The Policy Context of Telemedicine
In their early days, most telemedicine programs had relatively low profiles politically. Clinical applications generally did not cross state borders, or if they did, they involved federal government agencies that were not bound by state licensure or liability policies. The programs did not provoke much legal controversy at either the state or federal level, and decisionmakers, evaluators, and advocates did not appear immediately concerned with possible jurisdictional problems (Shinn, 1975).
Today, even though interstate telemedicine is not necessarily a high priority for many users or potential users, jurisdictional issues relating to licensure and medical liability are generating considerable debate and anxiety. Privacy and confidentiality have emerged as significant policy issues as computer-based patient information systems and databases have proliferated. Public and private policies regarding payment for telemedicine services are regarded by many advocates of telemedicine as a major obstacle. Whether and how such policy concerns are resolved can affect both the benefits and the costs of telemedicine and, thus, the sustainability of telemedicine programs.
At the same time that some governmental policies have posed problems for telemedicine, others have been devised specifically to encourage telemedicine. Such policies include demonstration project
funding, technical assistance, research, and telecommunications infrastructure development that supports an array of applications, in particular, distance education.
Public policies, thus, present a mixed picture of incentives and disincentives for telemedicine that those contemplating telemedicine programs have good reason to examine carefully. Evaluators likewise may reasonably consider whether policy-related variables should be factored into evaluation plans. They may, for example, want to
- be alert for changes in a telemedicine program intended to achieve consistency with state or federal policies;
- enlarge the array of benefits, risks, and costs to be assessed;
- extend the search for possible determinants of telemedicine's acceptability, affordability, and even availability to clinicians, patients, and others; and
- investigate a broader range of practitioner or patient concerns about specific policy issues (e.g., privacy and confidentiality) to assess how well issues are understood and how important they are.
As additional context for the evaluation framework presented later in this report, this chapter examines several key policy issues. The first two sections briefly review federal and state initiatives to promote telemedicine either directly or as part of more general efforts to encourage communications and information technologies. The next three sections discuss professional licensure, malpractice, and the privacy and confidentiality of personal medical information. The final two sections consider telemedicine payment policies (a topic that covers both public and private actions) and the regulation of medical devices. Because it was not part of its charge, the committee did not make recommendations about the policy issues discussed here.
National Communications And Information Infrastructure Policy
The most newsworthy recent federal action affecting telemedicine is the Telecommunications Bill of 1996, a broad and far-reaching reform of communications law that is expected to alter dramatically
the telecommunications industry (Andrews, 1996). Among other changes, it allows long-distance telephone companies, cable companies, and other firms to compete with local phone companies, and local phone companies can, under some conditions, compete in the long-distance markets. Most rates for cable television will be deregulated within a period of three years.
Of greatest immediate relevance for telemedicine are provisions of the legislation directed specifically at assuring universal communications services at affordable rates for rural, high-cost, or low-in-come areas. One provision (Section 101 of the bill) states that "a telecommunications carrier shall … provide telecommunications services which are necessary for the provision of health care services in a State, including instruction relating to such services, to any public or nonprofit health care provider that serves persons who reside in rural areas" at rates comparable to those charged in urban areas. Telecommunications providers in rural areas are to be compensated through a fund designed to promote universal access to modern telecommunications services. This provision will be interpreted by regulations developed by the Federal Communications Commission. Other more general provisions of this major legislation may have quite significant implications for telemedicine, but their effects are difficult to predict very precisely at this early point.
Congressional interest in telemedicine is reflected in Senate/House Ad Hoc Steering Committee on Telemedicine, which sponsors brown bag lunches and other events for members and staff interested in telemedicine. The group cosponsored a 1994 conference to develop a consensus agenda for telemedicine and health informatics (Bashshur et al., 1994).
During the 103rd Congress, at least 22 pieces of legislation specifically related to telemedicine were introduced (Telemedicine Monitor, 1995). Most explicitly or implicitly supported telemedicine in one way or another. In the first 15 months of the 104th Congress, telemedicine figured in at least 15 pieces of proposed legislation (Arent Fox, 1996). These proposals would, among other things, fund additional pilot projects, establish a commission on telemedicine, and include some telemedicine applications in health insurance reforms. Despite such positive signs, telemedicine was still targeted for a share of budget cuts as part of House and Senate proposals to eliminate the federal deficit.
More generally, as discussed in Chapter 1, telemedicine is just one element of a developing National Information Infrastructure (NII) that combines telecommunications and information technologies (Lindberg and Humphreys, 1995b). The White House has sponsored a major NII initiative to identify and resolve policy questions and speed the development and dissemination of information and telecommunications technologies not only in the government but also throughout the private sector in manufacturing, commerce, education, health care, and other areas.
An Information Infrastructure Task Force, led by the Department of Commerce, has the lead in coordinating activities related to the NII initiative. It created the Health Information Applications Working Group that, in conjunction with the Department of Health and Human Services, established the Joint Working Group on Telemedicine, chaired by the deputy director of the Office of Rural Health Policy (Puskin et al., 1995).1 As discussed further in Chapter 5, one objective of that group has been to develop a broad evaluation framework for various federally funded demonstration projects and other telemedicine activities. The working group also has several additional areas of activity related to policy issues, safety standards, managed care, and development and maintenance of a telemedicine project inventory.
Another closely related federal enterprise is the High Performance Computing and Communications (HPCC) program (Lindberg and Humphreys, 1995a,b). This initiative emphasizes basic research and advanced technologies and networks. Some projects involve telemedicine and other health applications such as virtual reality tools for guiding or performing surgical procedures, computer-based patient records, and digital imaging software. These applications may raise a variety of regulatory issues, for example, FDA regulation of medical devices for safety and effectiveness.
In addition to these initiatives, several federal agencies have funded various kinds of demonstration projects intended to promote more specific agency agendas such as rural economic development. Appendix A provides information on several of these projects.
State Programs And Initiatives
Although some state policies or proposed policies have been viewed as barriers by those interested in making telemedicine applications less cumbersome, expensive, or risky, other state policies have directly supported telemedicine through legislation, demonstration funding, planning and coordination support, and other means (Lipson and Henderson, 1995; GAO, 1996). State initiatives in behalf of telemedicine have been concentrated in, but not limited to, the Midwest, Great Plains, and South. The objectives, scope, and depth of these initiatives are quite varied as are funding levels and sources. Some have focused primarily on medical education whereas others have emphasized service provision or support. The Western Governors Association (WGA) has expressed particular interest in initiatives to reduce some of the policy barriers to telemedicine described later in this chapter (WGA, 1995).
