Physician performance failures are not rare and pose substantial threats to patient welfare and safety. Few hospitals respond to such failures promptly or effectively. Failure to ensure the quality and safety of the performance of colleagues is a breach of medicine’s fiduciary responsibility to the public. A major reason for this deficiency is the hospitals’ lack of formal systems to monitor physician performance and to identify and correct shortcomings. To develop and implement these systems, hospitals need better performance measures and substantial expansion of external programs for assessment and remediation. This is a task well beyond the capacities of individual hospitals; a national effort is required. The authors call on the Federation of State Medical Boards, the American Board of Medical Specialties and The Joint Commission (organizations that already bear a fiduciary responsibility for ensuring safe, competent care) to collaborate on developing better methods for measuring performance and to expand programs for helping practitioners who are deficient.
To facilitate discussion of performance deficiencies that threaten patient safety, we must first define some key terms that are used to categorize professional behaviors. Professional competence has been defined as “the habitual and judicious use of communication, knowledge, technical skills, clinical reasoning, emotions, values and reflection in daily practice for the benefit of the individual and community being served”.1 The Accreditation Council for Graduate Medical Education and the American Board of Medical Specialties have divided competence into defined sets of “competencies” in specific domains, including those that apply to all physicians and those that are unique to each specialty. A deficiency in any of these domains can be referred to as a “dyscompetency,” which is a useful concept because no one is totally incompetent.
Mental and behavioral problems include depression, anxiety, substance abuse, personality disorders (for example, antisocial behavior) and disruptive behavior with colleagues, patients and subordinates. At the extreme are physicians who have severe psychopathologic manifestations, such as psychosis or suicidal behavior, but such cases are rare.
The term disruptive physician has been applied to physicians who exhibit abusive behavior that “interferes with patient care or could reasonably be expected to interfere with the process of delivering quality care”.2 Examples of disruptive behavior are provided in Table 1.3
Disruptive, intimidating, or abusive behavior may increase the likelihood of errors by leading nurses, residents or colleagues to avoid the disruptive physician, to hesitate to ask for help or clarification of orders, and to hesitate to make suggestions about patient care.4,5 Such behavior may also deflect the physician’s attention from the patient, thereby impairing clinical judgment and performance. When patients witness disruptive behavior, it undermines their confidence in the physician and the institution, as well as their willingness to partner in their own care.6 Consequently, disruptive behavior by physicians not only threatens patient safety, but also has a corrosive effect on morale, making life miserable for the nurses and residents who work closely with these physicians.
The term impaired has been defined by the American Medical Association as disability resulting from psychiatric illness, alcoholism or drug dependence. However, the term has sometimes been inappropriately applied to physicians who have returned to good health, are substance-free and in a monitoring program, or have successfully completed a knowledge or skill remediation course.
We will use the term performance problems to refer to all types of deficiencies, regardless of cause.
Physician performance problems can be usefully thought of as symptoms of underlying disorders, not as diseases. Underlying causes include mental and behavioral problems, including substance abuse or dependence (drugs or alcohol); physical illness, including age-related and disease- related cognitive impairment; and failure to maintain or acquire knowledge and skills. Contributing stressors include overwork, family strife, a dysfunctional working environment, supervisor pressure and anxiety. Categories frequently overlap. For example, declining surgical competence can be attributed to knowledge or skill deficits and to alcohol dependence, and both of these problems may reflect underlying mental illness, such as severe depression.
Contributing to these problems are fatigue, stress, isolation and easy access to drugs. The “normal” stress of medical practice has been compounded in recent years by large educational debt loads for graduating physicians, increasing malpractice premiums, decreasing reimbursement and the pressure to see more patients in a shorter amount of time. Stress can lead to isolation and cause physicians to acquire maladaptive coping strategies, including alcohol or drug abuse.
