Abstract
The teleoncology practice based at the University of Kansas Medical Center (KUMC) in Kansas City, Kansas, is one of the longest running practices of its kind worldwide. The practice began in 1995 and connected an oncologist at KUMC with a rural medical center in Hays, Kansas. Fifteen years later, the practice continues to thrive at Hays Medical Center and has also expanded to include two additional sites within the state—the Northeast Kansas Center for Health and Wellness in Horton and Goodland Regional Medical Center in Goodland—that offer regularly scheduled teleoncology clinics. While the KUMC practice has witnessed an expansion in service sites throughout its history, the practice has seen a significant decrease in the costs associated with providing such services since its inception. The cost decrease can, in part, be attributed to an increase in the number of teleoncology visits conducted through the practice since it began. In Fiscal Year 1995 (FY 1995), 103 teleoncology visits resulted in a cost per visit of $812. Five years later, the FY 2000 $410 per visit cost for 121 visits was almost half the cost identified in the initial cost analysis. The FY 2003 cost per visit for 219 visits saw another decrease to $401, and the most recent FY 2005 cost analysis yielded another decrease to $251 per visit for 235 visits. The data reported below are likely to be the best now available to track time trends in the cost of providing telemedicine or telehealth consultations. The Conclusion and Policy Recommendations at the end of this article will focus on both the cost–time profile and some other challenges and lessons learned.
Introduction
Since 1992, the Center for TeleMedicine and TeleHealth (KUCTT) at the University of Kansas Medical Center (KUMC) has provided telehealth services to medically underserved Kansans. What began as a single interactive televideo (ITV) connection spanning 300 miles between Kansas City and Hays, Kansas, for pediatric cardiology has become a telehealth network of 66 interoperable facilities over 35 different clinical specialties.
Some of the most clinically underserved and financially underprivileged areas of Kansas receive access to and enhanced quality healthcare through this network. Hospitals, mental health facilities, community clinics, and schools are participating providers. Demand to provide additional telehealth service continues, and, as in many states, responsibility for balancing healthcare quality with access and costs has gained the attention of the Governor and state legislators in Kansas.
The focus of attention in this article is the cost experience of the KUMC teleoncology project, which since 1995 connects a KUMC oncologist with Hays Medical Center (HMC) in Hays, Kansas. The time-span of a decade and a half and the continuity of this project, which has maintained the same KUMC oncologist over this period, allow us to capture more carefully service costs than is usual in telemedicine practices. This cost experience, as reported below, appears to be replicated in other Kansas practices, where there have been fewer opportunities to monitor costs and service delivery. Additionally, we suspect that the cost experience reported herein will be similar to that in other parts of the United States, particularly in rural underserved areas, and of interest to population health practitioners.
The teleoncology project began in response to a lack of cancer specialists in rural Kansas and the recognition that outpatients were being subjected to elevated risks of poor outcomes, considerable out-of-pocket expenses, and inconvenience through travel to distant treatment sites. The purpose of the telemedicine practice has not been to entirely replace traditional in-person consultation; rather, it offers underserved populations the opportunity to acquire quality medical care in a convenient and timely manner close to home. In this particular instance, the practice of teleoncology in Hays, Kansas, has mixed in-person visits with virtual visits with the KUMC oncologist. This mode of combined practice has been well received by residents of rural Kansas, 1,2 and most cancer patients at Hays select this combined mode of care.
Over the >15 years of teleoncology practice in Hays, the demographics of the area and Kansas have changed little. 3,4 Except for urban concentrations in eastern Kansas (the Kansas City metropolitan area) and in south-central Kansas around Wichita, the state has a relatively stable population of about 2.8 million with population densities decreasing sharply as one travels from east to west. About 58 of 105 counties in Kansas are medically underserved. 5 An implication of these demographics for medicine is that patients often reside up to 300 miles from medical specialists.
The data reported below are likely to be the best now available to track time trends in the cost of providing telemedicine or telehealth consultations. The Conclusion and Policy Recommendations at the end of this article will focus on both the cost–time profile and some other lessons that we have learned.
Background and Policy
Teleoncology is the evaluation and treatment of cancer patients via ITV systems and supportive electronic equipment. The practice between the KUMC and HMC combines an in-person traditional outreach clinic at which a KUMC oncologist flies to HMC for in-person visits and a telemedicine clinic whereby the same oncologist provides care via ITV visits. That is, patients commonly receive both traditional and virtual consultations. This practice pattern, population, and patient demographics have remained essentially the same for more than a decade; however, as will be seen below, the number of patients who utilize this service has grown.