Evaluators need to be concerned about state initiatives, particularly as they may affect analyses of the cost, cost-effectiveness, convenience, or sustainability of telemedicine. For example, major infrastructure projects may provide a more enduring base for ongoing programs than do one-time grants.
As recently as 1992, only five states had explicit telemedicine planning or development efforts. According to a study undertaken by the Intergovernmental Health Policy Program (IHPP), the number was at least three times that in 1995, and additional states were considering action (Lipson and Henderson, 1995). The IHPP study, which covered 28 states, grouped states into five categories: (1) those with relatively well-developed state programs; (2) those with less well-developed programs; (3) those on the verge of program development; (4) those with some activities but no program; and (5) those with little or no state presence. Table 4.1 presents these categories with illustrative examples of state activities.
Some states surveyed by the IHHP have active telecommunications programs to support rural schools and libraries, but they have not made medical linkages a state policy priority. For example, a recent General Accounting Office (GAO) report profiled three states, two of which, Iowa and North Carolina, have encouraged medical care as well as educational links whereas the third, Nebraska, has emphasized educational links by the University of Nebraska Medical
TABLE 4.1 Categorization of 28 State Telemedicine Programs in 1995 with Examples
Center, which was the site of some of the earliest medical education experiments with telemedicine, as described in Chapter 2 (ORHP, 1993b, and GAO, 1996).
In addition, some states not surveyed for the IHHP study do have some level of state support for telemedicine. For example, the state of Maryland's trauma system includes telemedicine applications, which were discussed with the committee during a site visit. More states can be expected to develop policies on telemedicine in the future.
Professional Licensure And Distance Medicine
Telemedicine challenges the traditional view of professional practice as involving a face-to-face encounter between clinician and patient. This encounter made "the place where medicine was practiced, and who was practicing, … obvious" (Gilbert, 1995b, p. 28). Telemedicine breaks that physical link and thus complicates decisions about where a telemedicine practitioner should be licensed if the practitioner and patient are located in different states.2
Many telemedicine programs are not affected by these complications because they operate entirely within a single state. In addition, telemedicine programs operated by the federal government are not restricted by state licensure laws. For example, clinicians and managers working in the military, the Department of Veterans Affairs, the U.S. Public Health Service, and the federal prison system can proceed with cross-state tests and applications of telemedicine without concern about state challenges or penalties.
In the United States, the responsibility for licensing and otherwise regulating health professionals lies with state governments. States originally adopted licensure laws and objective criteria for entry into designated health professions to protect people from charlatans and untrained individuals holding themselves out as qualified
medical personnel. All states require physicians, nurses, dentists, and certain other health care personnel to be licensed by the state to practice their profession. The penalties for practicing medicine without a license are significant and may include criminal as well as civil penalties.
State laws, which were adopted long before the advent of modern telemedicine, also require any out-of-state physician who diagnoses or treats a patient in the state to be licensed in that state. Most states, however, provide an exception that allows physicians licensed in that state to consult with the licensed physicians from other states (and sometimes other countries). This exception is intended to allow patients to have access to the talents and expertise of physicians from other states without having to travel to those states.
State consultation exceptions are not uniform. Some exceptions are broadly stated, but others limit the exception to one-time or occasional consultations or include other restrictions. Generally, consultation provisions do not appear to offer protection for an out-of-state clinician unless a consultation is requested by or otherwise involves an in-state clinician. A few states, including Louisiana and Pennsylvania, do not have an explicit consultation exception.
Recently, some states have amended or are considering amending their physician licensing statutes (or changing the interpretation of existing statutes) to prohibit out-of-state physicians from practicing without a license in that state (Gilbert, 1995b; Young and Waters, 1995; Richardson, 1996).3 Indiana, Texas, South Dakota, and Nevada have enacted statutes that require out-of-state telemedicine providers to be licensed in those states. The Texas statute states that "a person who is physically located in another jurisdiction but who, through the use of any medium, including an electronic medium, performs an act that is part of a patient care service initiated in this
state … and that would affect the diagnosis or treatment of the patient, is engaged in the practice of medicine in this state for the purposes of this Act" (Texas HB 2669, amending the Texas Medical Practice Act, Article 4495b of the Texas Civil Statutes). The amended statute does not contain a narrow exception for "episodic" consultations with physicians in the same medical specialty.
Other states, such as Florida, have narrowed statutory consultation exceptions by administrative regulation or by interpretation. In a widely cited action in May 1994, the Kansas Board of Healing Arts adopted a regulation requiring that—except as authorized by statute—any person "regardless of location" must be licensed in Kansas if he or she "treats, prescribes, practices, or diagnoses a condition, illness, ailment, etc., of an individual who is located in Kansas" (Kansas Regulation 100-26-1). The Kansas legislature, however, has not amended the statutory consultation exception (Kansas Statute 65-2802), which exempts from licensing out-of-state physicians engaged in the healing arts in consultation with a licensed Kansas physician. Thus, the effect of the Kansas regulation is not entirely clear.
Restrictive state positions have been challenged by those who argue that a telemedicine service involving an out-of-state clinician does not involve the practice of medicine in the state where the patient is located. In this view, the situation is simply equivalent to a medical visit for which the patient travels physically rather than electronically to another state. To date, this perspective has not prevailed.
If a reliance on state consultation provisions does not appear possible or prudent and if a telemedicine practitioner believes it necessary or sensible to seek licensure in other states, the requirements for doing so will vary from state to state. Some states provide for licensure by endorsement, which means they will grant a license to a clinician already licensed by another state with equivalent or stricter requirements. Licensure by endorsement may not, however, be easy. For example, South Carolina requires "a fee; two photos; an application completed in part by the physician's medical school, the state or national board which issued the original certificate by written exam, and if possible a medical society; the signatures of three South Carolina licensed physicians if at all possible; and a personal appearance by the applicant before the medical board" (Granade, 1995a, p. 6).
Why do licensure requirements create problems for clinicians and organizations involved in telemedicine practice across state lines? First, obtaining a professional license involves some expense, including initial and renewal licensing fees. Some states require a face-to-face interview, which entails expenses for travel.
Second, obtaining and maintaining multiple licenses can involve considerable time, research, and paperwork, which may, in turn, involve additional expenses (Collins and Charboneau, 1993). Because requirements for licensure vary from state to state, new analyses have to be undertaken and new forms completed for each state in which a clinician expects to provide telemedicine services. Moreover, requirements for maintaining a license also vary with respect to the time intervals for relicensure, provisions for continuing medical education, and other matters.