In our experience, the professional realm is usually the last area in one’s life that is affected by substance abuse and mental and behavioral issues.7 By the time these disorders manifest in the workplace, the physician’s relationships with significant others, nuclear family, extended family, friends and community have usually been “impaired” for a long time.
EXTENT OF THE PROBLEM
The media frequently cite the number of physicians disciplined by state medical boards as a measure of performance issues. In 2002, approximately 0.5 percent of practicing physicians in the United States were disciplined; 1739 physicians had their licenses revoked and state boards imposed restrictions on an additional 1218.8 It is difficult to know how to interpret these figures because physicians are disciplined for various reasons, some of which may be unrelated to performance (such as fraudulent activities involving third-party payers).
Concerning mental illness, a recent study found a 16 percent lifetime incidence of major depressive disorders in the general population.9 The rate in physicians may be even higher10 ; for example, the rate of suicide is 40 percent higher in male physicians and more than twofold higher in female physicians than in the general population.11
Estimates of alcohol dependence vary from eight percent to 15 percent,12–16 the latter being similar to the 13.5 percent rate for the adult population.16–18 The American Medical Association’s estimate for drug dependence is one percent to two percent.18 The Medical Board of California estimated that 18 percent of physicians in its state abuse alcohol or other drugs at some point during their career.19
Sound data are lacking for the incidence of disruptive behavior. Surveys of nurses suggest that most have witnessed episodes caused by four to five percent of the physicians at their institutions, but these data are flawed by low response rates.5–20 Surveys of physician executives indicate that the percentage of disruptive physicians ranges from one percent to five percent.21 Hickson and colleagues6 found that six percent of physicians received 25 or more complaints from patients over a six-year period. Our best estimate is that three to five percent of physicians present a problem of disruptive behavior.
We found no studies of the incidence of physical illness among practicing physicians, but a reasonable estimate is that at least 10 percent of physicians must restrict their practice for several months or more during their career because of a disabling physical illness (such as diabetes, heart disease or surgical procedures). Like everyone else, physicians are subject to cognitive decline with aging,22–24 but the extent has not been quantified.
Similarly, there are no overall estimates of the extent of knowledge and skill dyscompetencies. Results from one measure, recertification examinations, show that first-time failure rates in four specialties ranged from one percent to 14 percent (Table 2). Failure rates tend to be higher on subsequent examinations. We estimate that as many as 10 percent of physicians will demonstrate significant deficiencies in knowledge or skills at some point in their career.
When all conditions are considered, at least one-third of all physicians will experience, at some time in their career, a period during which they have a condition that impairs their ability to practice medicine safely; for a hospital with a staff of 100 physicians, this translates to an average of one to two physicians per year. Referral rates to state physician health programs suggest that most practitioners get little help. On the basis of our experience, even serious problems are often handled poorly at the hospital or practice level. However, ensuring high standards of professional conduct is arguably the greatest responsibility of a professional and one that the public, lacking an alternative mechanism for oversight, has a right to expect. We believe that our profession’s failure to ensure the quality and safety of our colleagues’ performance is a breach of its fiduciary obligation to the public.
AN INEFFECTIVE SYSTEM
Neither physicians nor hospitals have adequately addressed performance problems.3 Few organizations systematically monitor physician performance or have formal programs to identify problem doctors.3 Annual physical examinations are not required of physicians, and only the Department of Veterans Affairs25 performs random drug testing. State licensing boards have relied on continuing education attendance as evidence of maintenance of competence.
Once problems are identified, management is also frequently haphazard. In egregious cases, investigations have repeatedly revealed that institutions ignored numerous warning signs months or years before a serious incident occurred.26,27 Many physicians are reluctant to confront behavioral or competence problems. Independence is so highly valued that physicians are loath to evaluate or confront a colleague whom they perceive as having a problem.28,29 Doctors abhor making judgments about colleagues who may also be personal friends or practice partners. Department chairs often lack the training and skills needed for managing doctors who perform poorly. The hospital may need the physician’s revenue stream.30
As a result, managing these situations can be difficult and aggravating for all parties concerned. Offers of assistance may be spurned. If disciplinary action is needed because a physician’s performance is unsafe, it can be met with countercharges or a lawsuit, even when evidence is clear and due process is followed. At best, management is often a messy business; at worst, it can be hazardous to everyone involved.
