Physician Licensure and the Quality of Care
The Role of New Information Technologies

Paul B. Ginsburg and Ernest Moy

Paul B. Ginsburg is executive director of the Physician Payment Review Commission. Ernest Moy is assistant professor of medicine at the University of Maryland School of Medicine.

Licensure of professionals attempts to raise the quality of services consumers receive by barring those who do not meet a standard from providing those services. The process involves two distinct quality-control measures. First, a governing body establishes the basic qualifications the professional must meet to obtain a license to practice. Those qualifications often include educational and training requirements and demonstration of competence by passing one or more examinations. In some instances professionals must periodically renew their licenses to prove that they have maintained their skills. Second, the governing body can revoke the license if there is evidence that services are substandard or that privileges such as prescription writing are abused.

Economists have long questioned whether the goal of removing supposedly low-quality practitioners from the market is an appropriate one. Many have advocated replacing licensure with a program that provides consumers with information on practitioners' qualifications and allows them to choose whether to seek services from practitioners not meeting the standards for certification.

The argument for not excluding any practitioners from the marketplace is that some consumers would choose those with lower qualifications to get a lower price or to obtain an approach that is not accepted by the mainstream of the profession. Those consumers would benefit from some combination of lower cost and increased access to services. The argument continues that a minimum standard set by government permits the profession to influence the licensing process to erect a restrictive entry barrier and to limit the ways in which its members compete. Thus, consumers pay higher prices for all services.

Acknowledging that many consumers desire guidance on minimum standards, opponents of licensure suggest that certification can provide that guidance without impeding those not seeking such assistance. Certification can also provide additional information by setting multiple standards, either by a single organization offering a range of certificates or by competing certifying organizations.

Proponents of licensure use two distinct arguments. The more traditional one is that given full information, there may be a standard on which virtually all consumers would insist. Precluding practitioners who do not meet that standard from providing services would economize on the time and resources consumers need to collect the information to make their decisions (or to learn about certification and its meaning). Such may be particularly true in medical care, where consumers are ill-equipped to evaluate the technical competence of physicians. If an alternative process of certification resulted in most consumers' restricting themselves to certified practitioners, such a process, most likely run by the profession, could be even more restrictive of entry than licensure, which incorporates the views of other interests.

A newer argument for licensure is that conferring an economic rent on members of a profession through a barrier to entry will lead them to make additional efforts to provide services in a competent and ethical manner, since the penalties from losing the privilege through disciplinary action become very large.

Our goal in writing this article is not to enter this relatively old debate, but to discuss how recent changes in both knowledge about quality of medical care and ability to monitor quality are likely to have profound effects on what licensure or certification attempts to do--provide purchasers of medical care with information on physicians' quality and assist practitioners in improving their care. We explain how more extensive activities by purchasers of care are likely to diminish the relative importance of licensure as we know it today.

While much of the previous debate focused on licensure and certification, some of the most important developments have concerned the expansion of "privileging." In contrast to a model with patients as the sole users of information provided by certification, we find that hospitals, ambulatory care facilities, and various types of insurers--public and private, health maintenance organization, and indemnity--are establishing policies concerning which physicians are to be allowed to perform a procedure or to be paid for performing a procedure. Privileging can be based on formal certification or on other criteria, such as documented experience.


Medical Licensure and Certification

Medical licensure by the states has been common since the late nineteenth century, but certification is extensive as well. Physicians have been licensed by the states. The initial license is typically based on graduation from a medical school accredited by the Liaison Committee on Medical Education, completion of at least one year of graduate medical education in a residency program accredited by the Accreditation Council for Graduate Medical Education, and passing either the National Board of Medical Examiners Examinations or the Federal Licensing Examination. Renewal of the license often requires completion of a certain number of units of continuing medical education. A physician can have his license restricted, suspended, or revoked if the licensing board determines that the individual is unethical, impaired, or incompetent.

Physicians may also be certified by medical specialty boards. Those private organizations have established graduate medical educational (residency) requirements and give examinations to determine whether a licensed physician has met an additional standard. Over time, a larger proportion of physicians has become board-certified, as more physicians have sought to meet the requirements of their specialty and as additional specialties have established such a credential.

Other credentials exist as well. For example, a physician may have completed a course of study that "certifies" him to provide particular services, such as cardiopulmonary resuscitation or advanced cardiac life support, or to perform specific procedures. Many specialty societies offer honors such as fellowship in the society and awards such as the American Medical Association's Physician's Recognition Award.

The importance of credentials beyond licensure in the medical marketplace appears to be increasing. For example, HMOs and group practices use board certification in decisions about hiring physicians and in advertisements about quality. Hospitals use board certification and other credentials to determine whether a physician will be permitted to practice there and whether there will be any restrictions on procedures performed.