Third, licensure typically brings with it a number of obligations with which clinicians must be familiar. Although broadly similar, state laws may vary on specific points, such as confidentiality requirements (Eid, 1995). For example, Colorado and Minnesota "require doctors to disclose to the state the names of HIV-positive patients' sexual partners, [while] New York and California say those names can't be disclosed" (Richardson, 1996, p. A8). California gives individuals the right to review their medical records, but Maryland does not (Eid, 1995). What would be the situation of a Maryland patient with a California telemedicine consultant?4
Many of the current practical burdens of multiple licensure could be eased for individual practitioners if health care institutions assisted with expenses and provided administrative support as costs of doing business. Moreover, the development of national practitioner databases should make access to information about physicians much easier, although concerns about the accuracy and importance of included data have made such databases controversial. Nonetheless, the ultimate responsibility for adhering to licensure requirements resides with the clinician.
Some advocates of telemedicine have suggested national legislation that would either create a national telemedicine license (on grounds that it is a form of interstate commerce) or replace state licensure with a national system (Bashshur et al., 1994; Sanders and Bashshur, 1994; Richardson, 1996). The latter concept, national licensure, has surfaced periodically in response to concerns about physician maldistribution, underserved areas, constraints on professional mobility (especially for dual-career couples), and anticompetitive professional practices. To date, resistance at the state level has been sufficiently strong to limit the progress of such proposals.
Today, with federal authority under attack on many fronts (e.g., Medicaid and welfare policies) and renewed advocacy of states' rights, the prospects for national licensure would appear dim. In the future, however, if telemedicine proves itself despite regulatory obstacles, the constituency for change—especially in the form of a national telemedicine license under the Commerce Clause of the Constitution—could become strong enough to prompt national action.5
Other policy options were recently described in an analysis for the Western Governors Association (WGA) (Gilbert, 1995b). These options, several of which could be adopted by states voluntarily, include
- explicit exemption of telemedicine consultations under specific conditions (e.g., at the request of an in-state physician or when an in-state physician is physically present);
- determination by an authoritative state body that no additional license is required under the concept that telemedicine is analogous to consultations involving the physical transport of patients across state boundaries;
- determination by an authoritative state body that no additional license is required because the referring physician remains responsible for the patient;
For example, large multistate employers secured passage of the Employee Retirement Income Security Act to override state laws regulating most employer-provided retirement, health, and other benefits.
- participation in a uniform regional licensure program that would allow physicians to go through a single channel to obtain a limited license for any of the participating states; and
- provision for institutional or network licensure that would grant a "general telemedicine license" for all network physicians.
A different strategy for achieving greater consistency in licensure laws is to develop and promote state passage of model legislation. This is a well-established vehicle for encouraging uniformity in state statutes covering such diverse areas as insurance, family, and criminal law. The strategy involves the design of prototype legislative provisions for voluntary adoption by states. The lead is often taken by associations of state regulators (e.g., the National Association of Insurance Commissioners), who have direct involvement in the issues at stake, although anyone may propose model legislation.
In the case of telemedicine, the model legislation strategy acknowledges states' interest in protecting the quality of care provided their residents but seeks to avoid undue burdens on out-of-state clinicians and to improve residents' access to appropriate care. The WGA has itself called for "a task force of interested parties to draft a Uniform State Code for Telemedicine Licensure and Credentialing (similar in principle to the Uniform Commercial Code). … The task force could also explore the possibility of expanded state reciprocity in licensing and credentialing as an alternative to model code" (WGA, 1995).
As this report was being drafted, the Federation of State Medical Boards was involved in drafting, circulating for comment, and revising model licensure legislation. It had proposed legislation that would require physicians to apply individually for limited licenses in each state in which they wished to engage in telemedicine practice on a regular and consistent basis (Winn, 1995). The federation would define the practice of medicine to include the provision of treatment to a patient or the provision of written or otherwise documented medical opinions concerning diagnosis or treatment of a patient as a "result of transmission of individual patient data by electronic or other means" from within the state to an outside physician or physician agent. Each state would design the application form and set the licensure fee. If such a proposal were submitted to individual state boards, they alone would have the final vote on its implementation.
It is difficult to predict the number of states that might adopt model legislation, and it is also not clear whether the legislation would lower any barriers to telemedicine.
Among the comments on the draft model legislation are the following (Wood and Whalen, 1995; Gilbert, 1996). First, even though the draft excepts "physician to physician" consultations, some worry that long-established patterns of interstate radiology and pathology consultations conducted by express mail and other nonelectronic means could become vulnerable to varying state interpretations. Second, the model legislation would still leave a significant burden on telemedicine practitioners to acquire multiple licenses. Third, although adoption of the model legislation by states with more restrictive laws would modestly reduce licensure burdens, that could be offset if the model legislation were likewise adopted by states that now have less restrictive statutes. Fourth, the model legislation would still leave physicians to deal with significant variations in state laws concerning, for example, the definition of what constitutes the "regular and consistent" practice of medicine or what constitutes a breach of confidentiality.
The committee did not undertake an exhaustive search for other model legislation but notes that other approaches are possible. For example, the list of policy options offered earlier in this section includes a general exception for telemedicine consultations.
A Note on Credentialing
Credentialing as used here refers to institutional procedures for determining whether individual clinicians have the qualifications to be employed (e.g., as a staff nurse) or granted privileges (e.g., permission to provide medical services generally or subject to certain limits) (IOM, 1990c). State laws, the Joint Commission on Accreditation of Health care Organizations (JCAHCO), the National Commission on Quality Assurance, and other bodies include varying requirements or provisions for credentialing.
For telemedicine, one question is whether a consultant needs to be credentialed not only at his or her "home" institution but also at the satellite or remote institution from which attending physicians request consultation (Gilbert, 1995b). This question is relevant for the intra- as well as the interstate practice of telemedicine. Even within a single state, requiring credentialing beyond the consultant's
home institution could create significant administrative burdens for both practitioners and health care organizations.
The JCAHCO has no general policy on telemedicine and credentialing. It has issued one opinion holding that a telemedicine consultant who responds to a referring physician at another institution in the same state does not have to go through a credentialing process at that institution as long as the referring physician writes all the orders that are in the patient's chart (personal communication, Jay Sanders, 1995; Gilbert, 1995b). Under this opinion, if telemedicine consultants write orders for patients or have primary responsibility for patients, then they would need to be credentialed at the satellite site.
