of a teleradiology system significantly, as most DICOM converters cost between $17,000 and $22,000 per modality. Digital frame grabbers offer a cost-effective alternative, although image post-processing is limited to the pre-windowed eight-bit data displayed on the modality's console. Still, unless the radiologists do multiple bone and tissue windows, this latter method may suffice for most teleradiology applications.
Another consideration is the cost of the film digitizer. With laser digitizers averaging in the low $40,000 range, many groups have chosen to purchase digitizers based on CCD (charge-coupled device) technology, especially where film volumes are relatively low but are still a part of the mix of images that must be read. CCD-based digitizers are typically priced at about half the cost of laser digitizers, and recent advances have brought the performance of these systems close to that of laser digitizers. Laser digitizers typically add between $35 to $40 per day to the overall cost of a teleradiology site.
Network connection costs
Equipment costs are a significant investment, but certainly not the only cost to consider. Other costs help determine payback as well. Wide-area network (WAN) costs are recurring and play an increasingly important role in the overall operating cost of a teleradiology site. A radiology group should determine both the mix and volume of images from each site before determining the most cost-effective WAN to use. These factors also determine what level of data compression (if any) can or should be used.
If a group is receiving images from just one site, then even a moderately slow WAN such as an ISDN (integrated services digital network) or even a dial-up telephone line might be employed, provided the level of compression used isn't excessive. However, if a fairly high image volume is being transmitted, then frame relay with a robust CIR (committed information rate) of 256 KB or higher, or even a dedicated T-1 line, should be considered. Again, cost is a factor, as dial-up lines are available for as little as $20 a month, while T-1 lines can cost $1,000 or more per month, depending on distances between the sites.
Overall WAN use also plays a role in determining the cost of the connection. If a T-1 line can be segmented so that a portion of the bandwidth is used for voice and a portion for data (called a fractional T-1), and a site can eliminate long-distance charges, then the higher cost can be justified more easily. If, on the other hand, the connection is used only for teleradiology applications, then the image volume must be significant in order to justify the cost of a T-1.
The use of data compression is highly controversial, yet in most cases is required to improve the transfer speed of image data. Various levels of data compression used can be used, ranging from 2.5:1 (lossless) up to 25:1 (lossy), depending on image type, quantity of images, compression algorithm used, and amount of grayscale in the images. Most RPD teleradiology systems let users define the compression ratios, with most selecting a medium scale (6:1) for digital images and higher (upwards to 20:1) for general radiographic images.
Service and maintenance costs
Service and maintenance costs also play a role in the cost of a system. Most RPD applications mimic PACS networks in that they require a service contract due to the system's complexity (which may even include an interface to the results-reporting system). This typically adds about 10% per year to the system's purchase price. Most low-end teleradiology systems can be maintained on a time-and-materials basis, again depending on the complexity of the system's design.
System down time is another critical issue. In general, the more complex a teleradiology system, the higher the chance a component will fail. Each day of down time can have a catastrophic effect on the system's cost/revenue ratio. A mid-priced teleradiology system in the $100,000 range costs about $150 a day to operate over a three-year amortization period. The same system can easily generate $500 to $600 a day in gross revenue. This translates into a loss of about $400 for every day the system is down, or approximately $2,000 per week.
Not only is revenue loss critical, but the impact on the quality of service a group provides to its referring physicians is significant as well. Therefore, ensuring that a teleradiology vendor can provide prompt local service is the main priority in evaluating teleradiology and PAC systems.
Break-even periods
The break-even period and fee structure for teleradiology is determined by all of the factors discussed in this report. Imaging facilities considering a teleradiology purchase should also look at the competitive environment in which they operate, and evaluate the amount of time a radiologist really needs to spend on-site. As mentioned previously, teleradiology cannot supplant the need for an on-site radiologist, nor can a group use teleradiology without the benefit of an on-site radiologist who visits at least weekly. At best, proper application of the technology can significantly reduce the number of hours a radiologist spends on-site, and eliminate non-productive travel time while providing a higher level of service to referring hospitals.
Are hospitals willing to pay more for the higher service level teleradiology can provide? Most will not, considering the excess capacity of radiologists in the U.S. This overcapacity is estimated to reach as high as 40% by 2000. The most that hospitals are willing to pay per exam is the prevailing reimbursement rate, which can easily provide an ROI of over $100,000 per year per site for even a moderate-sized (75-bed) hospital. This estimate is predicated on that facility's generating only three CT exams, three ultrasound scans, and a dozen general radiographic exams per day.
Next to the cost of converting non-DICOM images into DICOM-compliant data sets, the largest cost of a teleradiology system surrounds the interpretation of plain films. These procedures also have the lowest reimbursement rates, averaging $12 to $15 per study. In contrast, a CT study typically generates between $125 and $150 in professional fees, with ultrasound scans generating $75 on average. Again, these fees vary by geographic area. Most cost-justification models use a return-on-investment period of three years, or a maximum of five years, even though the teleradiology system might be used for a longer period.
Saving money with system design
There are many ways to save money in designing a teleradiology system. The first is to evaluate the requirement for DICOM data sets. If the modalities being used are native DICOM, then certainly an imaging site should consider using DICOM data sets.
If the scanners are not native DICOM (as is the case in more than 80% of all small hospitals and several imaging centers), video frame-grabber boards might be a more logical and cost-effective solution. Though theoretically most frame-grabbers can handle up to four inputs, the practical maximum number of connections to a signal frame-grabber board is two, or at the most three. This may dictate a second image capture station. While the cost of the second station may equal one non-DICOM to DICOM conversion device, this device can handle multiple inputs and again help defray costs.
CCD-based film digitizers can save $20,000 to $25,000 in a typical system configuration (a figure that is half the cost of a laser digitizer) and provide image quality and resolution comparable to that of lasers. Ongoing operating and maintenance costs are also a fraction of those incurred with laser digitizers.
Using data compression can significantly cut down on networking requirements. This method allows a dial-up line, ISDN, or a moderate-sized (128 to 256 KB) frame-relay cloud to supplant the need for more expensive T-1 or other higher-priced line. There are some pricing tradeoffs, however, as the cost for a dedicated compression server can add anywhere from $5,000 to $10,000 to the overall system price. Still, when amortized, the cost is less than that of a T-1 line, which can easily exceed $1,500 per month depending on distance and the local infrastructure.
In areas where they are available, ADSL (asynchronous digital subscriber lines, otherwise known as DSL-Light) and cable modems offer exceptionally high performance-to-price ratios, averaging $50 to $100 per month for 1 to 10 MB/second throughputs. If a T-1 is required, consider using it for other applications as well, including voice and/or dedicated data sharing (centralized shared HIS, RIS, or transcription use between facilities, for example) to help defray the cost.
Conclusion
Teleradiology can be a significant moneymaker for a radiology group if the system is properly designed. By asking the right questions, purchasers of teleradiology systems can hone in on the right mix of price, performance, and ROI, and meet their goals of using teleradiology to improve efficiency and maximize profits.
About the author
Mr. Cannavo provides consulting services on the acquisition and assessment of PACS and teleradiology systems. He can be reached at:
Michael J. Cannavo
Image Management Consultants
A division of Healthcare Imaging Specialists
1545 Warrington Court
Winter Springs, FL 32708
USA
Phone: 407-359-6575
Fax: 407-359-6577
[email protected]