About half (52%) of the patients receiving care through this practice are female. The ages of the patients range from 20 to 86 years (mean of 66 years). Sixty-two percent of the patients seen through the KUMC-HMC practice reported a total household income of <$40,000 per year in 2006. Over the time that data were collected for this report (our most recent cost analysis) all of the patients receiving care through the KUMC-HMC practice were Caucasian. The demographics of western Kansas are fairly represented in this report.
In its early years, KUMC used Integrated Services Digital Network (ISDN) lines to transmit voice and video data over existing telephone wires to HMC from KUMC. Utilizing ISDN technology, the university-based oncologist was able to evaluate patients in Hays and treat a variety of oncologic and hematologic disorders. Peripheral devices such as an electronic stethoscope and a document stand for X-ray reviews allowed him to effectively conduct a clinical visit without being in close proximity to the patient. The patient's history was obtained in the same manner as it would during an in-person clinic visit; however, the physical distance imposed between the physician and patient during a teleoncology visit necessitated that a nurse be present alongside the patient to serve as a proxy examiner for the physician.
While the protocol for conducting teleoncology visits has not changed much since the 1995 inception of the KUMC-HMC practice, the technology has. Today, the oncologist uses an Internet-based videoconferencing system for clinical evaluation and treatment of patients in Hays. The real-time technology used allows patients to receive specialty healthcare more comparable to the care they would receive in an in-person setting. As appealing as it may be to reduce patient travel expenses, risks, and discomfort, it is also essential to know the costs associated with providing care via telemedicine.
When the KUMC-HMC teleoncology practice first began, it was quite costly to conduct clinical evaluations via telemedicine. 6 However, data collected from cost analyses over the past 12 years indicate that the costs of a teleoncology visit between KUMC and HMC have fallen significantly. Current data also suggest that, in terms of financial viability, the teleoncology practice is now at a self-sustaining point; third-party payers are reimbursing for this service, and HMC has been willing and able to contract with KUMC for consultations.
Data and Methods
Since 1995, four cost analyses have been conducted within the KUMC-HMC teleoncology practice. 6 –8 The first analysis looked at cost data from Fiscal Year 1995 (FY 1995), and the most recent analysis examined FY 2005 cost data; FY 2000 and FY 2003 cost data also have been analyzed and reported. The goal of each analysis was to document the provider costs associated with the practice by monitoring expenses for oncology services rendered via telemedicine. These analyses, therefore, have been primarily cost accounting studies, since the data required to undertake cost–effectiveness, cost–benefit, or cost–utility analyses have been too sparse. 9
One of the biggest challenges in executing these cost analyses was the actual collection of the data consistently at the two study sites. 9 –11 Working with individuals at both geographic ends of the practice and determining the correct contact persons for identification of systematic and specific expenses was often difficult. Additionally, while key staff remained the same across the decade of data collection, many of those assisting during the first cost analysis in FY 1995 were no longer available during the most recent analysis due to position or responsibility changes.
Reconciling hospital accounting data can be challenging and time consuming even at a single point in time much less over time. 10 Given that data were collected in a naturalistic setting where the clinical practices are located—as opposed to an environment specifically designed for the conduct of research—it was not possible to control for all personnel and staffing changes that occurred during the four data collection periods spread over more than a decade. Nevertheless, the lack of turn-over in key staff facilitated the collection of consistent data at the two practice sites.
The director and assistant director of the KUMC telemedicine department, the KUMC oncology services office manager, the HMC administrative director of oncology services, the HMC telemedicine coordinator and office manager, and the HMC teleoncology nurses have all contributed to the collection of cost data. 6,8 Cost-tracking and accounting methods used in the initial FY 1995 analysis have been used in each successive analysis, 6 and the same health economist was available for the entire study period.
Costs of operating the teleoncology practice were examined each year by and at the telemedicine clinic and the Hays outreach clinic. When expenses were incurred jointly by the two services, costs were allocated based upon the percentage of the total number of patient visits conducted via telemedicine and the percentage of the total number of patient visits conducted in the outreach clinic. Teleoncology practice costs assumed by both medical centers—KUMC and HMC—were recorded separately for each cost analysis. Costs incurred included technology-related expenses, such as telemedicine room rent, telemedicine equipment, telecommunication line charges, and or practice-related expenses, which encompassed such costs as salaries, billing and transcription expenses, telephone and fax machine, and miscellaneous supplies.
The telemedicine room costs were calculated by multiplying the size of the room in square feet by an average rental figure for commercial office space in the area at both KUMC and HMC. Telemedicine equipment and maintenance costs varied depending on whether new equipment was purchased or updates were made to the equipment during the fiscal year in which the analysis was completed. Additional information in the Appendix describes expenses for all four cost analyses; changes in practice associated with changes in accounting costs are also mentioned in the Appendix.