Medical malpractice is conventionally described as a deviation from the accepted medical standard of care that causes injury to a patient for whom a clinician has a duty of care (Kinney and Wilder, 1989; AMA, 1990). Definitions of the standard of care and the duty of care may be provided by case law (i.e., judicial precedent) or by state statute or regulation.
The nature and diversity of state statutory and case law on medical negligence, the uncertainties created by jury trials, and the high cost of malpractice insurance have made malpractice liability policy one of the most visible and contentious health policy issues (Bovjberg, 1989; Hall, 1991; Lawthers et al., 1992; Weiler et al., 1993; White, 1994). Proposals for reform abound at both the state and the national level (Sage et al., 1994; Rosenthal and Brennan, 1996). Most such proposals focus on issues such as limits on punitive damages and the use of alternative ways of resolving disputes. For telemedicine practitioners, the subject of malpractice presents complicated legal issues, some specific to the interstate practice of telemedicine but others applicable to in-state practitioners as well.
Malpractice liability is generally governed by state law. These laws vary in their definition of what constitutes the duty of care owed by clinicians to patients, the specification of the damages that patients may be awarded, and other matters.
One element of malpractice liability—the definition of the standard of care—that could have been a significant problem for interstate telemedicine has become less of an issue in recent years (Hall, 1991; IOM, 1992a). Traditionally, the accepted standard of care was described as the degree of care exercised by clinicians in good standing in the same or similar locality as the defendant physician. In recent years, statutory and case law have moved toward national standards of care—a shift that telemedicine could be expected to reinforce. Other factors, such as resource constraints and existence of a respectable minority opinion, may still be considered by the courts, although the latter concept is not well defined. Nonetheless, the shift toward national standards should help minimize the complexities of applying local standards of medical practice to interstate telemedicine.
One analyst has summarized several questions that practitioners involved in telemedicine have raised about malpractice liability:
1) Has the practitioner formed a physician-patient relationship via telemedicine (which would be the basis for a lawsuit)?; 2) Who is potentially liable for an incorrect diagnosis rendered via telemedicine, the consulting or the referring physician?; 3) If sued, will a telemedicine practitioner be presented with the same legal questions as in an ordinary malpractice case, or must he answer new standards of care?; 4) What are the potential risks and liabilities involved with the use of technologically advanced equipment?; 5) Where is the telemedicine practitioner practicing medicine, the patient's locality or his own, and consequently, where may the telemedicine practitioner be sued for malpractice? [Granade, 1995b, p. 36]
Case law suggests that the likely answer to the first question is yes, the consultant has a duty of care to the patient. Similarly, the likely answer to the second question is that the telemedicine consultant could be held liable for negligence (Granade, 1995b; Young and Waters, 1995). As in other areas of medical practice, the usual protections against liability—good judgment, good skills, good documentation, and good faith—apply, although they do not guarantee immunity from litigation (IOM, 1992a).
With respect to the use of advanced technologies, the basic issues for telemedicine do not appear to be unique. Telecommunications and other technologies may differ in their specifics, but questions
about equipment performance and practitioner proficiency have been raised in a number of areas, for example, in the context of laser surgery and mammography screening. Responses have included the development of professional standards or guidelines and, in some areas, governmental regulations. Systematic, evidence-based guidelines for clinical practice are gaining increasing legal recognition (IOM, 1992a). Although the American Medical Association and others have called for the development of guidelines for telemedicine, responses to date have been limited. As noted in Chapter 3, various standards for computer hardware and software and information exchange have been developed.
The question of whether "distance medicine" imposes a new standard of care has not apparently been litigated. The major uncertainties seem to involve absence of direct patient contact and the possibility that such contact might both provide additional crucial clues to a patient's condition and contribute to the creation of a therapeutic relationship (Granade, 1995b). Further, to the extent that face-to-face contact creates better understanding and sympathy between patient and physician, a telemedicine practitioner might be more vulnerable to suit from a patient who feels psychologically as well as physically distant.
The first four questions posed above would appear to be relevant for the intrastate as well as the interstate practice of telemedicine. The fifth question—where may a practitioner be sued—pertains only to the latter. The question is actually more complex, involving not only where a suit can be filed but also which state law governs. The Rules of Civil Procedure determine jurisdiction. In general, a defendant must either reside in or have had at least minimal contacts with a state to be subject to jurisdiction of that state's courts. Phone calls into a state for the purpose of telemedicine consultations may satisfy the requirement of minimal contact. The law governing decisions in a case, however, is not necessarily the law of the state in which the case is tried. For example, a telemedicine patient might bring suit in the state in which the consultant physician resides. Ordinarily, the law of the patient's home would still apply, but the circumstances of a particular case might prompt a dispute over which law should apply.
Some have argued that the laws of the practitioner's state should apply as they would if the patient was physically rather than "electronically"
transported or "virtually" present, but this argument seems largely theoretical at this time (Sanders and Bashshur, 1995). In any case, telemedicine practitioners face some additional vulnerability and uncertainty related to malpractice exposure in multiple states and would likely face additional expenses for malpractice insurance and for travel and other costs should a suit be filed in a distant state.
In addition to questions about individual liability, questions also arise about organizational liability for errors in transmitting, storing, or otherwise managing information transfer. Such questions might also arise with computer-aided diagnosis programs. These issues, however, have not figured prominently in discussions of state policies and telemedicine.
Although telemedicine poses liability concerns, it has also encouraged some interesting potential protections in the case of litigation. For example, video consultations can be taped and included in the patient record to provide documentation of the care provided.6 Patient-record software is now available to permit the incorporation and retrieval of video and other images and audio information (e.g., chest sounds) in computer-based patient records. Such records also may have educational and research uses. As discussed below, they may additionally intensify concerns about the security of personal medical information.
Just as a national licensure system or national provisions for telemedicine could resolve many concerns about state licensure as an impediment to telemedicine, so could similar provisions for malpractice. Although reform of state medical practice laws and imposition of some national requirements or limitations (e.g., limits on certain kinds of damages) have been health policy issues for several years, none of the major reform proposals would totally replace state with federal jurisdiction (Rosenthal and Brennan, 1996).
Telemedicine has not featured in broad reform proposals nor has
malpractice been generally addressed in bills dealing more specifically with telemedicine. Further, because telemedicine is relatively new, it has not yet figured noticeably in statutory or case law. In at least one state, Oregon, legislation has been proposed to require malpractice insurers to make telemedicine coverage available. Legislation has been introduced in Colorado to set minimum standards for teleradiology, including a requirement that radiologists carry appropriate malpractice insurance (Telemedicine Monitor, 1995).