Hospitals receive little help from regulators. Although they are required to have credentialing and disciplinary processes, the details of implementing such processes are left to the hospitals. There are few national or state standards of conduct or competence, or measures for monitoring performance.
State medical boards discipline physicians after the fact when unsatisfactory performance is reported by hospitals or patients or when malpractice settlements are reviewed. However, the state boards typically do not define prevention of injury as part of their responsibility.
Health care’s casual approach to monitoring physician performance contrasts markedly to that of other professions whose conduct affects the public welfare. Commercial pilots, for example, must pass both physical and performance examinations every year.
The challenge is clear: We need to identify problem doctors early and address the problems in a timely fashion. To do this, we require better measures for identifying physicians who need help and better programs for providing help to those who need it. Although performance problems are widespread, we suggest that the place to start is in hospitals, where a credentialing process is already in place.
IDENTIFYING PHYSICIANS WHOSE PERFORMANCE MAY ENDANGER PATIENTS
We propose that the current ad hoc, informal, reactive approach to physician performance problems be replaced with a routine, formal, proactive system of monitoring that uses validated measures to focus strictly on clinical and behavioral performance. The goal would be to identify problem doctors early, before they jeopardize patient safety.
This system would have three essential characteristics. First, it should be objective. A common criticism of current methods is that they are based on subjective judgments of personality, motivation or character instead of performance. The solution is to base evaluations on data, such as evidence of compliance with performance standards.
Second, the system should be fair. To avoid being viewed as stigmatizing or punitive, all physicians should be evaluated on an annual basis according to the same measures. The evaluation process must be open and unbiased, and it must comply with labor regulations.
Third, the system should be responsive. When physicians with problems are identified, they must be treated promptly. For some physicians, feedback and internal counseling may be all that is necessary. Others may need further assessment and referral to a program to help them to correct their deficiencies and enable them to continue to practice medicine, if possible.
A MODEL SYSTEM
How would such a system operate? We envision four stages to the process: adopting standards, requiring compliance, monitoring performance and responding to deficiencies.
First, an institution should adopt explicit performance standards of behavior and competence. These standards need to be developed at the national level; specialty boards are currently at work on such standards for competence. The Federation of State Medical Boards could develop behavioral standards by using currently available material. The Joint Commission should coordinate with these groups. During the interim, hospitals can develop their own performance standards (as many already have). These should address all aspects of professional behavior. For example, one standard might be, “All patients and personnel will be treated with respect.”
Second, all physicians should be required to acknowledge that they 1) have read and understand the standards, 2) have a responsibility to follow the standards, 3) know that adherence will be monitored, and 4) understand that persistent failure will lead to loss of privileges and dismissal. This acknowledgment should be given in writing as a condition of being granted clinical privileges.
Third, adherence to standards would be monitored annually by formal evaluations of all members of the staff using accepted and validated measures of competence and behavior. These should include confidential evaluations by colleagues and coworkers and analysis of complaints by patients or others.6,31
Fourth, results of the evaluations should be provided confidentially to each individual (with the identities of their colleagues concealed for protection). If significant deficiencies are identified, the department chairman should be responsible for ensuring a prompt, appropriate response. This response could involve evaluative testing, counseling, or referral for further assessment and treatment. In cases that threaten patient welfare, department chiefs and hospital leaders must take immediate action to limit practice during assessment and rehabilitation.
An essential element of this system is that everyone clearly understands his or her roles and responsibilities when a practitioner with performance issues has been identified; from the outset, all parties will know who is responsible for collecting data, who should receive reports and what actions are required by whom at each level. Such a system would provide accountability at all levels: physician to department chair, chair to hospital medical staff, medical staff to hospital board and hospital to state boards and The Joint Commission.