During FY 2005, HMC expenses associated with traditional medical visits on site were collected. Amounts reported included such expenses as personnel salaries, billing expenses, fax and phone expenses, transcription costs, supplies, clinic room and equipment expenses, and malpractice insurance. These expense data were obtained for any visit at HMC, including patients seen by any of the local oncologists or midlevel practitioners at HMC as well as patients seen by the KUMC oncologist during his fly in outreach clinics. Additionally, the KUMC oncologist has been on contract for clinical services for the duration of the teleoncology practice with HMC. HMC also pays a small fee annually to help cover expenses related to the teleoncology practice.
Study Results
We report costs of the telemedicine practice below in terms of the dollar costs incurred in the year under examination. Since costs of telemedicine equipment and services have decreased over time as quality has improved, our figures are conservative estimates of real cost reductions over the decade. We have made no attempts to quality-adjust the data for changes in cost relative to the quality of equipment, since estimating inflation and quality improvements for technology is a daunting task 12,13 and would not contribute much to this article.
Our FY 1995 analysis (the FY at both study sites runs from July 1 through June 30) revealed an average cost per teleoncology visit of $812. 6 Five years later, the FY 2000 analysis indicated that the cost of a teleoncology visit had been reduced by almost 50% to $410. 11 While the cost of a teleoncology visit remained steady over the next 3 years, our most recent analysis demonstrated another substantial decrease to $251 per patient visit in FY 2005, as shown in Table 1.
Teleoncology Practice Costs
The subcontract for physician services is included as an HMC, not a KUMC, expense, since payments were made to KUMC to cover costs of this service.
FY 1995, Fiscal Year 1995; HMC, Hays Medical Center; KUMC, University of Kansas Medical Center.
The four cost analyses of the teleoncology practice clearly illustrate that while the number of visits conducted has more than doubled since 1995, the cost per patient visit has decreased by almost 70% in that time.
Given that the number of teleoncology visits increased to 235 during FY 2005 and the average cost per visit decreased to $251, the question arises as to whether a teleoncology visit is cost competitive with a traditional in-person visit. We have found that the average cost of a traditional clinical oncology visit at HMC was $332 in FY 2005, as shown in Table 2; this cost figure was calculated based upon data obtained for any visit at HMC. Therefore, the $332 figure includes costs associated with patients seen by any of the local oncologists in Hays, by the KUMC oncologist when he flew in for outreach clinics, and by the midlevel practitioners at HMC.
Hays Medical Center Traditional Practice Cost per Visit: Fiscal Year 2005
USD, U.S. dollars.
Conclusion and Policy Recommendations
The most notable conclusion to be drawn from our four cost analyses is that the cost of providing cancer care via telemedicine has decreased substantially since the practice first began in 1995. This cost reduction arises from decreasing costs of technology and the increasing number of patient visits for which this technology is used. Learning-by-doing also has effected cost reductions through sensitive and more realistic adaptations of the KUMC-HMC physician service contract. Contract payments have come to more accurately project the numbers of consultations actually provided, patients seen, and real costs incurred by KUMC. Contracts are made with KUMC for a year before the actual delivery of services.
Since patient choice is involved (in-person versus telemedicine), the fact that more patients have become aware that a telemedicine offering is available along with their traditional consults, the number of telemedicine consultations has increased. As in other published studies, we have found that patients using telemedicine express high levels of satisfaction with the service and continue to use it as an alternative to more traditional service delivery. 1,2 Further, we believe that telemedicine can offer an excellent study environment for “testing” and careful assessment of patient–physician interactions and how clinical judgments are made. 14
In a span of only 10 years, the cost of a teleoncology visit fell from $812 to $251. These costs are average costs. While it would be desirable to know the marginal costs of providing this service to determine an optimal level of service provision, 15 data are unlikely to become available for such calculations soon due to both the relatively small numbers of services provided through telemedicine and the limited number of service sites.