Proposals to require telemedicine practitioners to carry appropriate malpractice insurance and to require malpractice insurers to make such coverage available highlight the relevance of private as well as public policies—in this case, those of malpractice insurers. It is the committee's understanding that virtually no discussions have been held with major malpractice insurers regarding telemedicine. One professional-liability insurer, St. Paul Fire & Marine, has, however, declared that its policies would cover cases arising from video consultations (Crump and Pfeil, 1995).
State malpractice laws present several potential complexities for telemedicine practice. Consider, for example, a Georgia physician who also practices in two other states and is thereby also subject to their jurisdiction.7 Imagine that these states differ. In one, physician-defendants tend to prevail, and physicians consider awards made to patient-plaintiffs to be reasonable. In another state, patient-plaintiffs tend to prevail, to sue frivolously, and to win awards that physicians consider unreasonable. Based on existing legal rules, even if the alleged malpractice occurred in the first state, the physician-defendant could be sued in the other state. How would the malpractice insurer of this Georgia physician deal with this situation? Would separate policies for each state be required? These questions remain unresolved (Granade, 1995c).
Privacy, Confidentiality, And Security
Serious questions have been posed about current legal protections for medical privacy and confidentiality and the relevance of these concepts to modern health care (see, for example, OTA, 1986a, 1993; Gostin et al., 1993; IOM, 1993b; Gostin, 1995; Schwartz,
1995). These questions are not peculiar to telemedicine or even to information technologies more generally. Conventional health care practices and paper medical records offer numerous opportunities for unintentional, careless, or deliberate infringements of medical privacy.
Nonetheless, the electronic recording, storage, transmission, and retrieval of patient information has complicated the situation and increased the opportunities for the privacy and confidentiality of personal medical information to be infringed (IOM, 1991, 1994b; WEDI, 1993; AMA, 1995; Eid, 1995; Gilbert, 1995a; Kolata, 1995; Woodward, 1995). Such personal medical information extends beyond the written word to include still images, audio records, and videos of patients.
These developments have required continuing reassessment of (a) the trade-offs between privacy concerns and other values such as convenient and quick access to information and (b) the practical realities of enforcing an agreed-upon balance of competing objectives. Some organizations are attempting to implement a tiered system that provides access to patient information on a "need to know" basis, and they have also developed confidentiality agreements to be signed by their own personnel, subcontractors, vendors, and others who have access to personal medical information (see, e.g., Dargahi et al., 1994, and Henkind et al., 1994). Other organizations appear to have balanced the scale heavily toward the administrative, financial, and clinical considerations so that anyone needing access to some information has access to virtually all electronic records for current and past patients. For example, patients who had received mental health services at the Harvard Community Health Plan discovered that their records were available to all health care personnel who provided them care and that the records were very detailed (Lewin, 1996; Page, 1996b). The plan has restructured access to records and developed a training program on confidentiality, but concerns remain that the records are too detailed.
As discussed in a 1994 IOM report on health data, privacy has several aspects related to (a) freedom of individual decisionmaking (e.g., about sexual or reproductive options), (b) protection of one's person or property from surveillance or search, and (c) control of personal information (IOM, 1994b). The last concept, which may be termed informational privacy, is the focus of the present discussion.
Confidentiality refers to a relationship (e.g., between patient and physician) that includes an expectation that personal information obtained as part of that relationship will be protected. The 1994 IOM report recommended a new conceptualization of data confidentiality as involving "data that have been declared to be sensitive and must be protected and handled as such" (p. 154). This concept focuses on the information itself rather than on its context in an interpersonal relationship such as that of patient and physician.
The 1994 report discussed data security in terms of defined structures and processes for protecting defined information for defined users and systems. As proposed by the earlier IOM committee, data security is a matter of technology and management whereas data confidentiality is a matter for law and regulation.
The National Library of Medicine has asked another National Academy of Sciences body, the Computer Science and Telecommunications Board (in collaboration with the IOM), to assess social and technical mechanisms for protecting privacy and security and to outline promising areas for future development and testing. It will issue an interim report in the latter part of 1996 and a final report in early 1997.
As noted in the discussion of licensure, state licensure laws for health professionals and health care institutions include varying provisions regarding the confidentiality of medical information (see also Gelman, 1995). States may, however, have no provisions regarding breaches of confidentiality by nonclinicians.
The U.S. Constitution offers protection from governmental infringement of privacy but not from infringement by private individuals or organizations. A common observation about current protections for medical information is that Congress has protected the privacy of video rental records to a greater extent than medical records (Gelman, 1995; Schwartz, 1995). As this report was being prepared, proposals to expand medical privacy protections were being debated in Congress (see discussion later in this section), but the outcome was still uncertain.
Existing federal legislation protects the confidentiality of certain patient information available to governments, and various Medicare regulations impose confidentiality requirements upon health professionals
and institutions. In addition, the Americans with Disabilities Act (ADA) was intended to safeguard certain kinds of employee medical information and protect employees with disabilities from discrimination, but the scope and interpretation of the law are not yet clearly defined. Although the law restricts employer access to information from preemployment physical examinations and certain other sources, it is silent about employer access to the much greater volume and sensitivity of information that is generated from employee health insurance programs.8 Another IOM committee has recommended that the protections of the ADA be extended to this latter information (IOM, 1993b).
Private organizations may also establish policies to protect the privacy of personal medical information. The JCAHO, for example, includes privacy protections in its accreditation standards for health care institutions. The American Medical Association has set forth ethical standards that may "take on the force of law when they are expressly incorporated, or implied in, physician licensure laws" (AMA, 1995, p.3).
Centuries ago, the Hippocratic Oath required that physicians keep silent what they learn about people, "counting such things to be as sacred secrets" (cited in IOM, 1994b, p. 148). The expectation of such secrecy has become increasingly fragile as physicians' control over information has diminished and as more and more individuals have claimed access to personal medical information for people with whom they have no personal connection and for whom they lack professional responsibility. Contemporary threats to informational privacy arise from several interrelated developments, including the computer-based patient record, the creation of large databases, new
applications of telecommunications technologies, and the growing commercial value of patient information.
The 1994 IOM report on health data warned that "unless security systems are designed to record access, the curious, entrepreneurial or venal can enter databases without leaving evidence of having done so" (IOM, 1994b, p. 141). It also observed that it can be expensive and inconvenient to protect personal medical information and that the harms as well as the benefits of restricting access should be considered in designing security policies and procedures.