Insurers are beginning to use credentialling to limit payment for particular services to approved practitioners. In some cases approval is based on board certification in a particular specialty. Insurers are considering the use of credentials more closely tied to the specific service, such as specialized training or experience. Medicare limits coverage of transplants to "centers of excellence" that can demonstrate levels of experience and success rates. Several Medicare carriers (private insurers that process claims for Medicare under contract) require physicians to meet specific procedure-related educational and training requirements to be eligible to be paid by Medicare.


Changing Role of Licensure

Licensure has been playing a less important role in the market for physicians over time. Not only has it become less significant an entry barrier over time, but the educational requirements for licensure prepare physicians for an increasingly small portion of clinical practice. Licensure's role in assuring quality though discipline of those practitioners who no longer meet standards has also declined.

The most significant change affecting entry barriers has been the increase during the 1960s in the size of U.S. medical school classes. Having concluded that physicians were in shortage, the federal government embarked on a program to increase their numbers. Capital grants and operating subsidies were given to both new medical schools and to existing ones that agreed to increase class size. In addition, physicians applying to immigrate to the United States were given preferential status.

That policy was very successful in increasing the supply of physicians, with the number of nonfederal physicians in patient care increasing from 124 per 100,000 civilians in 1965 to 198 per 100,000 civilians in 1990. Indeed, analysts have been debating whether the policy "overshot the mark" and turned a shortage into a surplus.

Many economists have suggested that the increased supply of physicians has been critical to many of the competitive developments in the market for physicians' services. They argue that without increases in supply, HMOs and preferred provider organizations would have had great difficulty recruiting adequate numbers of physicians to affiliate or contract with them.

At the same time as more persons have trained to be physicians, other health professionals have earned the privilege to perform certain services that had been reserved exclusively for physicians. For example, optometrists are generally permitted to dilate pupils as part of routine eye examinations. Nurse practitioners have been granted privileges to perform many primary care services under supervision of physicians. In areas of chronic physician shortages, they are permitted to practice independently. Certified nurse anesthetists can provide anesthesia services independently in several states.

Interestingly, many of the "turf battles" between physicians and other health professionals have been fought not in the context of professional practice acts, which determine those services for which a physician's license is required, but in the context of coverage policies by insurers, especially public insurers such as Medicare and Medicaid. With the extensiveness of health insurance, the ability to bill an insurer for a service has become almost as important as the license to provide it.

Insurers, especially private ones, have been more protective of physician entry barriers than some would have expected. Whereas some would suggest that an insurer would want to encourage expanded supply and lower prices by allowing the substitution of nonphysician practitioners, many insurers instead have been concerned that an increased supply of providers would instead increase outlays for services. Interestingly, empirical support for the notion that increased supply increases outlays comes from economists on both sides of the issue of whether classical economic models can explain the market for physicians' services.

Licensure's role has also declined because medicine has become more specialized. The requirements for a license--four years of medical school and a year's internship--prepare physicians for an increasingly limited part of modern clinical practice. Licensure requirements have changed little during a period in which medicine has become more specialized and requires large investments in specialized training. Indeed, many physicians decline to perform procedures or see patients with problems outside their specialty or subspecialty because they do not deem themselves to be competent or because others, such as hospitals, will not permit them to do so. Consequently, licensure's influence on who can provide medical care has decreased.

While economists have long focused on the role of licensure as an entry barrier, most of the resources in the process are devoted to monitoring the quality of the services of licensed physicians. Each state has a medical board empowered to limit or revoke a license on the basis of "unprofessional conduct." The boards receive complaints from patients and providers, investigate those complaints, and discipline the physicians when indicated.

Many question the effectiveness of those processes. The rate of disciplinary actions varies from state to state and is generally very low. From 1985 through 1987 the rate of licensing board actions reported to the Federation of State Medical Boards ranged from 2.6 to 32.2 per 1,000 licensees. A 1985 survey by physician-owned malpractice insurers showed that while 3.2 percent of physicians were sanctioned by the insurers for poor quality, state boards took reportable disciplinary action against only .54 percent for all reasons, less than 15 percent of which were for poor quality. That suggests that state boards disciplined a very small fraction of those physicians malpractice insurers identified as having quality problems.

The ineffectiveness of many boards at disciplining physicians who are performing poorly stems from a number of factors. First, the boards rely on complaints to bring cases to their attention, but few physicians file complaints about their peers. Second, boards have limited resources. With the increasing litigiousness of physicians facing sanctions, inadequate resources have severely limited the number of cases that can be pursued. Third, the standard of "clear and convincing" evidence needed by most state boards to discipline a physician is a difficult one to meet.

Effectiveness of medical boards may be increasing, however. Many states have passed mandatory reporting laws and have increased immunity protections to those physicians filing complaints. Some have lowered the standard for discipline to a "preponderance of evidence." From 1985 to 1987 the number of reportable actions increased 29 percent.