As Table 1 shows, the cost of conducting a teleoncology visit decreases as the number of visits increase. In an earlier peer reviewed article, we reported the first quantitative cost curves for a telemedicine service. While these were only pseudo-cost curves, in that the data were insufficient to estimate well-behaved quadratic cost curves, these curves allowed us to estimate both average and marginal costs for different levels of service provision. We found that the telemedicine practice then under examination needed to attain, maintain, or exceed a threshold level of 200 visits per year to provide service at the same cost as that of a traditional clinic. 16 The similarity between the number of visits and costs of our earlier finding and this report is highly suggestive. Telemedicine may be cost competitive with in-person consults when telemedicine consultations exceed 200 annually. Each of the cost analyses we have done demonstrates that the costs of providing services via teleoncology are closely tied to levels of utilization with steeply declining marginal cost curves. 8
In our FY 2003 and FY 2005 analyses, we were able to estimate additional (or ancillary) revenue streams entering HMC as a consequence of the KUMC-HMC “combined” practice. These additional revenue streams make the telemedicine practice even more attractive to HMC and, most likely, would to other rural providers. Estimated revenue from this combined practice flowing into HMC included billings for physician evaluation and management and hospital billings for services such as pharmacy; intravenous therapy; laboratory medical and surgical supplies and devices; diagnostic and therapeutic radiology; nuclear medicine; computed tomography scans and other imaging operating room; blood storage and processing; magnetic resonance image tests; chemotherapy drugs; electrocardiogram tests; and additional diagnostic and therapeutic services. Billings for these ancillary activities were $1,051,026 in 2003 and $809,451 in 2005. In both years, around 50% of billed charges were collected by HMC.
While other studies have examined patient and provider perceptions of teleoncology services, 17 –20 data from our four cost analyses offer a longitudinal perspective on how the average costs of a rural teleoncology service have evolved through the years—and might well evolve in other settings. Further work is required to better analyze costs, outcomes, and benefits of telemedicine services. We believe that a mixed modality of service can be acceptable to patients and providers and can improve quality, contain costs, and reduce the risks to which patients might be subjected through lengthy commutes for all their healthcare services.
In the future, we suggest that more attention be given to the costs and efficacy of care delivered to patients using a combined traditional-telemedicine mode of practice. The practice mode described herein arose over time through negotiations among providers, patients, and practitioners. Initially, rural residents were reluctant (indeed some were vocal and antagonistic) about substituting telemedicine for a real doctor. Yet, few of the communities served by the KUCTT can attract and retain medical specialists. A combined mode of practice may be a possible answer to service delivery problems, needs, and the high costs of specialty healthcare services (when they can be obtained at all) in rural areas. Cost and efficacy studies are needed to suggest optimal balances between modes of service, total costs of care, and efficacy to better serve patients in medically underserved areas.
Footnotes
Disclosure Statement
No competing financial interests exist.
Appendix
The expenses accounted for in Table 1 included the following: 1. Telemedicine room space at University of Kansas Medical Center (KUMC) and Hays Medical Center (HMC) estimated at the cost of office space at the study sites allocated on a percentage use basis by teleoncology patients. 2. Amortized costs of all telemedicine equipment and line costs. 3. Equipment maintenance costs. 4. Actual expense of use of technicians to handle problems associated with equipment use including training other staff. 5. Salaries for administrative support apportioned by use of teleoncology patients as a proportion of total patients seen. 6. Scheduler expense (salary-based) apportioned by teleoncology patient usage. 7. Fax, telephone, copying, and supplies based upon actual cost. 8. Office space based upon estimated cost of office space at KUMC and HMC, respectively, allocated on a percentage use basis by teleoncology patients.
All expenses accounted for in Table 2 come directly from the accounting systems at HMC. No adjustments were made in Tables 1 or 2 for constant dollars; therefore, the figures provided are incurred and allocated expenses in the year noted.
“HMC practice-related expenses” in Table 1 include a contract negotiated cooperatively between HMC and KUMC for physician services. Variations in the amounts reported reflect salary and fringe at KUMC, forecasted (or anticipated) levels of service, and some other technical support. All expenses reported under the physician contract are physician services provided physically at KUMC and received via telemedicine by the patient at HMC. This contract is recorded as a cost to HMC. This avoids double counting of expenses incurred and ties these expenses to the site at which services are physically received by patients. Any services provided by KUMC that are not covered explicitly under the “HMC practice-related expenses” physician contract are reported in Table 1 as “KUMC practice-related expenses.”
Increased costs between 2000 and 2003 in Table 1 reflect an anticipated and actual expansion of services, which included acquisition of some new equipment, increased technician time, and increased contracted costs for physician services. Substantial reductions in technician time and some disposable equipment and supply costs occurred in 2005. “HMC practice-related expenses” may be underestimates of true costs since the physician contract was negotiated anticipating the hiring of an oncologist at HMC; however, an oncologist was not hired at HMC for a number of reasons. This means that HMC patients may have received physician services through this contract that are not fully accounted. This, however, is difficult to determine since the number of consults may correlate modestly with the actual use of physician time. If the physician contract was increased to the value of 2003, the cost per consult would be about $281.