Early in developing this report, this committee recognized the important role that the computer-based patient record is playing and will play in the provision, assessment, and improvement of health care services. The ongoing shift from the paper-based to the computer-based patient record is not, however, without risks. Benefits and risks stem from the same basic capacities of the computer-based patient record, which are that it
makes searching records for specific information much more convenient;
allows for easier collection and updating from multiple sources of comprehensive information about a patient;
makes information about an individual patient potentially available to more people, including those far removed from that patient; and
potentially allows easier access to more information about a patient than the user either requests or needs.
Information networks and telecommunication links present further opportunities for infringements of privacy. The electronic transmission of data from computer to computer and site to site increases the opportunity for unauthorized interception of personal information. Most notable, perhaps, in their sensitivity are video transmitted and taped psychiatric consultations. With the electronic transmission of information, encounters that were once merely summarized on paper may be more likely to be recorded and stored for later access as part of quality review, education, or legal proceedings. In fact, such recording was cited earlier in this chapter as a potential benefit of telemedicine in malpractice litigation.
Even without the full implementation of the computer-based
patient record, large health databases have been created to provide easy electronic access to much personal medical information (IOM, 1994b; Gelman, 1995; Borzo, 1996b). These databases, some of which are community-based and others of which are proprietary, serve a variety of users and uses, including management of health care organizations and health plans; provider payment; employee benefits management; insurance marketing and underwriting; health services research; and health industry market research. In many instances, such uses of health information are essentially unregulated. Also largely unregulated is the private market for individual social security numbers that are used as identifiers in a wide variety of health and nonhealth applications involving medical, financial, credit, educational, and other records (Flynn, 1996).
In response to growing concerns about threats to medical privacy and confidentiality, several proposals have been put forth to substitute a uniform national policy for the current "patchwork" of state laws, many of which offer little if any protection to individuals, especially from actions by nonclinicians. As indicated in the discussion of licensure, the diversity of state laws poses a challenge for telemedicine because consultants may be held responsible for knowing and adhering to different laws for every state in which they consult.
During the unsuccessful health care reform initiative following the 1992 presidential election, a task force drafted privacy protections as part of proposed legislation (the Health Security Act of 1993). The IOM and the Office of Technology Assessment (OTA) also proposed several provisions to be considered in designing privacy legislation for health information (OTA, 1993; IOM, 1994b).
More recently, the Medical Records Confidentiality Act (S. 1360) and the Fair Health Information Practices Act (H.R. 435), both introduced in 1995, and the Medical Privacy in the Age of New Technologies Act of 1996 (H.R. 3482) would establish certain federal privacy protections and regularize certain procedures for the routine collection, maintenance, distribution, and use of personal medical information. S. 1360 would require patient authorization for disclosures of information for purposes of medical treatment or payment; allow patients to obtain, copy, and correct their records;
provide civil and criminal penalties for unauthorized disclosure of patient records; require that records be kept of disclosures of patient data; and require the Department of Health and Human Services to draft standards for organizational protection of personal medical information. Under certain conditions, the proposal would also allow researchers, public health authorities, and health oversight organizations (e.g., licensing bodies, claims review organizations) to gain access to personal medical information without patient authorization or notification and would provide law enforcement authorities access to information without patient authorization but with notification under some circumstances. Much of the controversy over the legislative proposal revolves around these exceptions (Kolata, 1995; Schwartz, 1995). H.R. 3482 would, in addition, allow states to set stricter privacy rules and let patients designate some information (e.g., HIV status) as specifically protected (Page, 1996b).
An alternative to federal legislation establishing more uniform protection for personal medical information is the voluntary adoption of model state legislation (Eid, 1995). The American Medical Association, for example, has proposed model privacy legislation (AMA, 1995). Earlier model legislation drafted by the National Association of Insurance Commissioners in 1985 has not been widely and consistently adopted by states (Gelman, 1995).
Technical and Administrative Options
In response to concerns from a wide variety of business, consumer, government, and other organizations, security systems and procedures have been developed to guard electronically stored and transmitted information against misuse (see, e.g., Hammond, 1992; OTA, 1993, 1995; WEDI, 1993; IOM, 1994b; Gilbert, 1995a; Young and Waters, 1995). In addition to protecting sensitive personal, commercial, and national security information, these procedures are also intended to protect organizations from intrusions that destroy, damage, or disrupt information and operating systems. They likewise help shield organizations from liability for damages resulting from lax security (e.g., if a hacker or disgruntled employee changed pharmacy orders or revealed information in patient records). Security measures include
- data encryption techniques that encode information so that it is not easily read without a code book or decryption key;
- authentication procedures to ensure that messages are received from the stated source exactly as they were sent;
- authorization procedures that determine whether a user is permitted access to particular information;
- auditing and tracking programs that provide information about those who have gained access to a system; and
- so-called "firewalls" that encompass a range of access control mechanisms that either block or permit access to one network from another.
Developers of security procedures are engaged in a constant struggle with those who seek to defeat such procedures either for the challenge of doing so or for some kind of economic or personal advantage. Similarly, proponents of these mechanisms are continually working to educate individuals and organizations of varying technical, legal, and social sophistication about the importance of good security measures.
Payment Policies For Telemedicine9
As noted in Chapter 3, insurer and health plan restrictions on fee-for-service payments to physicians for telemedicine consultations are viewed as a major barrier to telemedicine's growth. Environments not based on fee-for-service payments to providers tend to be seen as less constraining for telemedicine. For example, the use of telemedicine in the military, veterans' health, and prison health care systems, in which most care is provided by salaried or contracting clinicians, is not limited by fee-for-service payment policies or concerns.
Because so much medical care is provided to the elderly and paid for by Medicare on a fee-for-service basis and because private payers may follow Medicare's lead, fee-for-service payment policies established by the Health Care Financing Administration have been a dominant concern for telemedicine proponents. Any analysis of
payment for telemedicine services, however, needs to recognize the role of private policymakers in insurance companies, managed care plans, and similar organizations. In particular, the growth of private managed care organizations that use capitation or salaries to pay for physician services is likely to reduce the focus on fee-for-service payment, particularly should Congress undertake a major overhaul of Medicare designed to move more beneficiaries to such organizations.
In both public and private sectors and regardless of payment method, the major issue for policymakers and managers is whether any additional benefits provided to patients by telemedicine are worth any additional costs, including costs of possible increased use of both appropriate and inappropriate health services. Another issue is whether it is appropriate to ask patients to pay some or all of any higher costs associated with a particular telemedicine service, particularly if that service is more convenient for patients but adds to total health care costs.