New Technologies to Monitor Quality

The technology to monitor the quality of physicians' services has advanced dramatically in recent years. Hospitals and health care purchasers have been experimenting with building profiles of physicians' practices and comparing them with standards. Because of the limitations of the data that are now available, much of the effort has gone into feeding information back to physicians to guide but not dictate changes in practice. Increasingly, attempts are being made to evaluate the quality of a physician's care to improve decisions about staff privileges or inclusion in a network of physicians.

Ideally, monitoring the quality of a physician's services involves three components. First, how the physician is practicing must be documented. That includes information both on the patient's medical problems and other characteristics and on the services that the physician performs or orders. Second, information on desirable courses of diagnosis and treatment is needed. Third, the physician's practices must be compared with the standard.

Advances in data processing are making the first task--obtaining information on the patient and on the physician's practice--more feasible. Initial work has emphasized the administrative data used by insurers to process claims. Those data include information on the patient's age, sex, and diagnosis. They include detailed information on the services that the physician performs or supervises. The limitations of administrative data include the accuracy and specificity with which services are described, the quality of the diagnostic information, difficulties in identifying services provided on referral or ordered by other physicians, and difficulties in combining data from different insurers to build a profile of sufficient size to represent a physician's practice.

The Medicare program appears to be leading the way in developing claims-based data systems. Its Common Working File ensures that the various Medicare carriers adhere to common standards. A new system of unique physician identifier numbers provides for the first time a mechanism to distinguish those services that a physician provides directly to Medicare patients from those services that he provides on referral. Improved specialty designations will permit better sorting of physicians to develop profiles.

Extensive efforts are underway to develop standards for administrative data that would include private insurers. While the major stated motivation is to reduce administrative costs, such a measure is also a first step toward broadening the availability of administrative data for profiling. Experiments sponsored by the Hartford Foundation are developing such an infrastructure.

Clinical data are also useful to profile physicians, especially when profiles are used to judge physicians' quality. At this point, hospitals and HMOs have clinical data to profile their affiliated physicians, but little clinical data are available in usable form outside provider organizations. A notable exception is the Medicare program, whose peer review organizations abstract large numbers of hospital records to monitor the quality of care for patients. The orientation to date, however, has been to examine the quality of care provided to an individual patient in a particular encounter rather than to profile physicians across multiple encounters with many patients. The Health Care Financing Administration is currently experimenting with the Uniform Clinical Data Set, which systematically abstracts and screens data from hospital records.

Strides are being made in developing the standards with which physicians' practices can be compared. Practice guidelines provide recommendations for physicians about the appropriate use of medical services. They are based on knowledge of the effectiveness of medical practices and procedures.

During a short period of time, the leadership of the medical profession has gone from criticizing guidelines as "cookbook medicine" to embracing them as the core of public and private efforts to improve the quality of care and to contain costs. Extensive efforts by the Agency for Health Care Policy and Research, physician specialty societies, and health insurers are currently underway to develop guidelines that are based on both the research literature and clinical judgment. Approximately 1,000 guidelines exist at present, though many of these are not specific enough or not grounded enough in the scientific literature to be useful.

Over the next few years, more and better practice guidelines are likely to be developed. Additional knowledge about the effectiveness of various procedures and increased experience with methods of guideline development will likely make guidelines more relevant to more of medical practice and perhaps more proscriptive.

Despite the early stage in the development of physician profiling, evidence is accumulating that it can increase the quality of care. For years, the Maine Medical Assessment Foundation has profiled individual physicians, informed them about how they compare with their peers, and convened discussions about variations from optimal practice. Despite the lack of formal practice guidelines, important changes in practice patterns have been documented. In addition, a recent review of published studies of physician profiling found that 80 percent documented improved performance by physicians.


Using Profiling Technology to Improve Quality

Both private entities and state licensing boards theoretically have the ability to profile physicians, evaluate their quality of care, attempt to improve the quality of their care through feedback, and select which physicians should continue to practice. Private entities are likely to make much more extensive use of those tools, however.

Among private entities, hospitals and HMOs have made the most extensive use of profiling techniques to assess physicians' quality. Both have access to clinical data and can examine complication rates associated with particular procedures. They can compare patterns of care with various practice guidelines. To date, however, such activities have been limited.

Profiling data can be used in a number of ways. They can be used for educational purposes to improve the quality of care physicians deliver. They also can be used as a screen to initiate consideration of steps to restrict the scope of a physician's practice in an HMO or a hospital.

Private insurers' use of profiling is unlikely to be as effective. Since they have a less extensive relationship with physicians, using the results of profiling for educational purposes is less likely to be an option. In addition, with claims data on only a small portion of a physician's practice and minimal clinical data on them, insurers are at a disadvantage in using profiling to assess a physician's quality of care.