Finally, assessment and treatment programs must be available for management of all underlying causes of substandard performance: substance abuse, psychiatric problems, behavioral problems and dyscompetencies. Programs should be personalized to enable the individual to use his or her strengths and knowledge in productive ways, ultimately resuming practice if possible. If a physician refuses to accept education, treatment, monitoring, or necessary restrictions of practice, or if these interventions fail, the physician must be promptly referred to the state medical board for disposition.
We conclude that an effective system for managing physicians with performance issues is built on the ideas that 1) subpar performance can be objectively defined; 2) routine monitoring of all members of the medical staff is necessary to detect problems fairly and early; and 3) the responses to deficiencies should be prompt, constructive and sustained. The long-term objective is to enable physicians to continue to practice effectively and safely—not to “weed them out.” If the system works properly, that is, if physicians who perform poorly are identified before serious consequences arise, then one might expect referrals to the state medical board for disciplinary action to decrease.
Implementing such a system requires a national effort on several fronts. Three issues must be addressed: developing better measures for assessing performance, expanding the number of assessment programs for physicians with competence or behavioral problems, and developing and supervising remediation programs.
Some authorities have contended that performance assessment is not feasible.31 We disagree. However, the obstacles for setting assessment standards for behavioral problems differ from those for dyscompetencies.
For behavioral problems, several measures could be combined to provide earlier identification of physicians who need help. For example, Hickson and colleagues6 found that analysis of patient complaints can identify doctors with interpersonal problems and predict the likelihood of malpractice litigation. Physicians with four or more complaints over a six-year period were found to be 16 times more likely to have two or more risk management files opened than were physicians with no complaints.
The Physicians Achievement Review program, which is run by the College of Physicians and Surgeons of Alberta, assesses every physician in Alberta every five years.32 Patients, physician colleagues and nonphysician coworkers complete confidential questionnaires regarding the individual’s clinical knowledge and skills, communication skills, psychosocial management, office management and collegiality. Physicians are provided with detailed aggregate responses for their own practice and a report comparing their personal results with the summary profile of all physicians with similar types of practices. Surveys may also be suitable for identifying physicians with alcohol or drug dependence, physical impairment and mental illness.33 These measures need to be validated in the clinical setting.
Serious consideration should be given to implementing annual physical examinations and random drug testing for all physicians. More controversial, but clearly in need of investigation, is the feasibility of routine cognitive evaluations for older physicians.
Competence is a more difficult area to measure. Myriad instruments have been advanced over the years to measure competence, but most have proven too cumbersome or expensive to be implemented for all physicians on a regular basis. However, the Accreditation Council for Graduate Medical Education and the members of the American Board of Medical Specialties have advanced a massive national effort to define general and specialty-specific competencies. They have started to develop measures of these competencies to assess trainees for certification,34 and these same measures will be used to assess practicing physicians as part of maintenance of certification.
These tools could also be used by hospitals. Several specialty organizations, particularly the American Board of Internal Medicine, have already made substantial progress in developing measures35 (Cassel C. Personal communication, Dec. 18, 2005). These assessments might be financed through a combination of user fees and support from the national boards and specialty societies.
Expanding Assessment Programs
A second serious challenge is the need to greatly expand the number and capacity of assessment programs for physicians with competence or behavioral problems. All states except Nebraska have physician health programs for doctors with alcoholism or drug abuse, and 41 states offer assistance for physicians coping with mental health issues; however, few state programs address knowledge and skill deficits, clinical dyscompetencies or disruptive behavior (Table 3).
Nationwide, only 10 programs are available for assessing physicians’ skills or for education plans to correct deficiencies (Table 4). Five programs assess disruptive behavior. Others deal with relationship issues, communication skills, medical skills and knowledge, clinical reasoning and patient care documentation.