The following discussion considers three major types of payment schemes: fee-for-service; per case or bundled rates; and capitated or fixed budget payments. Although a specific payer's policies can blur the boundaries, these are still useful distinctions for policy discussions. Chapter 7 presents a fuller analysis of economic issues in telemedicine, including how to measure or estimate the actual cost of delivering telemedicine services and how to analyze cost impacts for different parties and decisionmakers.
Fee-for-Service Payment and Telemedicine
Despite the growth of other ways of paying for medical care, fee-for-service still dominates payment for personal health care services, and most payers relying on this method do not pay for most kinds of telemedicine consultations (Grigsby, 1995a). One of the major obstacles to coverage of telemedicine on a fee-for-service basis is the fear that costs would escalate because of higher use of both appropriate and inappropriate services and that utilization and quality review mechanisms would not be sufficient to control inappropriate use.
As a general policy, Medicare covers consultative services provided on a "face-to-face" basis and proscribes coverage for telephone consultations. The Health Care Financing Administration
(HCFA) and other insurers do, however, reimburse for teleradiology and telepathology consultations, which do not ordinarily involve direct patient contact in any case. Except for four cardiology services (e.g., transmitting an EKG), Medicare procedure (CPT) codes do not generally distinguish telemedicine from other consultations for specialties such as radiology and pathology that do not normally involve face-to-face contact with patients (PPRC, 1995). Thus, payers may not even know whether telecommunications technologies are being used by certain consultants and may not be able to generate data on claims and expenditures for such services. In a few states, including Georgia and Kansas, coverage by Blue Cross and Blue Shield plans or other payers has been negotiated for a broader range of telemedicine services (McIlrath, 1995b).
In its 1995 report to Congress, the Physician Payment Review Commission (PPRC) identified several issues in coverage decisionmaking, most of which were earlier identified by Grigsby and his colleagues in their reports to HCFA (see Grigsby et al., 1995). The issues include (PPRC, 1995, p. 135)
- lack of information on the value of telemedicine applications compared with the traditional services they would replace;
- uncertainty about whether various telemedicine payment methods might stimulate excess service use or otherwise affect service patterns;
- potentially large increases in utilization and costs if telemedicine services improve access to care; and
- barriers to the sustainability of telemedicine systems in some rural markets.
In establishing coverage policies for medical services, HCFA considers evidence that the service is effective. As discussed in later chapters of this report, the evidence of effectiveness for most of the array of clinical applications of telemedicine is sparse. HCFA, however, has funded analytic work on telemedicine and is sponsoring demonstration projects to provide information for policymaking (see Chapter 5 and Appendix A). Reimbursement policy changes do not appear imminent (Richardson, 1996).
To set payments for specific telemedicine services, HCFA or other payers would have to make several important decisions (Grigsby et al., 1994b; PPRC, 1995). These include decisions about
- which types of telemedicine services would be covered;
- how procedures would be coded for claims payment and other purposes;
- how the payment level would be set for specific services;
- how payment would be apportioned for a consultation involving an attending physician and a distant specialist;
- whether and how nonphysician practitioners would be paid; and
- whether and how payments would be divided into professional and facility components.
Payers would, for example, have to determine whether to pay the same rate for a service (e.g., psychiatric consultation) regardless of the mode of service (e.g., office visit versus telemedicine) or to set special rates for telemedicine. If special rates were negotiated, the amounts could depend on the payer's leverage vis-à-vis the provider, with the provider wanting rates as high as possible and the payer wanting rates as low as possible.
For physician payment rates based on resource costs (e.g., the Medicare relative value scale or RVS), a payer could set a higher payment rate for a telemedicine service if that service were more expensive to provide than alternatives and if the benefits were judged to be worth the extra costs. Conversely, if a telemedicine service were less expensive to provide than the alternative, the payer could lower the payment rate for telemedicine.
If payment rates exceed providers' costs, providers will be more inclined to use telemedicine in discretionary situations. This situation has characterized other innovative technologies and is inherent in the incentives of fee-for-service payment. Although payers may establish mechanisms to monitor appropriate utilization, such efforts have often proved difficult or unsatisfactory for a variety of clinical, administrative, political, and financial reasons (PPRC, 1988; IOM, 1989).
Per Case or Other Bundled Payment Methods
One alternative to traditional fee-for-service payment is payment of a fixed rate for a package of related services. This method has long been used for surgical services, for which one global payment covers physician services before, during, and after a procedure such as coronary artery bypass (CBO, 1986; OTA, 1986b; PPRC, 1987).
During the 1980s, Medicare adopted a diagnosis related group (DRG) payment scheme that applies the global payment concept more broadly to inpatient hospital care (ProPAC, 1986, 1996). After making certain adjustments related to patient characteristics and other factors, the program pays hospitals fixed amounts for the care of patients with a designated diagnosis regardless of the specific diagnostic and therapeutic services provided. Although methodological and other problems need to be resolved, the packaged payment approach could also be extended to additional episodes of care, such as ambulatory, home, or nursing home care for certain conditions (ProPAC, 1996). Such approaches are important elements in various state health proposals and managed care plans.
Just as fee-for-service payments may be allowed to vary geographically, a per-case payment approach can also adjust the schedule of payment rates by locality to reflect, for example, different costs of providing services in rural or urban areas, but differences in rates would not necessarily hinge on whether or not telemedicine was used. Payers could also adjust payment rates to reflect whether the use of telemedicine substantially raised expenses and was judged to be worth the greater health benefits achieved, or whether telemedicine lowered expenses to providers for delivering given services and health benefits. If telemedicine raised providers' costs for delivering the package of services and the benefits gained were worth the extra costs, the payer and providers could negotiate increased payment rates.
On the positive side, per case and similar payment methods reduce incentives for providers to provide more of certain services in order to generate more fee-for-service payments. Providers determine how best to marshal resources to deliver efficiently the services covered within the payment rate. On the negative side, because payment remains the same regardless of the services actually provided, providers also have an incentive to restrict the provision of needed services for a case covered by a global payment. The need for monitoring to ensure access to and quality of care would thus continue.
Capitation Payment/Fixed Budget
Capitation payment entails a predetermined fixed payment per person for all benefits covered during a specified time period, regardless
of the type and number of services actually used. The payment may be provided to a health plan or system, which may or may not pay providers of care on a capitated basis. Usually, patients pay only a modest copayment for services received from health care providers participating in a capitated plan or system. The committee heard suggestions that because telemedicine might reduce time, travel, and other costs for patients, patients might be expected to make some copayment for telemedicine services offered by a capitated health plan.