Potential developments in data collection will, however, increase the effectiveness of profiling by insurers. Some policymakers envision government's sponsoring the development of all-patient data bases, which would enable insurers to examine data on a physician's entire practice. Following the example of Medicare's Common Working File, insurers would submit standardized information from all claims to a host carrier, which in turn would provide insurers limited access to those files. The Hartford Foundation's demonstrations plan to develop such data to support local organizations' convening physicians to discuss differences in practices to improve the quality of care.

State licensing boards could potentially use profiling in two ways. Should all-patient data bases be developed, boards could supplement the process of complaints by searching the data base for physicians whose pattern of practice appears to be substandard. Such screens of profiles could be used to trigger investigations. For example, surgeons who have unusually high operative mortality or complication rates could be targeted for extensive review. Alternatively, private quality improvement entities using such data could be required to alert licensing boards to physicians whose practice profiles suggest serious deficiencies in the quality of care.

State boards could also incorporate profiling into the license renewal process. If an all-patient data base were constructed, the profiles of physicians applying for relicensure could be reviewed. New York State is considering a proposal under which physicians would be required to demonstrate competence every nine years to renew their licenses. One option available to physicians to demonstrate competence would involve opening their medical records for the construction of practice profiles and comparison of those profiles with state-endorsed practice guidelines.


The Future of Licensure

Given the various trends discussed above, what will the role of physician licensure be in the future? We expect that licensure will play a relatively diminished role in quality assurance, but that an important role will remain for the foreseeable future.

Profiling of individual physicians and the development and dissemination of practice guidelines will become the most significant activities that govern the quality of care. The distinction between their use by private entities and by licensing boards is that the former will focus most on distinctions within the mainstream of physicians while the latter will focus on physicians who are clear outliers. Private entities will be able to engage in education and to make judgment calls concerning hospital privileges and inclusion in a physician network. The boards will need more objective and clear evidence of incompetence to discipline physicians. When the activities of the private entities and the boards are combined, boards may substantially improve the quality of care.

Despite the advances in quality that private entities will be able to promote, the boards will still have important roles to play. Those activities are unlikely to play a significant role in the initial granting of the privilege to practice medicine. To make a judgment on a practitioner's competence through profiling, a substantial amount of practice experience must be analyzed. Thus, dispensation with educational requirements would mean a substantial period in which the practitioner would provide care before a judgment on competence could be made. One could argue that increased ability to assess practice means that initial requirements could be relaxed, but patients would probably be exposed to more risks than with current entry requirements.

In addition, the work by private entities is unlikely to be sufficient to protect the public from physicians who are impaired or who practice incompetently. Currently, much of the public is not enrolled in health plans that have the ability to profile physicians and include only those who measure up to some quality standard. Thus, a physician whose competence is judged substandard by payers making such assessments may have extensive opportunities to practice poorly on patients not enrolled in such health plans. It is difficult to project whether the health system will evolve to the point where all persons will be enrolled in health plans that screen physicians on the basis of quality.

Finally, the market processes outlined above may be too slow in dealing with physicians who are seriously impaired. If the market continues to be structured so that each physician deals with numerous payers, it might require a lengthy process before every purchaser determines that the physician ought to be excluded. Despite due process requirements, the state board might be able to bar impaired physicians from practice faster than the market processes can.

In conclusion, we expect that the public will continue to want physician licensure. What will change is that licensure will become even more than today the aspect of quality assurance that deals with impairment or gross incompetence. Private entities will perform the mainstream of quality assurance. They will become much more effective and pervasive as technology permits increasingly sophisticated profiling of physicians to determine the quality of care.

The federal government is likely to play an extensive role in supporting private efforts to improve quality. Already, the federal government has taken a major role in guiding and supporting various efforts to develop practice guidelines. As a major insurer (Medicare), it has taken the lead in developing administrative and clinical data bases that will permit better assessments of the quality of care. The federal government is beginning to confront issues of standardization and access to private insurer data that will open additional opportunities for private monitoring of quality.

Selected Readings

Langsley, D.G. "Medical Competence and Performance Assessment: A New Era." Journal of the American Medical Association, Vol. 266 (1991).

Office of Technology Assessment. The Quality of Medical Care: Information for Consumers. Washington, D.C.: 1988.

Reiser, S.J. "Consumer Competence and the Reform of American Health Care." Journal of the American Medical Association, Vol. 267 (1992).

Svorny, S. "Should We Reconsider Licensing Physicians?" Contemporary Policy Issues, Vol. 10 (1992).

Regulation is published four times a year by the Cato Institute. Editorial and business offices are located at 1000 Massachusetts Avenue, N.W., Washington, D.C., 20001. For subscription information, please write to Circulation Department, Cato Institute, same address, or call (202) 842-0200. Send email inquiries to, or subscribe online via the World Wide Web at:

| Regulation | Home | Order Regulation | Publications | Search