The Federation of State Medical Boards administers a standardized examination of clinical knowledge, the Special Purpose Examination, to physicians referred by state medical boards or by themselves. If results are questionable, the physician may undergo an additional assessment by The Institute for Physician Evaluation, which is a joint initiative of the Federation and the National Board of Medical Examiners. The Institute’s assessments include computer-based case simulations, structured interviews, multiple-choice examinations, cognitive function screening and interactive judgment analysis. The evaluations were previously administered in Philadelphia, Pa., and Dallas, Texas, but are now offered exclusively in cooperation with physician enhancement services at five locations: the Physician Assessment and Clinical Education Program (PACE) at the University of California, San Diego; the Clinical Competency and Assessment Training Program (CCAT) at Rush University Medical Center in Chicago, Ill.; the Upstate New York Clinical Competency Center at Albany Medical College in Albany, N.Y.; the Florida Competency Advancement Program (CAP) at the University of Florida in Gainesville, Fla.; and the Physician Assessment Program at the University of Wisconsin School of Medicine and Public Health, Madison, Wis.36 These collaborators use the Institute’s tools to assess competence, then strive to remedy the identified deficiencies.37
Developing Remediation Programs
The final challenge is to use the assessment results to construct successful remediation programs for those with skill deficits. The obstacles to progress in this area are substantial. The first obstacle is a lack of expertise to oversee such programs. Few national programs exist, and hospital level programs are often poorly organized. A major national effort is needed to develop additional programs. A second barrier is inadequate financing. If physicians who will already be losing practice income are responsible for the full cost, they may be unwilling to participate. Other potential sources of support include specialty societies, licensing boards, federal or state governments and liability insurers.
Yet another barrier is the reluctance of hospitals and physician colleagues to voluntarily guide, mentor and supervise remediation activities. Department chairs rarely have formal supervisory training and often lack the experience needed to effectively manage physicians with performance problems. In addition to time, cost and liability concerns, many institutions are uncomfortable asking patients to give informed consent for physicians with acknowledged deficits to care for them. Effective models need to be developed and tested.
Medical schools and their affiliated teaching hospitals should assume a leadership role in developing supervised clinical programs for physicians who have been found to have remediable knowledge and skill deficits. If developed and coordinated like traditional postgraduate training programs, these “mini-residencies” could effectively overcome the time, liability and informed consent issues.
A CALL TO ACTION
The responsibility for monitoring and ensuring acceptable physician performance must occur at the local level. However, hospitals lack the resources to develop the systems and measures that are needed. We believe the time has come for a major national effort to develop these systems and measures. The organizations that are best positioned to take on this task are those that already bear a fiduciary responsibility for ensuring safe, competent care: the state medical boards (represented by the Federation of State Medical Boards), the medical specialty boards (represented by the American Board of Medical Specialties) and The Joint Commission.
We call on these national organizations to collaborate in an effort to accomplish three goals: develop standards and measures for annual data-based assessment of physician performance and require that they be implemented by all hospitals, launch a major effort to develop better measures of competence and behavior and develop more state and regional centers for assessment and remediation of physicians with performance deficiencies. The initial specifications should be based on currently available data and should then be improved as better data become available. Demonstration of an individual’s compliance with these national standardized measures should be sufficient to satisfy state relicensing and certification requirements.
Performance failures of one type or another are not uncommon among physicians, posing substantial threats to patient welfare and safety. Few hospitals manage these situations promptly or well. It is time for a national effort to develop better methods for assessing performance and better programs for helping those who are deficient.
From the Harvard School of Public Health and Harvard Medical School, Boston, Mass.
Reprinted with permission from pages 107-115 of the Volume 144, Number 2, issue of the Annals of Internal Medicine. Published Jan. 17, 2006.
POTENTIAL FINANCIAL CONFLICTS OF INTEREST:
Grants received: L.L. Leape (Robert Wood Johnson Foundation).