Capitated health plans have had their greatest success in urban areas. The small population base, limited resources, and relative absence of competition in rural areas have made it difficult to extend such plans to rural markets. One of the potential attractions of telemedicine for capitated organizations is that it provides a mechanism for reaching into rural areas.
Because health plans and providers are financially at risk, capitation payment provides incentives for them to employ the most efficient means to deliver and manage services across a range of covered benefits for a defined population of enrollees. Because the use of telemedicine would be neither encouraged nor discouraged by payment method, decisions about whether to "move" specialized services and facilities to patients through telemedicine or to transport patients to those services and facilities can be based on a more straightforward comparison of benefits and costs. HCFA allows "at risk" Medicare HMOs to use telemedicine services, but it will not adjust their payments for such services (and none appear to be offering telemedicine services). The face-to-face consultation requirements, however, apply to HMO plans paid on the basis of their costs.
If telemedicine raised providers' costs of delivering care and the health benefits gained were worth the extra costs, payers, health plans, and providers would face negotiations to increase capitation rates accordingly. Conversely, if telemedicine lowered costs, payer pressure to cut health care could prompt negotiations on some reduction of capitation rates.
Like bundled payment methods, capitation payment provides incentives for efficiency, on the one hand, and for underprovision of appropriate care, on the other. Thus, monitoring is needed to ensure access to and quality of care.
Other Financing Mechanisms
As important as payment and coverage policies are, other avenues of public and private financing for telemedicine should not be ignored. As discussed earlier, both federal and state governments have helped support a telecommunications infrastructure that can be used for health care, educational, and other purposes. Demonstration projects may remain an important vehicle of government funding (preferably with strong provisions for evaluation). In addition, commercial organizations (e.g., software vendors) seeking to develop and expand product markets have provided both direct funding and contributions in-kind of equipment and technical assistance. If, however, telemedicine is not accepted as cost-effective and is not incorporated—one way or another—into the everyday financing of patient care services, it will be difficult to sustain telemedicine once grants expire, donated equipment becomes outdated, and routine operating as well as up-front capital costs have to be covered. Chapters 6, 7, and 8 discuss economic issues and sustainability concerns in more depth.
Regulation Of Medical Devices10
Some of the devices employed in telemedicine are subject to regulation by the federal Food and Drug Administration (FDA), primarily through its Center for Devices and Radiological Health (CDRH).11 CDRH regulates medical devices and radiation-emitting electronic products used for telemedicine, and it sets standards for mammography personnel, equipment, and practices. Its programs are intended to assure that medical devices are safe, effective, and properly manufactured; to promote quality in mammographic services including telemammography; and to control unnecessary human exposure to potentially hazardous radiation and ensure the safe, efficacious use of such radiation. CDRH is represented on the
This section is based on material drafted by Mel Greberman.
The FDA and CDRH currently use electronic systems to make information available to interested parties. For example, the CDRH Home Page on the World Wide Web includes device safety alerts; Federal Register reprints; information on premarket submissions, small manufacturers assistance, video conferencing, and electronic submissions; and other medical device oriented information (http://www.fda.gov/cdrh/cdrhhome.html).
federal Joint Working Group on Telemedicine and has been involved in the efforts to establish several health care informatics, medical imaging, and other standard-setting activities that were described in Chapter 3.
CDRH programs are conducted under the authority of the Federal Food, Drug, and Cosmetic Act (FD&C Act), the Medical Device Amendments of 1976, and several later amendments. According to the FD&C Act, a medical device is
… an instrument, apparatus, machine, contrivance, implant, in vitro reagent, or other similar or related article, including any component, part, or accessory, which is—
(1) recognized in the official National Formulary, or the United States Pharmacopeia, or any supplement to them, (2) intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease, in man or other animals, or (3) intended to affect the structure or any function of the body of man or other animals, and which does not achieve its primary intended purposes through chemical action within or on the body of man or other animals and which is not dependent upon being metabolized for the achievement of its primary intended purposes.
Medical devices include devices that produce medical images (e.g., radiographic and fluoroscopic x-ray machines, x-ray computed tomography and nuclear medicine scanners, diagnostic ultrasound devices). The software and equipment used to transmit, store, process, display, and copy medical images may be treated by CDRH as accessories to the imaging devices or as separate medical devices.
The CDRH is considering revisions in its requirements regarding these products that would exempt from significant regulatory requirements certain products that the FDA believes do not pose significant risk to patients (Zaremba and Phillips, 1993; Zaremba and Anderson, 1996). The revisions, which would be published in the Federal Register for public comment, would establish five generic categories of products related to medical image management and communications. These categories would cover medical image digitizers, communications devices, storage devices, hardcopy devices, and picture archiving and communications systems (PACS).
The PACS category would include both relatively small products (e.g., portable devices that transmit images over telephone lines to an on-call radiologist) and large, complex systems that utilize fiber optic networks and advanced communications methods to transmit
images for entire health care facilities. The category would also include the medical image workstation, which generally consists of a computer, video monitor, and storage device. The computer often utilizes software related to data communications, file management, image processing (e.g., edge enhancement), measurement, and special image displays (e.g., 3D surface and volume rendering).
Currently, some stand-alone software products (e.g., some decision support systems) also fit the definition of a medical device as described in draft policies developed in the 1980s (FDA, 1989). Other software products are not covered if they are intended only to serve traditional library, accounting, communications, or educational functions and are not used to diagnose or treat patients.
As this report was being drafted, the FDA was exploring new policies that would allow the agency to assess the risks of software devices and decide which might be exempt from certain regulatory requirements. It was also planning to hold a workshop to obtain public suggestions and comments on how the health risks of medical software devices might be assessed and how policies might be defined. In addition, the CDRH had established a Computer Aided Diagnosis (CADx) Working Group to develop guidelines for agency reviewers of CADx software devices.
The task for this committee was to develop an evaluation framework for clinical telemedicine—not to develop policy recommendations. The committee recognized, however, that policies related to licensure, malpractice, and other matters need to be considered in an evaluation framework because they may affect the availability, acceptability, effectiveness, and cost of telemedicine services.
By providing an overview of policy issues, this chapter along with Chapters 2 and 3 has attempted to provide background and context relevant to the task of developing an evaluation framework for clinical applications of telemedicine. The next chapter considers existing and planned evaluations of telemedicine.